PONTIFICIA UNIVERSIDAD CATOLICA DE CHILE INFLUENCIA DE …
134
1 PONTIFICIA UNIVERSIDAD CATOLICA DE CHILE Facultad de Ciencias Biológicas Programa de Doctorado en Ciencias Biológicas Mención Ecología INFLUENCIA DE LOS CAMBIOS DE USO/COBERTURA DEL SUELO Y EL CLIMA EN EL CICLO DEL NITRÓGENO EN DOS LAGOS COSTEROS DE CHILE CENTRAL A PARTIR DE LA COLONIZACIÓN ESPAÑOLA Tesis entregada a la Pontificia Universidad Católica de Chile en cumplimiento parcial de los requisitos para optar al Grado de Doctor en Ciencias con mención en Ecología Por: MARÍA MAGDALENA FUENTEALBA LANDEROS Director de Tesis : Dr. Claudio Latorre Hidalgo Co-director : Dr. Blas Valero-Garcés Junio, 2019
Transcript of PONTIFICIA UNIVERSIDAD CATOLICA DE CHILE INFLUENCIA DE …
1
PONTIFICIA UNIVERSIDAD CATOLICA DE CHILE
Facultad de Ciencias Bioloacutegicas
Programa de Doctorado en Ciencias Bioloacutegicas Mencioacuten Ecologiacutea
INFLUENCIA DE LOS CAMBIOS DE USOCOBERTURA DEL SUELO Y EL
CLIMA EN EL CICLO DEL NITROacuteGENO EN DOS LAGOS COSTEROS DE
CHILE CENTRAL A PARTIR DE LA COLONIZACIOacuteN ESPANtildeOLA
Tesis entregada a la Pontificia Universidad Catoacutelica de Chile en cumplimiento
parcial de los requisitos para optar al Grado de Doctor en Ciencias con mencioacuten
en Ecologiacutea
Por MARIacuteA MAGDALENA FUENTEALBA LANDEROS
Director de Tesis Dr Claudio Latorre Hidalgo
Co-director Dr Blas Valero-Garceacutes
Junio 2019
2
LA DEFENSA FINAL DE LA TESIS DOCTORAL TITULADA
ldquoINFLUENCIA DE LOS CAMBIOS DE USOCOBERTURA DEL SUELO Y EL CLIMA EN EL CICLO DEL NITROacuteGENO EN DOS LAGOS COSTEROS DE CHILE CENTRAL A
PARTIR DE LA COLONIZACIOacuteN ESPANtildeOLArdquo
Presentada por la Candidata a Doctor en Ciencias Bioloacutegicas Mencioacuten Ecologiacutea de la Pontificia Universidad Catoacutelica de Chile
SRA MARIacuteA MAGDALENA FUENTEALBA LANDEROS
Ha sido aprobada por el Tribunal Examinador constituido por los profesores
abajo firmantes calificaacutendose el trabajo realizado el manuscrito sometido
y la defensa oral con nota _________ (_______________________________)
DR JOSEacute MIGUEL FARINtildeA R Coordinador Comiteacute de Tesis
Facultad de Ciencias Bioloacutegicas-UC
DR JUAN A CORREA M Decano
Facultad de Ciencias Bioloacutegicas-UC
DR BLAS VALERO G Co-Director de Tesis
Consejo Superior de Investigaciones Cientiacuteficas
DR CLAUDIO LATORRE H Director de Tesis
Facultad de Ciencias Bioloacutegicas-UC
DR JUAN ARMESTO Z Miembro del Comiteacute de Tesis
Facultad de Ciencias Bioloacutegicas-UC
DR RICARDO DE POL H Profesor Invitado
Universidad de Magallanes
Santiago de Chile 30 de septiembre de 2019-
3
TABLA DE CONTENIDO
RESUMEN 9
ABSTRACT 11
INTRODUCCIOacuteN 13
Los ecosistemas mediterraacuteneos y el ciclo del N 15
Los lagos como sensores ambientales 16
El ciclo del N en lagos 18
Reconstruyendo el ciclo del N a partir de variaciones en δ15N 20
Referencias 25
CAPIacuteTULO 1 A COMBINED APPROACH TO ESTABLISHING THE TIMING AND MAGNITUDE OF ANTHROPOGENIC NUTRIENT ALTERATION IN A MEDITERRANEAN COASTAL LAKE- WATERSHED SYSTEM 29
Abstract 30 1 INTRODUCTION 31 2 STUDY SITE 35 3 RESULTS 38
31 Age Model 38 32 The sediment sequence 39 33 Sedimentary units 41 34 Isotopic signatures 42 35 Recent land use changes in the Laguna Matanzas watershed 44
4 DISCUSSION 45 41 N and C dynamics in Laguna Matanzas 45 42 Recent evolution of the Laguna Matanzas watershed 48
5 CONCLUSIONS 54 6 METHODS 55
References 58
CAPIacuteTULO 2 STABLE ISOTOPES TRACK LAND USE AND COVER CHANGES IN A MEDITERRANEAN LAKE IN CENTRAL CHILE OVER THE LAST 600 YEARS 71
Abstract 73 1 Introduction 73
4
2 Study Site 77 3 Methods 79 4 Results 84
41 Geochemistry and PCA analysis 84 42 Sedimentary units 86 43 Recent seasonal changes of particulate organic matter on water column 88 44 Stable isotope values across the Lake Vichuqueacuten watershed 89 45 Land use and cover change from 1975 to 2014 90
5 Discussion 93 51 Seasonal variability of POM in the water column 93 52 Stable isotope signatures in the Lake Vichuqueacuten watershed 95 53 Recently land use and cover change and its influences on N inputs to the lake 97 54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years 98
6 Conclusions 101
LITERATURE CITED 110
DISCUSION GENERAL 117
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central 119
CONCLUSIONES GENERALES 130
Referencias 133
5
A mis padres Arturo y Malena
A mis hijos Xavi y Panchito
6
AGRADECIMIENTOS
Quiero agradecer a mi tutor y mentor Dr Claudio Latorre por brindarme su
apoyo sin el cual no habriacutea logrado concluir esta tesis de doctorado Claudio tu
apoyo constante incentivo y el fijarme metas que a veces me pareciacutean imposibles
de alcanzar no solo han dado forma a esta tesis sino tambieacuten me ha hecho maacutes
exigente como cientiacutefica Claudio destacas no solo por ser un gran cientiacutefico si no
tambieacuten por tu gran calidad humana eres un gran ejemplo
Quiero agradecer Dr Blas Valero-Garceacutes por nuestras numerosas
conversaciones viacutea Skype que incluiacutean vacaciones y fines de semana para discutir
los resultados de la tesis y que han dado forma a esta investigacioacuten principalmente
al primer capiacutetulo Ademaacutes por haberme acogido como un miembro maacutes en el
laboratorio de Paleoambientes Cuaternarios durante las estancias que he realizado
en el transcurso de estos antildeos Blas eres un ejemplo para miacute conjugas ciencia de
calidad calidez y dedicacioacuten por tus estudiantes
A quienes han financiado mi doctorado la Comisioacuten Nacional de
Investigacioacuten Cientiacutefica y Tecnoloacutegica (CONICYT) con sus becas de manutencioacuten
doctoral (2013) gastos operacionales pasantiacutea (2016) postnatal (2017) y termino
de tesis doctoralrdquo (2013) A FONDECYT a traveacutes del proyecto 1160744 de C
Santoro Al Departamento de InvestigacioacutenAl Instituto de Ecologiacutea y Biodiversidad
(IEB) a traveacutes de del PIA financiamiento basal 170008 la Pontificia Universidad
Catoacutelica de Chile por la beca incentivo para tesis interdisciplinaria para doctorandos
(2015)
Agradezco a mis compantildeeros del laboratorio de Paleoecologiacutea y
Paleoclimatologiacutea Karla Matias Dani Carolina Mauricio y Pancho que han hecho
grato mi tiempo en el laboratorio Agradecimientos especiales a Carolina Matiacuteas y
Leo por acompantildearme a terreno
7
Agradezco a los miembros del laboratorio de Paleoambientes Cuaternarios
(IPE) en especial a Raquel Elena Fernando Ana Penelope Graciela Miguel
Sevilla Mariacutea y Miguel Bartolomeacute
Agradezco a los miembros de la comisioacuten Dr Joseacute Miguel Farintildea Dr Juan
Armesto y Dr Ricardo De Pol-Holz por la gran ayuda durante el desarrollo de mi
doctorado en especial por las correcciones finales de la tesis
Al departamento de Ecologiacutea en especial a Rosario Galaz Edita Luengo
Daniela Mora y Valeria Cavallero por su apoyo
A mis compantildeeros de doctorado Beleacuten Gallardo Pancho Diacuteaz Bryan Bularz
Bian Latorre Renato Borras Juan Carlos Hernaacutendez y Paulina Troncoso con
quienes estudieacute aprendiacute y compartimos muy buenos momentos durante los
primeros antildeos del doctorado
A Pablo Sarricolea mi compantildeero de vida Gracias por tu constante e
incondicional apoyo Sin ti durante este proceso todo habriacutea sido cuesta arriba
A mis padres Arturo y Malena gracias por su carintildeo apoyo y compromiso
Papaacute aunque no esteacutes a mi lado siempre te recuerdo con mucho carintildeo A mi
madre que ha sido un constante apoyo sobre todo en el cuidado de mis hijos xavi
y panchito
A mis hermanos Rodrigo y David por estar presentes durante toda esta
etapa Siempre con carintildeo y hermandad
A Rodrigo David Matiacuteas Jany Paola y Julio por cuidar a mis hijos con carintildeo
siempre que estuve ausente por el doctorado
8
ABREVIATURAS
N Nitroacutegeno (Nitrogen)
DIN Nitroacutegeno Inorgaacutenico Disuelto (Dissolved Inorganic Nitrogen)
C Carbono (Carbon)
TOC Carbono Inorgaacutenico Total (Total Organic Carbon)
TIC Carbono Inorgaacutenico Total (Total inorganic Carbon)
TC Carbono Total (Total Carbon)
TS Azufre Total (Total Sulfur)
LUCC Cambios de UsoCobertura del Suelo (Land Use and Cover Change)
OM Materia Orgaacutenica (Organic Matter)
POM Particulate Organic Matter (materia orgaacutenica particulada)
CE Common Era
BCE Before Common Era
Cal BP Calibrado en antildeos radiocarbono antes de 1950
ie id est (esto es)
e g Exempli gratia (por ejemplo)
9
RESUMEN
El ldquoAntropocenordquo se caracteriza por pertubaciones de origen humano que
conllevan impactos globales Por ejemplo los cambios de usocobertura del suelo
(LUCC) son conocidos por perturbar el ciclo del N a partir de la Revolucioacuten Industrial
pero especialmente desde la Gran Aceleracioacuten (1950 CE) en adelante Sin
embargo existen incertezas asociadas a la magnitud del impacto y su efecto
acoplado con los cambios climaacuteticos actuales (ie megasequiacutea disminucioacuten de las
precipitaciones) y acoplamientos con otros ciclos biogeoquiacutemicos (eg ciclo del
Carbono) Este impacto ha cambiado la disponibilidad de N en los ecosistemas
terrestres y acuaacuteticos y sus enlaces Los sedimentos lacustres contienen
informacioacuten de las condiciones paleoambientales del lago y su cuenca en el
momento en que se depositaron y el anaacutelisis de los isoacutetopos estables de N (δ15N)
en los sedimentos permiten reconstruir los cambios en la disponibilidad de N a
traveacutes del tiempo En este trabajo usamos una aproximacioacuten multiproxy que incluye
10
anaacutelisis sedimentoloacutegicos geoquiacutemicos e isotoacutepicos (δ15N δ13C) de los sedimentos
lacustres columna de agua y suelo-vegetacioacuten de la cuenca y la reconstruccioacuten de
los LUCC entre 1975 y 2016 a partir de imaacutegenes satelitales El objetivo de esta
tesis fue evaluar el rol de LUCC como principal control del ciclo del N en el sistema
cuenca-lago durante las uacuteltimas centurias en Chile central Nuestros principales
resultados muestran que valores maacutes positivos de δ15N en los sedimentos lacustres
estaacuten relacionados con grandes aportes de N de la cuenca que a su vez son
mayores cuanto mayor la superficie agriacutecola yo los pastizales mientras que tanto
las plantaciones forestales como los bosques favorecen la retencioacuten de nutrientes
en la cuenca (lo que se ve reflejado en un δ15N maacutes negativo) Esta tesis plantea
un cambio de estado en la dinaacutemica del N asociado a la introduccioacuten y expansioacuten
de las plantaciones forestales o bosques los que retienen el Nitroacutegeno en las
cuencas mientras que el clima juega un rol secundario
11
ABSTRACT
The Anthropocene is characterized by human disturbances at the global
scale For example changes in land use are known to disturb the N cycle since the
industrial revolution but especially since the Great Acceleration (1950 CE) onwards
This impact has changed N availability in both terrestrial and aquatic ecosystems
However there are some important uncertainties associated with the extent of this
impact and how it is coupled to ongoing climate change (ie megadroughts rainfall
variability and shifting stationarity) or to other nutrient cycles (e g carbon cycle)
Lake sediments contain paleoenvironmental information regarding the conditions of
the watershed and associated lakes and which the respective sediments are
deposited Nitrogen stable isotope analysis (δ15N) on lake sediments allows us to
reconstruct the changes in N availability through time Here we used a multiproxy
approach that uses sedimentological geochemical and isotopic analyses on
lacustrine sediments water column and soilvegetation from the watershed as well
12
as land use and cover change (LUCC) from 1975 to 2016 obtained from satellite
images The goal of this thesis is to evaluate the role of LUCC as the main driver for
N cycling in a coastal watershed system of central Chile over the last centuries Our
main results show that more positive δ15N values in lake sediments are related to
higher N contributions from the watershed which in turn increase with increased
agricultural andor pasture cover whereas either forest plantations or native forests
can favor nutrient retention in the watershed (δ15N more negative) This thesis
proposes that N dynamics are mainly driven by the introduction and expansion of
forest or tree plantations that retain nitrogen in the watershed whereas climate plays
a secondary role
13
INTRODUCCIOacuteN
El N es un elemento esencial para la vida y limita la productividad en
ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al 1997) Las actividades
humanas han tenido un profundo impacto sobre el ciclo del N global principalmente
a partir la revolucioacuten industrial (IPCC 2013 2007) Las concentraciones de N se
han incrementado por la fijacioacuten industrial de N atmosfeacuterico (viacutea el proceso Haber-
Bosch Erisman et al 2008) por el uso de fertilizantes sobre la base de N para
mejorar el rendimiento de los cultivos la produccioacuten ganadera y ademaacutes de los
cambios en el uso del suelo (abreviado en ingleacutes- LUCC) (Robertson y Vitousek
2009 Galloway et al 2008 Battye et al 2017 Xu et al 2019) Estas actividades
contribuyen significativamente al incremento de la disponibilidad bioloacutegica de N
cuyas consecuencias para los ecosistemas incluye la perdida de diversidad
modificacioacuten de la disponibilidad de nutrientes y acidificacioacuten de los suelos entre
otras Sin embargo se desconoce la cantidad destino y el alcance que ha tenido
14
el N movilizado entre los ecosistemas generado por la influencia de las actividades
humanas (Vitousek et al 1997)
La entrada natural de N en los ecosistemas terrestres y acuaacuteticos es viacutea
fijacioacuten atmosfeacuterica la que es llevada a cabo por microorganismos muchos de ellos
en relaciones simbioacuteticas (eg Rhizobium en leguminosas) y las algas (Vitousek et
al 1997 Battye et al 2017) Las principales salidas estaacuten relacionadas con la
desnitrificacioacuten volatilizacioacuten del amoniaco y por escurrimiento superficial y
subterraacuteneo (Galloway et al 2008 Diacuteaz et al 2016) En el caso de los sistemas
lacustres ademaacutes estaacuten la resuspencioacuten de N del fondo del lago los aportes desde
la cuenca (entradas de N) La depositacioacuten de MO en el fondo del lago que marca
la salida de N de la columna de agua Estas relaciones de intercambio de N tienen
un control geograacutefico (eg latitud exposicioacuten relieve etc) climaacutetico y biotico
(McLauchlan et al 2013) La alteracioacuten provocada por los LUCC no solo genera
las peacuterdidas de nutrientes del suelo por erosioacuten y escurrimiento superficial sino que
tambieacuten modifica la disponibilidad y transferencia de N entre los ecosistemas
terrestres y acuaacuteticos (Elbert et al 2012 McLauchlan et al 2013) Por ejemplo el
reemplazo de la vegetacioacuten nativa por la agricultura yo plantaciones forestales
altera el ciclo del N debido al despeje de los terrenos viacutea quema de la vegetacioacuten
pero tambieacuten debido al reemplazo de la vegetacioacuten preexistente por un
monocultivo (eg trigo pino) y el uso ineficiente de fertilizantes para mejorar el
rendimiento agriacutecola Estas actividades favorecen un incremento de los aportes de
N (y otros nutrientes) viacutea sistemas de drenajes a lagos los que actuacutean como
sumideros El incremento del N derivado de las actividades humanas tanto en los
ecosistemas terrestres como acuaacuteticos genera incertezas relacionados con la
trayectoria y la magnitud de la disponibilidad de N en ambos ecosistemas (Batye et
15
al 2017 Mclauchlan et al 2017) Por lo tanto se requiere de una liacutenea base de
largo plazo para saber coacutemo estas perturbaciones impactan la disponibilidad de N
en las uacuteltimas deacutecadassiglos en los ecosistemas lacustres y por lo tanto el alcance
real que los LUCC han tenido en el ciclo del N
Los ecosistemas mediterraacuteneos y el ciclo del N
Los ecosistemas mediterraacuteneos son especialmente sensibles a los LUCC
pues estos se encuentran sometidos a un estreacutes hiacutedrico durante largas temporadas
estivales y las precipitaciones se concentran en eventos puntuales y a veces con
altos montos pluviomeacutetricos Las precipitaciones tienen un alto potencial de arrastre
de sedimentos y nutrientes los que actuacutean como fertilizantes al llegar a los
ecosistemas lacustres A su vez un incremento en la disponibilidad de N puede
generar un desbalance con otros nutrientes (eg Fosforo) con efectos sobre la
productividad y diversidad de los lagos (Pentildeuelas et al 2012 Sardans et al 2012
McLauchlan 2013) En este sentido en Chile central se observa una disminucioacuten
de las precipitaciones especialmente fuerte en las uacuteltimas deacutecadas por lo que se ha
denoacuteminado como Megasequiacutea (Garreaud et al 2017) y el efecto de las
precipitaciones esporaacutedicas pueden generar un arrastre de nutrientes que no ha
sido evaluado
Los ecosistemas mediterraacuteneos concentran el 20 de la biodiversidad global
(Myers et al 2000) pero existe una escasez de conocimiento respecto a los
efectos del incremento de N en los cuerpos de agua como consecuencia de las
actividades humanas en el pasado histoacuterico (Ochoa-Hueso et al 2011) y coacutemo la
disponibilidad de N ha variado en el tiempo Los LUCC han aumentado el flujo de
N en las cuencas hidrograacuteficas viacutea escurrimiento superficial y lixiviacioacuten
16
favoreciendo la peacuterdida de sedimentos y nutrientes del suelo en el mundo entero
(Elser et al 2011 McLauchlan et al 2013 Xu et al 2017 y 2019) Ello ha
contribuido a la acidificacioacuten y eutrofizacioacuten generalizada de riacuteos y lagos
(McLauchlan et al 2013 Schindler et al 2008)
El ecosistema mediterraacuteneo de Chile central es una regioacuten ampliamente
intervenida desde al menos la colonizacioacuten espantildeola (1600-1800 CE) Los LUCC
han tenido efectos negativos en la disponibilidad de agua especialmente
observados con el desarrollo de las actividades minera (Prieto et al 2019) Aunque
se sabe de los impactos en los ecosistemas de bosque en teacuterminos de cobertura
debidos al desarrollo silvo-agropecuarias (Armesto et al 2010) se desconoce el
impacto de estas actividades sobre el ciclo del N en los cuerpos de agua o en queacute
momento cobran fuerza suficiente para alterar el flujo de N hacia los lagos de Chile
Central Por lo tanto en esta tesis nos centramos en entender coacutemo los LUCC han
afectado la disponibilidad de N en los lagos costeros de Laguna Matanzas y Lago
Vichuqueacuten desde la llegada de los espantildeoles a Chile hasta el presente
Los lagos como sensores ambientales
Los sedimentos lacustres son buenos sensores de cambios en los aportes
de nutrientes contaminacioacuten y eutroficacioacuten de los cuerpos de agua ya que son
capaces de preservar sentildeales quiacutemicas y fiacutesicas al momento de su formacioacuten y
ademaacutes con alta resolucioacuten y a largo plazo (Leng et al 2006) Por lo tanto
constituyen una herramienta uacutetil para reconstruir el flujo de N entre ecosistemas
terrestres y acuaacuteticos en el pasado sus consecuencias (eg incremento de la
productividad lacustre) y el rol del clima y los LUCC en el pasado (McLauchlan et
al 2013 von Gunten et al 2009) Por ejemplo el cultivo de maiacutez por parte de los
17
nativos iroqueses en Canadaacute provocoacute la eutrofizacioacuten del Lago Crawford (43degN)
durante los siglos XII y XV Entre las consecuencias registradas se documentoacute un
claro incremento de la productividad primaria y cambios en la estructura comunitaria
de diatomeas foacutesiles (incremento de Stephanodiscus complex y disminucioacuten de
Cyclotella bodanica en Ekdahl et al 2007) Otro ejemplo demuestra como las
actividades agriacutecolas en la cuenca del riacuteo Mississippi modificaron los patrones de
sedimentacioacuten y aporte de nutrientes al lago Pepin EE UU (44degN) a partir del
asentamiento europeo en 1840 CE y principalmente desde 1940 CE (Engstrom et
al 2009) Para Chile von Gunten et al (2009) a partir de indicadores
limnogeoloacutegicos evidencioacute la contaminacioacuten y eutroficacioacuten de cinco lagos cercanos
a la ciudad de Santiago (33ordmS) como consecuencia de la deposicioacuten atmosfeacuterica
de metales pesados (especialmente Cu) quema de combustible foacutesil y flujo de
nutrientes durante los uacuteltimos 200 antildeos
Caracteriacutesticas limnoloacutegicas de los lagos
Los procesos limnoloacutegicos afectan la distribucioacuten y abundancia de los
organismos en los lagos Estaacuten influenciados por forzamientos externos por
ejemplo la precipitacioacuten o la penetracioacuten de la luz y la morfometriacutea del lago En este
sentido el nivel del lago dependeraacute del balance entre las entradas y salidas de agua
(precipitaciones escurrimiento superficial y subterraacuteneo y la evaporacioacuten) y la forma
de la cuenca (profundidad pendiente aacuterea del espejo de agua)
En funcioacuten de la profundidad de penetracioacuten de la luz es posible diferenciar
dos aacutereas que determinan la distribucioacuten de los organismos la zona foacutetica (donde
penetra la luz y permite la fotosiacutentesis) y la zona afoacutetica (subyacente a la zona
foacutetica) Al depender de la radiacioacuten la zona foacutetica variacutea estacionalmente Ademaacutes
18
puede variar por el grado de turbidez la morfometriacutea del lago y la produccioacuten de
materia orgaacutenica en la columna de agua
Otro factor que influye en la productividad es el reacutegimen de mezcla de la
columna de agua pues favorece la oxigenacioacuten y el reciclado de nutrientes La
mezcla ocurre en condiciones de gran inestabilidad usualmente relacionado con el
reacutegimen de viento Por el contrario un lago estratificado resulta de grandes
diferencias en temperatura o salinidad entre la superficie (epilimnion) y el fondo del
lago (hipolimnion) que separa las masas de agua superficial y de fondo por una
termoclina (si la estratificacioacuten estaacute relacionada con diferencias de temperatura de
las masas de agua) o haloclina (diferencias de salinidad) En funcioacuten del reacutegimen
de mezcla los lagos se pueden clasificar en (Lewis 1983)
1 Amiacutecticos no hay mezcla vertical de la columna de agua
2 Monomiacutecticos friacuteos y caacutelidos se mezcla una vez al antildeo
3 Dimiacutecticos la columna de agua se mezcla dos veces al antildeo
4 Oligomiacutecticos Lagos estratificados la mayor parte del tiempo pero se mezclan a
intervalos irregulares mayores a 1 antildeo
5 Polimiacutecticos friacuteos y caacutelidos la columna de agua se mezcla varias veces al antildeo
El ciclo del N en lagos
Al estar presente en aacutecidos nucleicos aminoaacutecidos y proteiacutenas el N es un
nutriente esencial de los organismos Por lo tanto su disponibilidad en la columna
de agua es clave para la produccioacuten composicioacuten y acumulacioacuten de la MO a traveacutes
del tiempo (Olson 1998 Talbot 2002) Aunque el principal reservorio de N estaacute en
19
la atmosfera (N2) solo unos pocos organismos (eg cianobacterias) pueden fijarlo
directamente Asiacute el Nitroacutegeno Inorgaacutenico Disuelto (DIN) representa la principal
fuente de N disponible para la produccioacuten primaria en los ecosistemas acuaacuteticos
(Talbot et al 2002) El DIN estaacute compuesto por nitrato (NO3-) nitrito (N02
-) y amonio
(NH4+) y los cambios en su disponibilidad condicionan el tipo de produccioacuten primaria
(eg fijadores vs asimiladores de N ver Scott y Grant 2013 Scott 2008)
La Figura 1 resume los principales componentes en lagos del ciclo del N y
sus interacciones La principal entrada natural de N es viacutea fijacioacuten de N atmosfeacuterico
y es llevado a cabo principalmente por bacterias y algas verde-azules capaces de
romper el triple enlace entre los dos aacutetomos de N a un alto costo energeacutetico (Torres
et al 2012) El N fijado es reducido a NH3 Tras la muerte de los organismos el N
es liberado de sus tejidos por hongos y bacterias implicados en la descomposicioacuten
de la MO Una parte se deposita en el fondo del lago y la otra queda disponible para
ser asimilada por el fitoplancton como amonio mediante el proceso de
amonificacioacuten (Talbot 2002) La amonificacioacuten resulta de la reduccioacuten bacteriana
del amoniaco y se da preferentemente en condiciones anoacutexicas La volatizacioacuten del
amoniaco es un proceso por medio este se incorpora a la atmoacutesfera Este proceso
se da preferentemente en lagos aacutecidos (pH gt 85) El nitrato es otra forma de N
bioloacutegicamente disponible para el fitoplancton En los lagos el NO3 del DIN estaacute
compuesto por los aportes desde la cuenca hidrograacutefica en la que se circunscriben
por la resuspensioacuten desde el fondo del lago favorecido por procesos de mezcla
(descrito en seccioacuten anterior) y por los nitratos de la columna de agua Estos uacuteltimos
son el resultado de la nitrificacioacuten proceso realizado por bacterias aeroacutebicas
mediante el cual el amonio se oxida para formar nitrito y nitrato El ciclo se completa
20
con la desnitrificacioacuten que es la reduccioacuten bacteriana de nitrato al N atmosfeacuterico
Este proceso se da preferentemente en condiciones anoacutexicas
Figura 1 Materia Orgaacutenica sedimentaria y su relacioacuten con el ciclo del N y las
variaciones de δ15N (elaborado a partir de Talbot et al 2002) En la figura se
representan los factores clave en la acumulacioacuten de la MO sedimentaria y su
relacioacuten con el ciclo del N El δ15N de la MO es resultado de los aportes de MO
desde la cuenca MO de macroacutefitas y de la productividad bioloacutegica La productividad
en ecosistemas lacustres estaacute condicionada por DIN y la fijacioacuten de N atmosfeacuterico
El δ15N del pool del DIN disponible se va enriqueciendo por la asimilacioacuten
preferencial de 14N por parte del fitoplancton Ademaacutes el DIN remanente se va
enriqueciendo por accioacuten microbiana (amonificacioacuten nitrificacioacuten desnitrificacioacuten)
Reconstruyendo el ciclo del N a partir de variaciones en δ15N
La sentildeal isotoacutepica de N (δ15N) en los sedimentos lacustres puede ser usada
para reconstruir los cambios pasados del ciclo N la transferencia de N entre
ecosistemas terrestres y acuaacuteticos y el estado troacutefico del lago (Bruland y Mackenzie
2015 McLauchlan et al 2013 Woodward et al 2012 von Gunten et al 2009
Leng et al 2006 Meyers y Teranes 2001) La Figura 1 resume los principales
procesos y fuentes que afectan la sentildeal isotoacutepica δ15N (total o ldquobulkrdquo) en la MO de
21
los sedimentos lacustres Entre los que destacan las fuentes de MO (aloacutectona vs
autoacutectona) la bioproductividad lacustre y las entradas de N (ie fijacioacuten bioloacutegica
de N2) (Torres et al 2012) Estos procesos conducen a un fraccionamiento
isotoacutepico cineacutetico que favorece la disociacioacuten entre sus isotopos ldquopesadosrdquo (15N) y
ldquoligerosrdquo (14N) Asiacute por ejemplo los valores de δ15N de la MO se enriquecen en 15N
en la medida que los sedimentos lacustres estaacuten sometidos a perdidas de 14N viacutea
desnitrificacioacuten (Bruland y Mackenzie 2015 Diaz et al 2016 Talbot et al 2002)
Por otra parte el fitoplancton en la medida que aumenta su biomasa (eg
durante la estacioacuten caacutelida) puede llegar a asimilar el DIN hasta agotarse En este
caso el N remanente se va empobreciendo en 14N si no hay reposicioacuten de N (eg
aporte de NO3 de la cuenca) y la MO queda enriquecida en 15N Esta disminucioacuten
induce a una limitacioacuten de fitoplancton por N y los organismos diztroacuteficos pueden
verse estimulados por lo que se incrementa la entrada de N viacutea fijacioacuten de N2 (Scott
y Grantsz 2013) como resultado la MO oscila en valores maacutes bajos (en torno a 0permil)
La cantidad de MO que se deposita en el fondo del lago depende del
predominio de contribuciones provenientes desde la cuenca (aloacutectona) versus las
producidas dentro del mismo lago (autoacutectona) (Torres et al 2012) Aunque en
general los lagos reciben permanentemente aportes de sedimentos y MO desde
su cuenca los lagos pequentildeos tienden a reflejar maacutes los procesos que ocurren
solamente en su cuenca directa y reflejan los valores isotoacutepicos de la misma (Gu et
al 2006) Woodward et al (2012) realizaron un estudio en 50 lagos en Irlanda que
les permitioacute correlacionar diferentes patrones de LUCC (bosques de coniacuteferas
agricultura ganaderiacutea entre otros) con los valores de δ15N (y δ13C) en los
sedimentos superficiales de los lagos Encontraron que los valores de δ15N son maacutes
negativos en cuencas poco impactadas y maacutes positivos en aquellas con alto
22
impacto humano (eg usos de suelo agriacutecola y ganadero) McLauchlan et al (2007)
encontraron valores maacutes positivos de δ15N en los sedimentos del Mirror Lake New
Hampshire (29ordm N) a partir de la desforestacioacuten de su cuenca en 1790 CE (inicio
del asentamiento europeo en el lago) para el desarrollo de la actividad agriacutecola
Estos valores se volvieron maacutes negativos hacia valores similares al pre-
asentamiento europeo despueacutes del cese de la agricultura en la cuenca y la
recuperacioacuten del bosque a partir de 1929 CE
El anaacutelisis de δ15N en los sedimentos lacustres es una herramienta casi sin
explorar en Chile central Proponemos reconstruir la dinaacutemica de transferencia de
N cuenca-lago como consecuencia de los LUCC desde la colonizacioacuten espantildeola en
los lagos costeros de Laguna Matanzas (34ordmS) y Lago Vichuqueacuten (36ordmS) Como son
muacuteltiples los procesos que pueden afectar la sentildeal isotoacutepica de N (medido como
δ15N) en los sedimentos lacustres existen muchos problemas para su
interpretacioacuten paleoambiental (Leng et al 2006) Por ello en esta tesis utilizamos
un enfoque multiproxy que consiste en integrar el anaacutelisis limnoloacutegico y geoquiacutemico
de los sedimentos lacustres con los valores isotoacutepicos modernos de la columna de
agua (mediante monitoreo del POM) la relacioacuten vegetacioacuten-suelo de la cuenca y la
reconstruccioacuten de los principales usos de suelo de las cuencas desde 1975 CE
mediante el uso de imaacutegenes satelitales Considerando estas muacuteltiples liacuteneas de
evidencia nos planteamos utilizar las variaciones del δ15N para reconstruir los
cambios en la disponibilidad de N en los lagos costeros de Chile central y establecer
coacutemo y desde cuaacutendo las actividades humanas en la cuenca son lo suficientemente
importantes como para modificar el ciclo del N en los lagos desde la colonizacioacuten
espantildeola (siglo XVII)
23
Como hipoacutetesis planteamos que la dinaacutemica de transferencia de sedimentos
y nutrientes entre la cuenca y el lago tiene controles geoloacutegicos climaacuteticos y
bioloacutegicos que interactuacutean en el tiempo registraacutendose en la acumulacioacuten de MO de
los sedimentos lacustres (Fig1) Bajo este marco los LUCC modifican esta
dinaacutemica alterando la transferencia de N a los lagos ya que las actividades agriacutecolas
y ganaderas tienden a aportar maacutes N (aumentando δ15N) que la actividad forestal
(la que disminuye δ15N)
En el primer capiacutetulo titulado ldquoA combined approach to establishing the timing
and magnitude of anthropogenic nutrient alteration in a mediterranean coastal lake-
watershed systemrdquo se evaluoacute el rol de los LUCC en el control de la descarga de N
y sedimentos a Laguna Matanzas en un sistema cuenca-lago desde el siglo XVII
Para ello se realizoacute un anaacutelisis de la dinaacutemica sedimentaria lacustre mediante el
anaacutelisis de muacuteltiples proxys sedimentoloacutegicos (ie minerales y textura)
geoquiacutemicos (eg geoquiacutemica orgaacutenica e isotoacutepica) y bioloacutegicos (ie diatomeas) de
Laguna Matanzas que cubren los uacuteltimos 200 antildeos Ademaacutes se realizoacute una
reconstruccioacuten de los usos de suelo entre 1975 y 2016 a partir de imaacutegenes de
sateacutelites y se colectaron muestras de suelo de las principales coberturas de la
cuenca a los cuales se midioacute el δ15N
Entre los principales resultados obtenidos se destaca la influencia de la
ganaderiacutea y la denitrificacioacuten lo que explica los altos valores de δ15N evidenciados
por el registro lacustre durante la colonizacioacuten espantildeola e inicios de la repuacuteblica A
partir de la Gran Aceleracioacuten (1950 CE) la desforestacioacuten y el reemplazo de la
ganaderiacutea por plantaciones forestales tienen un correlato en el registro
sedimentario con valores de δ15N maacutes bajos Estos resultados demuestran que los
LUCC son el factor de primer orden para explicar los cambios observados en
24
nuestro registro de δ15N y a su vez implica un rol secundario del clima como posible
control de la disponibilidad de N en Laguna Matanzas Este capiacutetulo fue sometido
a la revista Scientific Reports y estaacute en revisioacuten desde el 26 de marzo de 2019 En
la elaboracioacuten del mauscrito han colaborado Claudio Latorre Blas Valero-Garceacutes
Matiacuteas Frugone Pablo Sarricolea Santiago Giralt Manuel Contreras-Lopez
Ricardo Prego y Patricia Bernardez
El segundo capiacutetulo se titula ldquoStable isotopes track land use and cover
changes in a mediterranean lake in central Chile over the last 600 yearsrdquo Aquiacute
evaluamos la dinaacutemica del N actual en el sistema cuenca-lago considerando los
valores isotoacutepicos modernos tanto de la vegetacioacuten como del suelo ademaacutes de los
cambios estacionales del POM de la columna de agua Esta informacioacuten se utiliza
como un anaacutelogo moderno para conocer el rol de los LUCC en la disponibilidad de
N en los uacuteltimos 600 antildeos Para ello se colectaron muestras filtradas de la columna
de agua a 2 5 y 20 m dos veces por antildeo entre noviembre de 2015 y julio de 2018
y se obtuvo el δsup1⁵N del POM Ademaacutes se colectoacute muestras de vegetacioacuten y suelo
de la cuenca diferenciando entre especies nativas plantaciones forestales y
vegetacioacuten herbaacutecea Esta informacioacuten se analizoacute en conjunto con la reconstruccioacuten
de los LUCC entre 1975 y 2014 y el anaacutelisis limnoloacutegico del lago Ello nos permitioacute
evaluar coacutemo el δ15N de los sedimentos lacustres refleja los valores δ15N de la
cuenca en el Lago Vichuqueacuten y el control que ejerce el clima en la sentildeal isotoacutepica
de POM Este capiacutetulo seraacute sometido a la revista Science of total environmet
proximamente En la elaboracioacuten del mauscrito han colaborado Claudio Latorre
Blas Valero-Garceacutes Matiacuteas Frugone Pablo Sarricolea Leonardo Villacis y M Laura
Carrevedo
25
Entre los principales resultados encontramos que el δ15N en los sedimentos
lacustres es maacutes positivo con presencia de agricultura en la cuenca y maacutes negativo
cuando esta es reemplazada por cobertura arboacuterea bien sea plantaciones
forestales bien sea bosque nativo A su vez durante la estacioacuten seca y caacutelida la
mayor entrada al lago es viacutea fijacioacuten de N atmosfeacuterico (bajos valores de δ15NPOM)
Durante la estacioacuten huacutemeda en cambio prevalecen los aportes de la cuenca con
altos valores de δ15NPOM Ello estariacutea evidenciando un cambio estacional en la
composicioacuten de especies en verano ligado a especies fijadoras de N y en invierno
las algas y microorganismos que consumen el DIN de la columna de agua
Referencias
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea AM Marquet PA 2010 From the Holocene to the Anthropocene A historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the
next carbon Earth rsquo s Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409 httpsdoiorg102134jeq20090005
Diacuteaz FP Frugone M Gutieacuterrez RA Latorre C 2016 Nitrogen cycling in an
extreme hyperarid environment inferred from δ15 N analyses of plants soils and herbivore diet Sci Rep 6 1ndash11 httpsdoiorg101038srep22226
Ekdahl EJ Teranes JL Wittkop CA Stoermer EF Reavie ED Smol JP
2007 Diatom assemblage response to Iroquoian and Euro-Canadian eutrophication of Crawford Lake Ontario Canada J Paleolimnol 37 233ndash246 httpsdoiorg101007s10933-006-9016-7
Elbert W Weber B Burrows S Steinkamp J Buumldel B Andreae MO
Poumlschl U 2012 Contribution of cryptogamic covers to the global cycles of carbon and nitrogen Nat Geosci 5 459ndash462
26
httpsdoiorg101038ngeo1486 Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506
httpsdoiorg101126science1215567 ARTICLE Engstrom DR Almendinger JE Wolin JA 2009 Historical changes in
sediment and phosphorus loading to the upper Mississippi River Mass-balance reconstructions from the sediments of Lake Pepin J Paleolimnol 41 563ndash588 httpsdoiorg101007s10933-008-9292-5
Erisman J W Sutton M A Galloway J Klimont Z amp Winiwarter W (2008)
How a century of ammonia synthesis changed the world Nature Geoscience 1(10) 636
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892
httpsdoiorg101126science1136674 Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J Christie
D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592 Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
IPCC 2013 Climate Change 2013 The Physical Science BasisContribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change Cambridge University Press Cambridge United Kingdom and New York NY USA
httpsdoiorg101017CBO9781107415324 Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007 Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32iumliquestfrac12S 3050iumliquestfrac12miumliquestfrac12asl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470
27
httpsdoiorg101073pnas0701779104 McLauchlan KK Gerhart LM Battles JJ Craine JM Elmore AJ Higuera
PE Mack MC McNeil BE Nelson DM Pederson N Perakis SS 2017 Centennial-scale reductions in nitrogen availability in temperate forests of the United States Sci Rep 7 1ndash7 httpsdoiorg101038s41598-017-08170-z
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Myers N Mittermeler RA Mittermeler CG Da Fonseca GAB Kent J
2000 Biodiversity hotspots for conservation priorities Nature httpsdoiorg10103835002501
Pentildeuelas J Poulter B Sardans J Ciais P Van Der Velde M Bopp L
Boucher O Godderis Y Hinsinger P Llusia J Nardin E Vicca S Obersteiner M Janssens IA 2013 Human-induced nitrogen-phosphorus imbalances alter natural and managed ecosystems across the globe Nat Commun 4 httpsdoiorg101038ncomms3934
Prieto M Salazar D Valenzuela MJ 2019 The dispossession of the San
Pedro de Inacaliri river Political Ecology extractivism and archaeology Extr Ind Soc httpsdoiorg101016jexis201902004
Robertson GP Vitousek P 2010 Nitrogen in Agriculture Balancing the Cost of
an Essential Resource Ssrn 34 97ndash125 httpsdoiorg101146annurevenviron032108105046
Sardans J Rivas-Ubach A Pentildeuelas J 2012 The CNP stoichiometry of
organisms and ecosystems in a changing world A review and perspectives Perspect Plant Ecol Evol Syst 14 33ndash47 httpsdoiorg101016jppees201108002
Schindler DW Howarth RW Vitousek PM Tilman DG Schlesinger WH
Matson PA Likens GE Aber JD 2007 Human Alteration of the Global Nitrogen Cycle Sources and Consequences Ecol Appl 7 737ndash750 httpsdoiorg1018901051-0761(1997)007[0737haotgn]20co2
Scott E and EM Grantzs 2013 N2 fixation exceeds internal nitrogen loading as
a phytoplankton nutrient source in perpetually nitrogen-limited reservoirs Freshwater Science 2013 32(3)849ndash861 10189912-1901
28
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Talbot M R (2002) Nitrogen isotopes in palaeolimnology In Tracking
environmental change using lake sediments (pp 401-439) Springer Dordrecht
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable
isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K
Yeager KM 2016 Different responses of sedimentary δ15N to climatic changes and anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
29
CAPIacuteTULO 1 A COMBINED APPROACH TO ESTABLISHING THE TIMING
AND MAGNITUDE OF ANTHROPOGENIC NUTRIENT ALTERATION IN A
MEDITERRANEAN COASTAL LAKE- WATERSHED SYSTEM
30
A combined approach to establishing the timing and magnitude of anthropogenic
nutrient alteration in a mediterranean coastal lake- watershed system
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugoneabc Pablo
Sarricolead Santiago Giralte Manuel Contreras-Lopezf Ricardo Prego g Patricia
Bernaacuterdez g Blas Valero-Garceacutesch
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Institute of Earth Sciences Jaume Almera (ICTJA-CSIC) CLluis Solegrave Sabaris sn Barcelona E-
08028 Spain
f Facultad de Ingenieriacutea y Centro de Estudios Avanzados Universidad de Playa Ancha Traslavintildea
450 Vintildea del Mar Chile
g Instituto de Investigaciones Marinas (CSIC) CEduardo Cabello 6 36208 Vigo Spain
h Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding author
E-mail address
clatorrebiopuccl magdalenafuentealbagmailcom
Abstract
Since the industrial revolution and especially during the Great Acceleration (1950
CE) human activities have profoundly altered the global nutrient cycle through land
use and cover changes (LUCC) However the timing and intensity of recent N
variability together with the extent of its impact in terrestrial and aquatic ecosystems
and coupled effects of regional LUCC and climate are not well understood Here
we used a multiproxy approach (sedimentological geochemical and isotopic
31
analyses historical records climate data and satellite images) to evaluate the role
of LUCC as the main control for N cycling in a coastal watershed system of central
Chile during the last few centuries The largest changes in N dynamics occurred in
the mid-1970s associated with the replacement of native forests and grasslands for
livestock grazing by plantations of introduced Monterey pine (Pinus radiata) and
eucalyptus (Eucalyptus globulus) Lake productivity increased and is evidenced by
an increase trend in δ15N values Our study shows that anthropogenic land
usecover changes are key in controlling nutrient supply and N availability in
Mediterranean watershed ndash lake systems and that large-scale forestry
developments during the mid-1970s likely caused the largest changes in central
Chile
Keywords Anthropocene Organic geochemistry watershedndashlake system Stable
Isotope Analyses Land usecover change Nitrogen cycle Mediterranean
ecosystems central Chile
1 INTRODUCTION
Human activities have become the most important driver of the nutrient cycles in
terrestrial and aquatic ecosystems since the industrial revolution (Gruber and
Galloway 2008 Galloway et al 2008 Holtgrieve et al 2011 Fowler et al 2013
Goyette et al 2016) Among these N is a common nutrient that limits productivity
in terrestrial and aquatic ecosystems (Vitousek and Howarth 1991 McLauchlan et
al 2013) With the advent of the Haber-Bosch industrial N fixation process in the
early 20th century total N fluxes have surpassed previous planetary boundaries
32
(Galloway et al 2008 Elser 2011) reaching unprecedented values (ie tipping
points) in the Earth system especially during what is now termed the Great
Acceleration (which began in the 1950s) which intensified in the 1970s (Howarth
2004 Steffen et al 2015) Yet dramatic changes in LUCC have occurred in the last
few centuries mainly linked to forestry and agro-pastoral activities (Poraj-Goacuterska et
al 2017 Ge et al 2019 Cheng et al 2019) These have had large impacts on N
(and Carbon -C-) availability in terrestrial and aquatic ecosystems but the synergic
effect with climate change and global N dynamics has not been established
(Vitousek et al 1997 Mclauchlan et al 2007 2013 Pongratz et al 2010
Woodward et al 2012 Mclauchlan et al 2017)
The onset of the Anthropocene poses significant challenges in mediterranean
regions that have a strong seasonality of hydrological regimes and an annual water
deficit (Stocker et al 2013) Mediterranean climates occur in all continents
(California central Chile Australia South Africa circum-Mediterranean regions)
providing a unique opportunity to investigate global change processes during the
Anthropocene in similar climate settings but with variable geographic and cultural
contexts The effects of global change in mediterranean watersheds have been
analyzed from different perspectives hydrology (Giorgi 2006 Arnell and Gosling
2013 Prudhomme et al 2014) vegetation dynamics (Lenihan et al 2003 Vicente-
Serrano et al 2013 Matesanz and Valladares 2014) sediment dynamics (Garciacutea-
Ruiz et al 2013 Garciacutea-Ruiz 2010 Syvitski et al 2005) changes in
biogeochemical cycles (Thomas et al 2013 Fowler et al 2013 Frank et al 2015)
carbon storage (Muntildeoz-Rojas et al 2015) and biodiversity (Hooper et al 2012) A
recent review showed an extraordinarily high variability of erosion rates in
mediterranean watersheds positive relationships with slope and annual
33
precipitation and the paramount effect of land use (Garciacutea-Ruiz et al 2015)
However the temporal context and effect of LUCC on nutrient supply to
mediterranean lakes has not been analyzed in much detail
Major LUCC in central Chile occurred during the Spanish Colonial period
(17th-18th centuries) (Armesto 1994 Gurnell et al 2001 Figueroa et al 2004
Martiacuten-Foreacutes et al 2012 Contreras-Loacutepez et al 2014) with the onset of
industrialization and mostly during the mid to late 20th century (von Gunten et al
2009 Gayo et al 2012) Recent copper pollution caused by 20th century mining
and industrial smelters has been documented in cores throughout the Andes
(Laguna el Ocho and Laguna Ensuentildeo von Gunten et al 2009) and also from our
surveys in the central valley (Batuco wetland) coastal range (Cordillera de Name)
and along the coast itself (Bucalemito and Colejuda) (Valero-Garceacutes et al 2010
unpublished data)
Paleolimnological studies have shown how these systems respond to
climate LUCC and anthropogenic impacts during the last millennia (Jenny et al
2002 2003 Villa Martiacutenez et al 2002 Frugone-Aacutelvarez et al 2017 Contreras et
al 2018) Furthermore changes in sediment and nutrient cycles have also been
identified in associated terrestrial ecosystems dating as far back as the Spanish
Conquest and related to fire clearance and wood extraction practices of the native
forests (Armesto 1994 Contreras-Loacutepez et al 2014) Nevertheless pollen and
limnological evidence argue for a more recent timing of the largest anthropogenic
impacts in central Chile For example paleo records show that during the mid-20th
century increased soil erosion followed replacement of native forest by Pinus
radiata and Eucalyptus globulus plantations at Laguna Matanzas Aculeo and
34
Vichuqueacuten lakes (Jenny et al 2002 Villa-Martiacutenez et al 2002 2003 Frugone-
Aacutelvarez et al 2017)
Lakes are a central component of the global carbon cycle Lakes act as a
sink of the carbon cycle both by mineralizing terrestrially derived organic matter and
by storing substantial amounts of organic carbon (OC) in their sediments (Anderson
et al 2009) Paleolimnological studies have shown a large increase in OC burial
rates during the last century (Heathcote et al 2015) however the rates and
controls on OC burial by lakes remain uncertain as do the possible effects of future
global change and the coupled effect with the N cycle LUCC intensification of
agriculture and associated nutrient loading together with atmospheric N-deposition
are expected to enhance OC sequestration by lakes Climate change has been
mainly responsible for the increased algal productivity since the end of the 19th
century and during the late 20th century in lakes from both the northern (Ruumlhland et
al2015) and southern hemispheres (Michelutti et al 2015 Carrevedo et al 2015)
but many studies suggest a complex interaction of global warming and
anthropogenic influences and it remains to be proven if climate is indeed the only
factor controlling these transitions (Catalaacuten et al 2013) Alternative causes for
recent N (Galloway et al 2008) increases in high altitude lakes such as catchment
mediated processes cannot be ruled out (Camarero and Catalan 2012 Fritz and
Anderson 2013) Few lake-watershed systems have robust enough chronologies of
recent changes to compare variations in C and N with regional and local processes
and even fewer of these are from the southern hemisphere (McLauchlan et al
2007 Holtgrieve et al 2011)
In this paper we present a multiproxy lake-watershed study including N and
C stable isotope analyses on a series of short cores from Laguna Matanzas in
35
central Chile focused in the last 200 years We complemented our record with land
use surveys satellite and aerial photograph studies Our major objectives are 1) to
reconstruct the dynamics among climate human activities and changes in the N
cycle over the last two centuries 2) to evaluate how human activities have altered
the N cycle during the Great Acceleration (since the mid-20th century)
2 STUDY SITE
Laguna Matanzas (33ordm45rsquoS 71ordm 40acuteW 7 m asl Fig 1) is a coastal lake located
in central Chile near to a large populated area (Santiago gt6106 inhabitants) The
lake has a surface area of 15 km2 with a max depth of 3 m and a watershed of 30
km2 The lake basin is emplaced over Pleistocene-Holocene aeolian and alluvial fan
deposits (SERNAGEOMIN 2003) A recent phase of dune activity occurred from the
mid to late Holocene which mostly sealed off the basin from the ocean (Villa-
Martinez 2002) Climate in Laguna Matanzas is characterized by cool-wet winters
and hot-dry summers with annual precipitation of ~510 mm and a mean annual
temperature of 12ordmC Central Chile is in the transition zone between the southern
hemisphere mid-latitude westerlies belt and the South Pacific Anticyclone (or SPA)
(Garreaud et al 2009) In winter precipitation is modulated by the north-west
displacement of the SPA the northward shift of the westerlies wind belt and an
increased frequency of storm fronts stemming off the Southern Hemisphere
Westerly Winds (SWW) (Frugone-Aacutelvarez et al 2017) Austral summers are
typically dry and warm as a strong SPA blocks the northward migration of storm
tracks stemming off the SWW
36
Historic land cover changes started after the Spanish conquest with a Jesuit
settlement in 1627 CE near El Convento village and the development of a livestock
ranch that included the Matanzas watershed After the Jesuits were expelled from
South America in 1778 CE the farm was bought by Pedro Balmaceda and had more
than 40000 head of cattle around 1800 CE (Contreras-Lopez et al 2014) The first
Pinus radiata and Eucalyptus globulus trees were planted during the second half of
the 19th century and were mostly used for dune stabilization (Albert 1900 Gibson
1972) However the main plantation phase occurred 60 years ago (Villa-Martinez
2002) as a response to the application of Chilean Forestry Laws promulgated in
1931 and 1974 and associated state subsidies
Major land cover changes occurred recently from 1975 to 2008 as shrublands
were replaced by more intensive land uses practices such as farmland and tree
plantations (Schulz et al 2010) Laguna Matanzas is part of the Reserva Nacional
Humedal El Yali protected area (Ramsar Ndeg 878) Despite this protected status the
lake and its watershed have been heavily affected by intense agricultural and
farming activities during the last decades The main inlet ldquoLas Rosasrdquo has been
diverted for crop irrigation causing a significant loss of water input to the lake
Consequently the flooded area of the lake has greatly decreased in the last couple
of decades (Fig 1b) Exotic tree species cover a large surface area of the
watershed Recently other activities such as farms for intensive chicken production
have been emplaced in the watershed
37
Figure 1 Laguna Matanzas a) Digital Elevation Model showing the watershed and
the surface hydrologic connection with Colejuda and Cabildo lakes b) Climograph
depicting the warm dry season in austral summer c) Annual precipitation from
1965-2015 (arrow shows start of the most recent ldquomega-droughtrdquo- see Garreaud et
al 2017) d) A decline of lake area occurs between 2007 and 2018 Lake surface
area decreased first along the western sector (in 2007) followed by more inland
areas (in 2018)
38
3 RESULTS
31 Age Model
The age model for the Matanzas sequence was developed using Bacon software
to establish the deposition rates and age uncertainty (Blaauw and Christen 2011)
It is based on 210Pb and 137Cs dates and two 14C dates (Table 2) According to this
age model the lake sequence spans the last 1000 years (Fig 2) A major breccia
layer (unit 3b) was deposited during the early 18th century which agrees with
historic documents indicating that a tsunami impacted Laguna Matanzas and its
watershed in 1730 CE (Contreras-Loacutepez et al 2014) Here we focus in the last 200
years were the most important changes occurred in terms of LUCC (after the
sedimentary hiatus caused by the tsunami) The Spanish colonial period (17th -18th
century) brought new forms of territorial management along with an intensification
of watershed use which remained relatively unchanged until the 1900s
39
Figure 2 a) Age-depth model obtained for the Laguna de Matanzas sedimentary
sequence A hiatus occurs at 80 cm depth (unit 3b) The section of core used for our
analysis is highlighted in a red rectangle b) Close up of the age model used for
analysis of recent anthropogenic influences on the N cycle c) Information regarding
the 14C dates used to construct age model
Lab code Sample ID
Depth (cm) Material Fraction of modern C
Radiocarbon age
Pmc Error BP Error
D-AMS 021579
MAT11-6A 104-105 Bulk Sediment
8843 041 988 37
D-AMS 001132
MAT11-6A 1345-1355
Bulk Sediment
8482 024 1268 21
POZ-57285
MAT13-12 DIC Water column 10454 035 Modern
Table 2 Laguna Matanzas radiocarbon dates
32 The sediment sequence
Laguna Matanzas sediments consist of massive to banded mud with some silt
intercalations They are composed of silicate minerals (plagioclase quartz and clay
minerals) with relatively high TOC content (Fig 3) Pyrite is a common mineral
indicating dominant anoxic conditions in the lake sediments whereas aragonite
occurs only in the uppermost section Mineralogical analyses visual descriptions
texture and geochemical composition were used to characterize five main facies
(Fig 3) F1 (organic-rich mud) represents baseline sedimentation in a shallow well-
mixed brackish highly productive lake and F1acute (dark orange) is a less organic facies
than F1 (more details see table in the supplementary material) F2 (massive to
banded silty mud) indicates periods of higher clastic input into the lake but finer
(mostly clay minerals) likely from suspension deposition associated with flooding
40
events Aragonite (up to 15 ) occurs in both facies but only in samples from the
uppermost 30 cm and it is interpreted as endogenic linked to higher MgFe waters
and elevated biologic productivity
Figure 3 Sedimentary facies and units mineralogy grain size elemental geochemical
and isotopic composition of Laguna Matanzas core Values highlighted in gray indicate
that these are above average
The banded to laminated fining upward silty clay layers (F3) reflect
deposition by high energy turbidity currents The presence of aragonite suggests
that littoral sediments were incorporated by these currents Non-graded laminated
coarse silt layers (F4) do not have aragonite indicating a dominant watershed
41
sediment source Both facies are interpreted as more energetic flood deposits but
with different sediment sources A unique breccia layer with coarse silt matrix and
cm-long soft clasts (F5) is interpreted as a high-energy event (ie a tsunami)
capable of eroding the littoral zone and depositing coarse clastic material in the
distal zone of the lake Similar coarse breccia layers have been found at several
coastal sites in Chile and interpreted as tsunami-related deposits (Vargas et al
2005 Le Roux et al 2008)
33 Sedimentary units
Three main units and six subunits have been defined (Fig 3) based on
sedimentary facies and sediment composition We use ZrTi as an indicator of the
mineral fraction transported from the watershed (Marzecoacuteva et al 2011) as higher
ZrTi (F3 and F4) is commonly associated with coarser sediments (Cuven et al
2010) A high correlation among Br BrTi and TOC (r = 046-087 p value=0)
supports the use of BrTi as an indicator of lake productivity (Marzecoacuteva et al 2011
Frugone-Aacutelvarez et al 2017) The MnFe ratio is indicative of lake bottom
oxygenation (Naecher et al 2013) as under reducing conditions Mn mobilizes more
than Fe leading to a decreased MnFe ratio (Marzecoacuteva et al 2011) SrTi indicates
periods of increased aragonite formation as Sr is preferentially included in the
aragonite mineral structure (Veizer et al 1971) (See supplementary material)
The basal unit (3c 129 to 99 cm) is relatively organic-rich (TOC mean=26
BioSi mean=5) and composed by F1 (without aragonite) with some coarser F4
flood layers (ZrTi mean=025) Unit 3b (98 to 80 cm) is interpreted as a tsunami or
storm surge deposit (breccia F5 grading into massive to banded silt F2) Unit 3a
(79 to 73 cm) is characterized by relatively low productivity (TOC=2 BrTi=002
42
BioSi=4) under anoxic conditions (MnFe=001) Unit 2a (72 to 31cm) has
relatively less organic content and more intercalated clastic facies F3 and F4 The
top of this unit (43-30 cm) has elevated TS values The Subunit 1b (30-20 cm)
shows increasing TOC BioSi and BrTi values (TOC mean = 29 BioSi mean =
54 BrTi mean=004) and the upper subunit 1a (19-0 cm) has the highest TOC
(mean = 64) and BioSi (mean = 56) high BrTi mean = 010 and the presence
of aragonite More frequent anoxic conditions (MnFe lower than 001) during units
3 and 2 shifted towards more oxic episodes (MnFe = 003) in unit 1 (Fig 3)
34 Isotopic signatures
Figure 4 shows the isotopic signature from soil samples of the major land
usescover present in the Laguna Matanzas used as an end member in comparison
with the lacustrine sedimentary units δ15N from cropland samples exhibit the
highest values whereas grassland and soil samples from lake shore areas have
intermediate values (Fig 4) Tree plantations and native forests have similarly low
δ15N values (+11 permil SD=24) All samples (except those from the lake shore)
exhibit low δ13C values (from ndash285permil to ndash298permil) CNmolar from agriculture land
lakeshore area and non-vegetation areas samples display the lowest values (about
18) CNmolar from tree plantations and native forest have the highest values (383
and 267 respectively)
43
Figure 4 C-N stable isotope plot showing a comparison of lake sediments grouped
by sedimentary units (MAT11-6A) with the soil end members of present-day (lake
shore and land usecover) from Laguna Matanzas
The δ15N values from sediment samples (MAT11-6A) range from ndash15 and
+53permil (mean= +35 SD=05) δ13C values range from ndash266permil to ndash202permil (mean=
ndash240 SD=14) In unit 3c δ15N and δ13C show relatively high values (mean=
+41permil and ndash233permil respectively) δ15N and δ13C from unit 3b and 3a fluctuate at
slightly lower values than in 3c (mean δ15N from 3a= +38permil and 3b mean= +39permil
mean δ13C from 3a= ndash242permil and 3b mean= ndash244permil) In unit 2a δ15N values are
relatively high (mean= +38permil) but show a slightly decreasing trend (from +52 to
+24permil at 66 cm and 36 cm respectively) δ13C also decreases (mean= ndash242permil)
reaching minimum values at 45 cm (ndash266permil) with a sharp increase towards the top
of this unit with maximum values ca ndash210permil Unit 1 exhibits the lowest δ15N values
(1a mean= +11permil 1b mean= +28permil) with negative values in the uppermost
44
sediments (ndash04permil at 14 cm) δ13C values show a decreasing trend over most of
subunit 1b and increase only near the very top of this unit
35 Recent land use changes in the Laguna Matanzas watershed
Major LUCC from 1975 (unit 1b) to 2016 CE in the Laguna Matanzasacutes
watershed is summarized in Figure 5 The watershed has a surface area of 30 km2
of which native forest (36) and grassland areas (44) represented 80 of the
total surface in 1975 The area occupied by agriculture was only 02 and tree
plantations were absent Isolated burned areas (33) were located mostly in the
northern part of the watershed By 1989 tree plantations surface area had increased
to 5 burned areas to 17 and agricultural fields to 9 (a 45-fold increase) and
native forest and grassland sectors decreased to 23 and 27 respectively By
2016 agricultural land and tree plantations have increased to 17 of the total area
whereas native forests decreased to 21
45
Figure 5 Land uses and cover changes from 1975 to 2016 in Laguna Matanzas
watershed from natural cover and areas for livestock grazing (grassland) to the
expansion of agriculture and forest plantation
4 DISCUSSION
41 N and C dynamics in Laguna Matanzas
Small lakes with relatively large watersheds such as Laguna Matanzas would
be expected to have relatively high contributions of allochthonous C to the sediment
OM (Gu et al 2006) Terrestrial C3 plants (δ13C =ndash26 to ndash28permil Meyers and Terranes
2001 Ku et al 2007) are dominant in the Laguna Matanzas watershed Likewise
our soil samples ranged across similar although slightly more negative values
46
(δ13C = ndash30 to ndash28permil Fig 4) to those proposed by Meyers and Terranes (2001)
and are used here as terrestrial end members oil samples were taken from the lake
shore (δ13C = ndash22permil) and POM from surface water (δ13C = ndash24permil) were more
positive than the terrestrial end member and are used as lacustrine end members
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from terrestrial vegetation and more positive δ13C values have increased
aquatic OM (Gu et al 2006 Torres et al 2012) Phytoplankton preferentially uptake
12C leaving the DIC pool enriched in 13C (Gu et al 2006) especially when there are
no important external sources of C (eg decreased C input from the watershed)
Therefore during events of elevated primary productivity the phytoplankton uptakes
12C until its depletion and are then obligated to use the heavier isotope resulting in
an increase in δ13C Changes in lake productivity thus greatly affect the C isotope
signal (Torres et al 2012) with high productivity leading to elevated δ13C values
(Torres et al 2012 Gu et al 2006)
In a similar fashion the N isotope signatures in Laguna Matanzas reflect a
combination of factors including different N sources (autochthonousallochthonous)
and lake processes such as productivity isotope fractionation in the water column
and sediment denitrification Elevated δ15N values from a POM sample (+22permil) and
average values from the lake shore (mean=+34permil SD=028) are used as aquatic
end members whereas terrestrial samples have values from +10 +24 (tree species)
to +77permil +35 (agriculture) which we use as terrestrial end members (Fig 4)
Autochthonous OM in aquatic ecosystems typically displays low δ15N values
when the OM comes from N-fixing species Atmospheric fixation of N2 by
cyanobacteria ends in OM with δ15N values close to 0permil (Leng et al 2006)
Phytoplankton preferentially uptake 14N from Dissolved Inorganic Nitrogen (DIN) in
47
the water column and derived OM typically have δ15N values lower than DIN values
When productivity increases the remaining DIN becomes depleted in 14N which in
turn increases the δ15N values of phytoplankton over time especially if the N not
replenished (Torres et al 2012) Thus high POM δ15N values from Laguna
Matanzas reflect elevated phytoplankton productivity with a 14N- depleted DIN In
addition N-watershed inputs also contribute to high δ15N values Heavily impacted
watersheds by human activities are often reflected in isotope values due to land use
changes and associated modified N fluxes For example the input of N runoff
derived from the use of inorganic fertilizers leads to the presence of elevated δ15N
(between ndash4 to +4permil) values in the water bodies (Elliott and Brush 2006 Diebel and
Vander Zanden 2009) Widory et al (2004) reported a direct relationship between
elevated δ15N values and increased nitrate concentration from manure in the
groundwater of Brittany (France) Terranes and Bernasconi (2000) found a good
correlation between augmented nutrient loading and a progressive increase in δ15N
values of sedimentary OM related to agricultural land use
Post-depositional diagenetic processes can further affect C and N isotope
signatures Several studies have shown a decrease in δ13C values of OM in anoxic
environments particularly during the first years of burial related to the selective
preservation of OM depleted in 13C (Hollander and Smith 2001 Lehmann et al
2002 Gӓlman et al 2009 Torres et al 2012) Diagenetic processes can also lead
to post-burial N isotope enrichment of the sediments Indeed 14N is consumed more
rapidly than 15N by denitrifying bacteria which intensifies under anoxic conditions
(Diebel and Vander Zanden 2009) Thus the remaining OM pool will be enriched
in 15N leaving to elevated δ15N values (Granger et al 2008 Menzel et al 2013)
48
In summary the relatively high δ15N values in sediments of Laguna Matanzas
reflect N input from an agriculturegrassland watershed with positive synergetic
effects from increased lake productivity enrichment of DIN in the water column and
most likely denitrification The increase of algal productivity associated with
increased N terrestrial input andor recycling of lake nutrients (and lesser extent
fixing atmospheric N) and denitrification under anoxic conditions can all increase
δ15N values (Fig 3) In addition elevated lake productivity without C replenishing
(eg by terrestrial C input) produces shifts towards positive δ13C values whereas C
input from the watershed generates more negative δ13C values
42 Recent evolution of the Laguna Matanzas watershed
Sedimentological compositional and geochemical indicators show three
depositional phases in the lake evolution under the human influence in the Laguna
Matanzas over the last two hundred years Although the record is longer (around
1000 years) we analyzed the last two centuries (unit 2 and 1) to provide a recent
historical context for the large changes detected during the 20th century
The first phase lasted from the beginning of the 19th century until ca 1940
(unit 2a Fig 3 4) and was characterized by moderate productivity with elevated
sediment input from the watershed as indicated by our geochemical proxies (BrTi
= 004 AlTi gt 01 Fig 6) Higher lake levels and dominant anoxic bottom conditions
(MnFe = 001 Fig3) can be related to increased precipitation (mean= 557 mmyr)
and lower temperatures (summer annual temperature lt19ordmC) During the Spanish
colonial period the Laguna Matanzas watershed was used as a livestock farm
(Jesuits 1627-1780 unit 3 followed by the Pedro Balmaceda era 1780-1810- unit
2a) (Contreras-Loacutepez et al 2014) The main buildings and farm facilities at the El
49
Convento village During this period livestock grazing and lumber extraction for
mining would have involved extensive deforestation and loss of native vegetation
(eg Armesto et al 1994 2010) However the Matanzas pollen record does not
show any significant regional deforestation during this period (Villa Martiacutenez 2002)
suggesting that the impact may have been highly localized
Lake productivity sediment input and elevated precipitation (Fig 6) all
suggest that N availability was related to this increased input from the watershed
The N from cow manure and soil particles would have led to higher δ15N values
(Elliot and Brush 2016) In addition anoxic lake bottom conditions would lead to
even further enrichment of buried sediment N The δ13C values lend further support
to our interpretation of increased sediment input -and N- from the watershed
Decreased δ13C values reach their lowest values in the entire record (ndash270permil) at
ca 1910 CE (Fig 4 6)
During most of the 19th century human activities in Laguna Matanzas were
similar to those during the Spanish Colonial period However the appearance of
Pinus radiata and Eucalyptus globulus pollen ca 1890 CE (Gibson 1972) the dune
stabilization-afforestation program which began ca 1900 CE (Albert 1900) and the
application of the Chilean forestry law of 1931 (DFL nordm265) contributed to an
increased capacity of the surrounding vegetation to retain nutrients and sediments
The law subsidized forest plantations in areas devoid of vegetation and prohibited
the cutting of forest on slopes greater than 45ordm These land use changes were coeval
with decreased sediment inputs (AlTi trend) from the watershed slightly increased
lake productivity (BrTi from 001 to 003) and a decrease in annual precipitation
(Fig 6) N isotope values become more negative during this period although they
remained high (from +49permil to +37permil) whereas the δ13C trend towards more
50
positive values reflects changes in the N source from watershed to in-lake dynamics
(e g increased endogenic productivity)
The second phase started after 1940 and is clearly marked by an abrupt
change in the general trend of δ15N that oscillates around +41permil (SD = 04permil) during
the first phase to +24permil (SD = 14permil) during the second These δ15N trends reflect
the lowest watershed nutrient and sediment inputs (based on the AlTi record)
decreased precipitation (mean = 318 mm year) and a slight increase in lake
productivity (increased BrTI) Depositional dynamics in the lake likely crossed a
threshold as human activity intensified throughout the watershed and lake levels
decreased
During the Great Acceleration δ15N values shifted towards higher values to
ca 3permil with an increase in δ13C values that are not reflected either in lake
productivity or lake level As the sediment input from the watershed increased and
precipitation remained as low as the previous decade δ15N values during this period
are likely related to watershed clearance which would have increased both nutrient
and sediment input into the lake
The δ13C trend to more positive values reaching the peaks in the 1960s (ndash
212permil) at the same time as the peaks in temperature (196 ordmC mean annual) with a
downward trend in precipitation A shift in OM origin from macrophytes and
watershed input influences to increased lake productivity could explain this trend
(Fig 4 1b)
In the 1970s the Laguna Matanzasacute watershed was mostly covered by native
forest (36) and grassland areas were intended for livestock grazing (44 Fig 5)
Both soils have δ15NSoil lt 5permil (Fig 4) Farming fields occupied a very small area and
tree plantations were almost nonexistent The decreasing trend in δ15N values seen
51
in our record is interrupted by several large peaks that occurred between ca 1975
and ca 1989 when the native forest and grassland areas fell by 23 and 27
respectively largely due to fires affecting 17 of the forests Agriculture fields
increased by 4 and forest plantations increased by 9 (unit 1a) Concomitantly
sediment ndash and likely N - inputs from the watershed decreased (as indicated by the
trend in AlTi) the precipitation was still relatively low (Fig 6) These changes are
likely related to the increase of vegetation cover especially of tree plantations (which
have more negative δ15N values) The small increase in productivity in the lake could
have been favored by increased temperature (von Gunten et al 2009) After 1989
the increase in agriculture land (17 in 2016) correlates with increasing δ15N δ13C
and TOC trends in spite of declining rainfall The increase of forest plantations was
mostly in response to the implementation of the Law Decree of Forestry
Development (DL 701 of 1974) that subsidized forest plantation After 1989 the
increase in agricultural land (17 in 2016) is synchronous with increasing δ15N
δ13C TOC and MnFe trends despite the decline in rainfall and overall lower lake
levels as more water is used for irrigation
The third phase started c 1990 CE (unit 1a) when OM accumulation rates
increase and δ13C δ15N decreased reaching their lowest values in the sequence
around 2000 CE Afterward during the 21st century δ13C and δ15N values again
began to increase The onset of unit 1 is marked by increased lake productivity and
decreased sediment input (AlTi lt 02) synchronous with intensive farming replacing
forestry and extensive agriculture (Fig 5 6)
A change in the general trend of δ15N values which decreased until 1990
(unit 1a) as the native forest and grassland area fell to 23 and 27 respectively
is most likely due to deforestation and fires Agriculture surface increased to 4 and
52
forest plantations increased by 9 (Fig 5) Concomitantly sediment ndash and likely N
ndash inputs from the watershed decreased probably related to the low precipitation (Fig
1b) and the increase of vegetation cover in the watershed in particularly by tree
plantations (with more negative δ15N Fig 4)
At present agriculture and tree plantations occupy around 34 of the
watershed surface whereas native forests and grassland cover 21 and 25
respectively Lake productivity as indicated by BrTi (Fig6) is higher and generates
OM with lower δ15N and higher δ13C values (about +2permil and ndash20permil ca 2000 CE
respectively) due to in-lake processes (ie biological N fixation and nutrient
recycling) and driven by changes in the arboreal cover which diminishes nutrient
flux into the lake (Fig6)
53
Figure 6 Anthropogenic and climatic forcing and lake dynamics response
(productivity sediment input N and C cycles) at Matanzas Lake over the last two
54
centuries Mean annual precipitation reconstructed and temperatures (von Gunten
et al 2009) Vertical gray bars indicate mega-droughts
5 CONCLUSIONS
Human activities have been the main factor controlling the N and C cycle in
the Laguna Matanzas during the last two centuries The N isotope signature in the
lake sediments reflects changes in the watershed fluxes to the lake but also in-lake
processes such as productivity and post-depositional changes Denitrification could
have been a dominant process during periods of increased anoxic conditions which
were more frequent prior to Great Acceleration (1950 CE) Higher δ15N and lower
δ13C values are associated with increased nutrient input from the watershed due to
increased livestock grazing and agriculture pressure (phase 1 Fig 7a) whereas
lower isotope values occurred during periods of increased forest plantations (phase
3 Fig 7c) During periods of increased lake productivity - such as in the last few
decades - δ15N values increased significantly
The most important change in C and N dynamics in the lake occurred after the
1950s (Phase 2 and 3) as tree plantations dominated the watershed Recent
changes in N dynamics can be explained by the higher nutrient contribution
associated with intensive agriculture (i e fertilizers) since the 1990s Although the
replacement of livestock activities with forestry and farming seems to have reduced
nutrient and soil export from the watershed to the lake the inefficient use of fertilizer
(by agriculture) can be the ultimate responsible for lake productivity increase during
the last decades
55
Figure 7 Schematic diagrams illustrating the main factors controlling the
isotope N signal in sediment OM of Laguna Matanzas N input from watershed
depends on human activities and land cover type Agriculture practices and cattle
(grassland development) contribute more N to the lake than native forest and
plantations Periods of higher productivity tend to deplete the dissolved inorganic N
in 14N resulting in higher δ15N (OM) The denitrification processes are more effective
in anoxic conditions associated with higher lake levels
6 METHODS
Short sediment cores were recovered from Laguna Matanzas using an Uwitec
gravity piston corer in 2011 and 2013 (Fig 1a) Sediment cores (MAT11-6A 129 cm
MAT13-2A 149 cm MAT13-3A 33 cm MAT13-4A 99 cm) were split
photographed sub-sampled and stored at the Pyrenean Institute of Ecology (IPE-
CSIC Spain) Core MAT11-6A was obtained from the central sector of the lake and
56
was selected for detailed multiproxy analyses (including elemental geochemistry C
and N isotope analyses XRF and 14C dating)
The isotope analyses (δ13C and δ15N) were performed at the Laboratory of
Biogeochemistry and Applied Stable Isotopes (LABASI-PUC Chile) using a Delta
V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via a
Conflo IV interface Isotope results are expressed in standard delta notation (δ) in
per mil (permil) relative to the standards Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Sediment samples
for δ13Corg were pre-treated with 50 ml of HCl and 50 ml of deionized water and
dried at 60ordmC for 4 hr to remove carbonates (Harris et al 2011)
Total Carbon (TC) Total Organic Carbon (TOC) Total Inorganic Carbon (TIC)
and Total Sulphur (TS) were measured every cm with a LECO SC 144 DR at IPE-
CSIC XRF measurements were carried out every 4 mm in MAT11-1A-1G core using
an AVAATECH X-ray Fluorescence II core scanner at the University of Barcelona
(Spain) Results are expressed as element intensities in counts per second (cps)
Tube voltage was operated at 30 kV and 10 kV to obtain the abundances of 15
elements (Al Si S Cl K Ca Ti V Mn Fe Br Rb Sr Y Zr) with an average at
least of 1600 cps (less for Br=1000)
Biogenic silica content mineralogy and grain size were measured every 4
cm Biogenic silica was measured following Mortlock and Froelich (1989) and
Bernaacuterdez et al (2005) using an Auto Analyzer Technicon AAII for dissolved silicate
analysis Mineralogy was analyzed with a Siemens D-500 X-ray diffractometer (Cu
kα 40 kV 30 mA graphite monochromator) at the ICTJA-CSIC (Spain) Grain size
analyses were performed in a Beckmann Coulter LS 13 320 Particle Size Analyzer
57
at the IPE-CSIC The samples were classified according to textural classes as
follows clay (lt2 μm) silt (20-2 microm) and sand (gt2 microm) fractions
The age-depth model for the Laguna Matanzas sedimentary sequence was
constructed using 210Pb and 137Cs dating techniques (MAT13-4A) as well as two 14C
AMS dates on bulk sediment samples (MAT11-6A fig 2) We dated the dissolved
inorganic carbon (DIC) in the water column and no significant reservoir effect is
present in the modern-day water column (10454 + 035 pcmc Table 2) An age-
depth model was obtained with the Bacon R package to estimate the deposition
rates and associated age uncertainties along the core (Blaauw and Christen 2011)
To estimate LUCC of Laguna Matanzasacutes watershed we use satellite images
Landsat MSS for 1975 Landsat TM for 1989 and Landsat OLI for 2016 all taken in
summer or autumn (Table 1) We performed supervised classification of land uses
(maximum likelihood algorithm) for each year (1975 1989 and 2016) and the results
were mapped using software ArcGIS 102 in 2017
Satellite Images Acquisition Date
Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat OLI 20160404 30 m
Table 1 Landsat imagery
Surface water samples were filtered for obtained particulate organic matter In
addition soil samples from the main land usecover present in the Laguna Matanzas
watershed were collected Elemental C N and their corresponding isotopes from
POM and soil were obtained at the LABASI and used here as end members
Daily precipitation at Santo Domingo (33ordm39`S 71ordm3 6`W)- the nearest weather
station to Laguna Matanzasndash was compiled using the redPrec R package (Fig1 d
Serrano-Notivoli et al 2017 Sarricolea et al 2019) To extend our precipitation
58
reconstruction back to 1824 we correlated this dataset with that available for
Santiago The Santiago data was compiled from data published in the Anales of
Universidad of Chile (1850) for the years 1824 to 1849 Almeyda (1940) for the years
1849 to 1864 and the Quinta Normal series from 1866 to the present (Direccioacuten
Meteoroloacutegica de Chile) We generated a linear regression model between the
presentday Santo Domingo station and the compiled Santiago data with a Pearson
coefficient of 087 and p-valuelt 001
Acknowledgments This research was funded by grants CONICYT AFB170008
to the Institute of Ecology and Biodiversity (IEB) and FONDECYT 1150763 (to CL)
Doctoral grant Becas Chile 21150224 MEDLANT (Spanish Ministry of Economy
and Competitiveness grant CGL2016-76215-R) Additional funding was provided
by the Laboratorio Internacional de Cambio Global (LINCGlobal PUC-CSIC) We
thank R Lopez E Royo and M Gallegos for help with sample analyses We thank
the Laboratory of Biogeochemistry and Applied Stable Isotopes (LABASI) of the
Department of Ecology (PUC) for sample analyses
References
Albert F 1900 Las dunas o sean arenas volantes voladeros arenas muertas invasion de las arenas playas I meacutedanos del centro de Chile comprendiendo el litoral desde el liacutemite norte de la provincia de Aconcagua hasta el liacutemite sur de la de Arauco Memorias cientiacuteficas I Lit
Anderson NJ W D Fritz SC 2009 Holocene carbon burial by lakes in SW
Greenland Glob Chang Biol httpsdoiorg101111j1365-2486200901942x
Armesto J Villagraacuten C Donoso C 1994 La historia del bosque templado
chileno Ambient y Desarro 66 66ndash72 httpsdoiorg101007BF00385244
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A
59
historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Arnell NW Gosling SN 2013 The impacts of climate change on river flow
regimes at the global scale J Hydrol 486 351ndash364 httpsdoiorg101016JJHYDROL201302010
Bernaacuterdez P Prego R Franceacutes G Gonzaacutelez-Aacutelvarez R 2005 Opal content in
the Riacutea de Vigo and Galician continental shelf Biogenic silica in the muddy fraction as an accurate paleoproductivity proxy Cont Shelf Res httpsdoiorg101016jcsr200412009
Blaauw M Christen JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474 httpsdoiorg10121411-BA618
Brush GS 2009 Historical land use nitrogen and coastal eutrophication A
paleoecological perspective Estuaries and Coasts 32 18ndash28 httpsdoiorg101007s12237-008-9106-z
Camarero L Catalan J 2012 Atmospheric phosphorus deposition may cause
lakes to revert from phosphorus limitation back to nitrogen limitation Nat Commun httpsdoiorg101038ncomms2125
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego
R Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and environmental change from a high Andean lake Laguna del Maule central Chile (36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Catalan J Pla-Rabeacutes S Wolfe AP Smol JP Ruumlhland KM Anderson NJ
Kopaacuteček J Stuchliacutek E Schmidt R Koinig KA Camarero L Flower RJ Heiri O Kamenik C Korhola A Leavitt PR Psenner R Renberg I 2013 Global change revealed by palaeolimnological records from remote lakes A review J Paleolimnol httpsdoiorg101007s10933-013-9681-2
Chen Y Y Huang W Wang W H Juang J Y Hong J S Kato T amp
Luyssaert S (2019) Reconstructing Taiwanrsquos land cover changes between 1904 and 2015 from historical maps and satellite images Scientific Reports 9(1) 3643 httpsdoiorg101038s41598-019-40063-1
Contreras S Werne J P Araneda A Urrutia R amp Conejero C A (2018)
Organic matter geochemical signatures (TOC TN CN ratio δ 13 C and δ 15 N) of surface sediment from lakes distributed along a climatological gradient on the western side of the southern Andes Science of The Total Environment 630 878-888 httpsdoiorg101016jscitotenv201802225
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia UC
60
de V 2014 ELEMENTOS DE LA HISTORIA NATURAL DEL An Mus Hist Natulas Vaplaraiso 27 51ndash67
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-010-9453-1
Diebel M Vander Zanden MJ 201209 Nitrogen stable isotopes in streams
stable isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19 1127ndash1134 httpsdoiorg10189008-03271
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506 Espizua LE Pitte P 2009 The Little Ice Age glacier advance in the Central
Andes (35degS) Argentina Palaeogeogr Palaeoclimatol Palaeoecol 281 345ndash350 httpsdoiorg101016JPALAEO200810032
Figueroa JA Castro S A Marquet P A Jaksic FM 2004 Exotic plant
invasions to the mediterranean region of Chile causes history and impacts Invasioacuten de plantas exoacuteticas en la regioacuten mediterraacutenea de Chile causas historia e impactos Rev Chil Hist Nat 465ndash483 httpsdoiorg104067S0716-078X2004000300006
Fowler D Coyle M Skiba U Sutton MA Cape JN Reis S Sheppard
LJ Jenkins A Grizzetti B Galloway N Vitousek P Leach A Bouwman AF Butterbach-bahl K Dentener F Stevenson D Amann M Voss M 2013 The global nitrogen cycle in the twenty- first century Philisophical Trans R Soc B httpsdoiorghttpdxdoiorg101098rstb20130164
Frank D Reichstein M Bahn M Thonicke K Frank D Mahecha MD
Smith P van der Velde M Vicca S Babst F Beer C Buchmann N Canadell JG Ciais P Cramer W Ibrom A Miglietta F Poulter B Rammig A Seneviratne SI Walz A Wattenbach M Zavala MA Zscheischler J 2015 Effects of climate extremes on the terrestrial carbon cycle Concepts processes and potential future impacts Glob Chang Biol httpsdoiorg101111gcb12916
Fritz SC Anderson NJ 2013 The relative influences of climate and catchment
processes on Holocene lake development in glaciated regions J Paleolimnol httpsdoiorg101007s10933-013-9684-z
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
61
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS) implications for past sea level and environmental variability J Quat Sci 32 830ndash844 httpsdoiorg101002jqs2936
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892 httpsdoiorg101126science1136674
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924 httpsdoiorg104319lo20095430917
Garciacutea-Ruiz JM 2010 The effects of land uses on soil erosion in Spain A
review Catena httpsdoiorg101016jcatena201001001
Garciacutea-Ruiz JM Loacutepez-Moreno JI Lasanta T Vicente-Serrano SM
Gonzaacutelez-Sampeacuteriz P Valero-Garceacutes BL Sanjuaacuten Y Begueriacutea S Nadal-Romero E Lana-Renault N Goacutemez-Villar A 2015 Los efectos geoecoloacutegicos del cambio global en el Pirineo Central espantildeol una revisioacuten a distintas escalas espaciales y temporales Pirineos httpsdoiorg103989Pirineos2015170005
Garciacutea-Ruiz JM Nadal-Romero E Lana-Renault N Begueriacutea S 2013
Erosion in Mediterranean landscapes Changes and future challenges Geomorphology 198 20ndash36 httpsdoiorg101016jgeomorph201305023
Garreaud RD Vuille M Compagnucci R Marengo J 2009 Present-day
South American climate Palaeogeogr Palaeoclimatol Palaeoecol httpsdoiorg101016jpalaeo200710032
Gayo EM Latorre C Jordan TE Nester PL Estay SA Ojeda KF
Santoro CM 2012 Late Quaternary hydrological and ecological changes in the hyperarid core of the northern Atacama Desert (~21degS) Earth-Science Rev httpsdoiorg101016jearscirev201204003
Ge Y Zhang K amp Yang X (2019) A 110-year pollen record of land use and land
cover changes in an anthropogenic watershed landscape eastern China Understanding past human-environment interactions Science of The Total Environment 650 2906-2918 httpsdoiorg101016jscitotenv201810058
Goyette J Bennett EM Howarth RW Maranger R 2016 Global
Biogeochemical Cycles 1000ndash1014 httpsdoiorg1010022016GB005384Received
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and
oxygen isotope fractionation during dissimilatory nitrate reduction by
62
denitrifying bacteria 53 2533ndash2545 httpsdoiorg10230740058342
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
Gurnell AM Petts GE Hannah DM Smith BPG Edwards PJ Kollmann
J Ward J V Tockner K 2001 Effects of historical land use on sediment yield from a lacustrine watershed in central Chile Earth Surf Process Landforms 26 63ndash76 httpsdoiorg1010021096-9837(200101)261lt63AID-ESP157gt30CO2-J
Heathcote A J et al Large increases in carbon burial in northern lakes during the
Anthropocene Nat Commun 610016 doi 101038ncomms10016 (2015) Hollander DJ Smith MA 2001 Microbially mediated carbon cycling as a
control on the δ13C of sedimentary carbon in eutrophic Lake Mendota (USA) New models for interpreting isotopic excursions in the sedimentary record Geochim Cosmochim Acta httpsdoiorg101016S0016-7037(00)00506-8
Holtgrieve GW Schindler DE Hobbs WO Leavitt PR Ward EJ Bunting
L Chen G Finney BP Gregory-Eaves I Holmgren S Lisac MJ Lisi PJ Nydick K Rogers LA Saros JE Selbie DT Shapley MD Walsh PB Wolfe AP 2011 A coherent signature of anthropogenic nitrogen deposition to remote watersheds of the Northern Hemisphere Science (80) 334 1545ndash1548 httpsdoiorg101126science1212267
Hooper DU Adair EC Cardinale BJ Byrnes JEK Hungate BA Matulich
KL Gonzalez A Duffy JE Gamfeldt L Connor MI 2012 A global synthesis reveals biodiversity loss as a major driver of ecosystem change Nature httpsdoiorg101038nature11118
Horvatinčić N Sironić A Barešić J Sondi I Krajcar Bronić I Borković D
2016 Mineralogical organic and isotopic composition as palaeoenvironmental records in the lake sediments of two lakes the Plitvice Lakes Croatia Quat Int httpsdoiorg101016jquaint201701022
Howarth RW 2004 Human acceleration of the nitrogen cycle Drivers
consequences and steps toward solutions Water Sci Technol httpsdoiorg1010382Fscientificamerican0490-56
Jenny B Valero-Garceacutes BL Urrutia R Kelts K Veit H Appleby PG Geyh
M 2002 Moisture changes and fluctuations of the Westerlies in Mediterranean Central Chile during the last 2000 years The Laguna Aculeo record (33deg50primeS) Quat Int 87 3ndash18 httpsdoiorg101016S1040-
63
6182(01)00058-1
Jenny B Wilhelm D Valero-Garceacutes BL 2003 The Southern Westerlies in
Central Chile Holocene precipitation estimates based on a water balance model for Laguna Aculeo (33deg50primeS) Clim Dyn 20 269ndash280 httpsdoiorg101007s00382-002-0267-3
Leng M J Lamb A L Heaton T H Marshall J D Wolfe B B Jones M D
amp Arrowsmith C 2006 Isotopes in lake sediments In Isotopes in palaeoenvironmental research (pp 147-184) Springer Dordrecht
Le Roux JP Nielsen SN Kemnitz H Henriquez Aacute 2008 A Pliocene mega-
tsunami deposit and associated features in the Ranquil Formation southern Chile Sediment Geol 203 164ndash180 httpsdoiorg101016jsedgeo200712002
Lehmann M Bernasconi S Barbieri a McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66 3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Lenihan JM Drapek R Bachelet D Neilson RP 2003 Climate change
effects on vegetation distribution carbon and fire in California Ecol Appl httpsdoiorg101890025295
McLauchlan K K Gerhart L M Battles J J Craine J M Elmore A J
Higuera P E amp Perakis S S (2017) Centennial-scale reductions in nitrogen availability in temperate forests of the United States Scientific reports 7(1) 7856 httpsdoi101038s41598-017-08170-z
Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32ordmS 3050 masl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
Martiacuten-Foreacutes I Casado MA Castro I Ovalle C Pozo A Del Acosta-Gallo
B Saacutenchez-Jardoacuten L Miguel JM De 2012 Flora of the mediterranean basin in the chilean ESPINALES Evidence of colonisation Pastos 42 137ndash160
Marzecovaacute A Mikomaumlgi a Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54 httpsdoiorg103176eco2011105
Matesanz S Valladares F 2014 Ecological and evolutionary responses of
Mediterranean plants to global change Environ Exp Bot httpsdoiorg101016jenvexpbot201309004
64
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Menzel P Gaye B Wiesner MG Prasad S Stebich M Das BK Anoop A
Riedel N Basavaiah N 2013 Influence of bottom water anoxia on nitrogen isotopic ratios and amino acid contributions of recent sediments from small eutrophic Lonar Lake central India Limnol Oceanogr 58 1061ndash1074 httpsdoiorg104319lo20135831061
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Michelutti N Wolfe AP Cooke CA Hobbs WO Vuille M Smol JP 2015
Climate change forces new ecological states in tropical Andean lakes PLoS One httpsdoiorg101371journalpone0115338
Muntildeoz-Rojas M De la Rosa D Zavala LM Jordaacuten A Anaya-Romero M
2011 Changes in land cover and vegetation carbon stocks in Andalusia Southern Spain (1956-2007) Sci Total Environ httpsdoiorg101016jscitotenv201104009
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Poraj-Goacuterska A I Żarczyński M J Ahrens A Enters D Weisbrodt D amp
Tylmann W (2017) Impact of historical land use changes on lacustrine sedimentation recorded in varved sediments of Lake Jaczno northeastern Poland Catena 153 182-193 httpdxdoiorg101016jcatena201702007
Pongratz J Reick CH Raddatz T Claussen M 2010 Biogeophysical versus
biogeochemical climate response to historical anthropogenic land cover change Geophys Res Lett 37 1ndash5 httpsdoiorg1010292010GL043010
Prudhomme C Giuntoli I Robinson EL Clark DB Arnell NW Dankers R
Fekete BM Franssen W Gerten D Gosling SN Hagemann S Hannah DM Kim H Masaki Y Satoh Y Stacke T Wada Y Wisser D 2014 Hydrological droughts in the 21st century hotspots and uncertainties from a global multimodel ensemble experiment Proc Natl Acad Sci httpsdoiorg101073pnas1222473110
65
Ruumlhland KM Paterson AM Smol JP 2015 Lake diatom responses to
warming reviewing the evidence J Paleolimnol httpsdoiorg101007s10933-
015-9837-3
Sarricolea P Meseguer-Ruiz Oacute Serrano-Notivoli R Soto M V amp Martin-Vide
J (2019) Trends of daily precipitation concentration in Central-Southern Chile Atmospheric research 215 85-98 httpsdoiorg101016jatmosres201809005
Sernageomin 2003 Mapa geologico de chile version digital Publ Geol Digit Serrano-Notivoli R de Luis M Begueriumlia S 2017 An R package for daily
precipitation climate series reconstruction Environ Model Softw httpsdoiorg101016jenvsoft201611005
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Stine S 1994 Extreme and persistent drought in California and Patagonia during
mediaeval time Nature 369 546ndash549 httpsdoiorg101038369546a0
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL
Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Syvitski JPM Voumlroumlsmarty CJ Kettner AJ Green P 2005 Impact of humans
on the flux of terrestrial sediment to the global coastal ocean Science 308(5720) 376-380 httpsdoiorg101126science1109454
Thomas RQ Zaehle S Templer PH Goodale CL 2013 Global patterns of
nitrogen limitation Confronting two global biogeochemical models with observations Glob Chang Biol httpsdoiorg101111gcb12281
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Urbina Carrasco MX Gorigoitiacutea Abbott N Cisternas Vega M 2016 Aportes a
la historia siacutesmica de Chile el caso del gran terremoto de 1730 Anuario de Estudios Americanos httpsdoiorg103989aeamer2016211
Vargas G Ortlieb L Chapron E Valdes J Marquardt C 2005 Paleoseismic
inferences from a high-resolution marine sedimentary record in northern Chile
66
(23degS) Tectonophysics httpsdoiorg101016jtecto200412031 399(1-4) 381-398httpsdoiorg101016jtecto200412031
Valero-Garceacutes BL Latorre C Morelliacuten M Corella P Maldonado A Frugone M Moreno A 2010 Recent Climate variability and human impact from lacustrine cores in central Chile 2nd International LOTRED-South America Symposium Reconstructing Climate Variations in South America and the Antarctic Peninsula over the last 2000 years Valdivia p 76 (httpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-yearshttpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-years
Vicente-Serrano SM Gouveia C Camarero JJ Begueria S Trigo R
Lopez-Moreno JI Azorin-Molina C Pasho E Lorenzo-Lacruz J Revuelto J Moran-Tejeda E Sanchez-Lorenzo A 2013 Response of vegetation to drought time-scales across global land biomes Proc Natl Acad Sci httpsdoiorg101073pnas1207068110
Villa Martiacutenez R P 2002 Historia del clima y la vegetacioacuten de Chile Central
durante el Holoceno Una reconstruccioacuten basada en el anaacutelisis de polen sedimentos microalgas y carboacuten PhD
Villa-Martiacutenez R Villagraacuten C Jenny B 2003 Pollen evidence for late -
Holocene climatic variability at laguna de Aculeo Central Chile (lat 34o S) The Holocene 3 361ndash367
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Human alteration of the global nitrogen cycle sources and consequences Ecological applications 7(3) 737-750 httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Rein B Urrutia R amp Appleby P (2009) A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 The Holocene 19(6) 873-881 httpsdoiorg1011770959683609336573
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Widory D Kloppmann W Chery L Bonnin J Rochdi H Guinamant JL
2004 Nitrate in groundwater An isotopic multi-tracer approach J Contam Hydrol 72 165ndash188 httpsdoiorg101016jjconhyd200310010
67
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
68
Supplementary material
Facie Name Description Depositional Environment
F1 Organic-rich
mud
Massive to banded black
organic - rich (TOC up to 14 )
mud with aragonite in dm - thick
layers Slightly banded intervals
contain less OM (TOClt4) and
aragonite than massive
intervals High MnFe (oxic
bottom conditions) High CaTi
BrTi and BioSi (up to 5)
Distal low energy environment
high productivity well oxygenated
and brackish waters and relative
low lake level
F2 Massive to
banded silty clay
to fine silt
cm-thick layers mostly
composed by silicates
(plagioclase quartz cristobalite
up to 65 TOC mean=23)
Some layers have relatively high
pyrite content (up to 25) No
carbonates CaTi BrTi and
BioSi (mean=48) are lower
than F1 higher ZrTi (coarser
grain size)
Deposition during periods of
higher sediment input from the
watershed
69
F3 Banded to
laminated light
brown silty clay
cm-thick layers mostly
composed of clay minerals
quartz and plagioclase (up to
42) low organic matter
(TOC mean=13) low pyrite
and BioSi content
(mean=46) and some
aragonite
Flooding events reworking
coastal deposits
F4 Laminated
coarse silts
Thin massive layers (lt2mm)
dominated by silicates Low
TOC (mean=214 ) BrTi
(mean=002) MnFe (lt02)
TIC (lt034) BioSi
(mean=46) and TS values
(lt064) and high ZrTi
Rapid flooding events
transporting material mostly
from within the watershed
F5 Breccia with
coarse silt
matrix
A 17 cm thick (80-97 cm
depth) layer composed by
irregular mm to cm-long ldquosoft-
clastsrdquo of silty sediment
fragments in a coarse silt
matrix Low CaTi BrTi and
MnFe ratios and BioSi
Rapid high energy flood
events
70
(mean=43) and high ZrTi
(gt018)
Table Sedimentological and compositional characteristics of Laguna Matanzas
facies
71
CAPIacuteTULO 2 STABLE ISOTOPES TRACK LAND USE AND COVER
CHANGES IN A MEDITERRANEAN LAKE IN CENTRAL CHILE OVER THE
LAST 600 YEARS
72
Stable isotopes track land use and cover changes in a mediterranean lake in
central Chile over the last 600 years
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugone-Aacutelvarezabc Pablo
Sarricolead Leonardo Villaciacutese M Laura Carrevedob Blas Valero-Garceacutescg
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Departamento de Ecologiacutea Universidad de ChileLas Palmeras 3425 Nuntildeoa Santiago Chile
f Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding authors E-mail addresses clatorrebiopuccl magdalenafuentealbagmailcom
Key words Anthropocene lake sediment δsup1⁵N δsup1sup3C Great Acceleration organic
geochemistry watershedndashlake system Stable Isotope Analyses land usecover
change Nitrogen cycle mediterranean ecosystems central Chile
73
Abstract
Nutrient fluxes in many aquatic ecosystems are currently being overridden by
anthropic controls especially since the industrial revolution (mid-1800s) and the
Great Acceleration (mid-1900s) Land use and cover changes (LUCC) alter the
availability and fluxes of nutrients such as nitrogen that are transferred via runoff
and groundwater into lakes By altering lake productivity and trophic status these
changes are often preserved in the sedimentary record Here we use
geolimnological proxies and stable isotope analyses (δsup1⁵N δsup13C) on lake sediments
to reconstruct the lake-watershed nutrient transfer in a central Chilean lake (Lago
Vichuqueacuten) over the past 600 years We compare our multiproxy record from recent
lake sediments to the soilvegetation relationship across the watershed as well as
land usecover changes from 1975 to 2014 derived from satellite images Our results
show that lake sediment δsup1⁵N values increased with meadow cover but decreased
with tree plantations suggesting increased nitrogen retention when trees dominate
the watershed Our δsup1⁵N record from central Chile can thus be interpreted as a proxy
for nutrient availability over the last 600 years mainly derived from land use changes
coupled with climate drivers Although variable sources of organic matter and in situ
fractionation often hinder straightforward environmental interpretations of stable N
isotope signatures our study shows that lake sediment δsup1⁵N can be a useful tool for
assessing the contribution of past human activities in nutrient and nitrogen cycling
1 Introduction
Nitrogen (N) is a limiting nutrient in terrestrial and aquatic ecosystems (Vitousek
et al 1997) Changes in its availability can drive eutrophication and increase
pollution in these ecosystems (McLauchlan et al 2013) Although recent human
74
impacts on the global N cycle have been significant the consequences of increased
anthropogenic N on these ecosystems are unclear (Gerhart and Mclauchlan 2014
Battye et al 2017) Isotopic signatures are key to understand N dynamics in lakes
nevertheless in situ andor diagenetic fractionation along with multiple sources of
organic matter (OM) often hinder straightforward environmental interpretations from
isotopes Monitoring δ15N and δ13C values as components of the N cycle
specifically those related to the link between terrestrial and aquatic ecosystems can
help differentiate between effects from processes versus sources in stable isotope
values (eg from Particulate Organic Matter -POM- soil and vegetation) and
improve how we interpret variations in δ15N (and δ13C) values at longer temporal
scales
The main processes controlling stable N isotopes in bulk lake OM are source
lake productivity diagenesis and denitrification (Talbot 2001 Leng et al 2006
Fariacuteas et al 2009 Torres et al 2012 Brahney et al 2014) N sources depend on
contributions from the watershed (ie soil and biomass) the transfer of atmospheric
N to the lake (ie atmospheric N2 fixation) and nutrient recycling (Leng et al 2006)
Atmospheric N2 (isotopically defined as δ15N =0) is fixed by cyanobacteria with
minimal fractionation resulting in δ15NOM values that fluctuate from -2 to +2permil (Fogel
and Cifuentes 1993 Talbot 2001 Elliot and Brush 2006) Besides N2 fixation by
cyanobacteria phytoplankton species assimilate dissolved inorganic nitrogen (DIN)
and values typically range from +7 to +10permil (Xu et al 2016 Meyers et al 2003) In
addition seasonal changes in POM occur in the lake water column Gu et al (2006)
sampled the water column of Lake Wauberg (Florida USA) during a hydrologic year
and found a higher development of N fixing species during the summer A major
factor behind this increase are human activities in the watershed which control the
75
inputs of the sediment and nutrients to the lake (Woodward et al 2011) Some
studies have shown higher δ15N values in lake sediments from watersheds that are
highly affected by human activities (Bruland and Mackenzie 2010 Woodward et al
2011) Syntenic fertilizers displayed δ15N values between -4 to +4permil cattle manure
around +8 to +22permil and surface and groundwater OM between 0 to +10permil (Elliott
and Brush 2006 Leng et al 2006) Although relatively low δ15N values from
fertilizers constitute major N input to human-altered watersheds the elevated loss
of 14N via volatilization of ammonia and denitrification leaves the remaining total N
input enriched in 15N (Bruland and Mackenzie 2010)
In addition to the different sources and variations in lake productivity early
diagenesis at the sedimentndashwater interface in the sediment can further alter
sediment OM isotope values (Brahney et al 2014 Galman et al 2009) During
diagenetic loss of N in lacustrine systems kinetic fractionation occurs and the
remaining N is enriched in 15N leading to higher δ15N values (Galman et al 2006
Talbot 2001) Denitrification tends to enrich lake sediments in 15N due to the
assimilation of nitrates by denitrifying bacteria (Granger et al 2008) and is more
prevalent in anoxic environments (Brahney et al 2016 Lehman et al 2002)
Carbon isotopes in lake sediments can also provide useful information about
paleoenvironmental changes OM origin and depositional processes (Meyers et al
2003) Allochthonous organic sources (high CN ratios) produce isotope values
similar to values from catchment vegetation Autochthonous organic matter (low CN
ratio) is influenced by fractionation both in the lake and the watershed leading up to
carbon assimilation by lacustrine biota (Woodward et al 2011) Changes in
productivity greatly affect δ13COM values (Torres et al 2012) as during C uptake
plankton preferentially assimilate 12C leaving the dissolved inorganic carbon (DIC)
76
pool enriched in 13C (Gu et al 2006) with phytoplankton averaging ca 20 permil lower
than the respective δ13CDIC values (Leng et al 2006) Under conditions of low to
moderate primary productivity plankton preferentially uptake the lighter 12C
resulting in low δ13C values displayed by the OM (Torres et al 2012) Conversely
during high primary productivity phytoplankton will uptake 12C until its depletion and
is then forced to assimilate the heavier isotope resulting in an increase in δ13C
values Higher productivity in C-limited lakes due to slow water-atmosphere
exchange of CO2 also results in high δ13C values (Galman et al 2009) In these
cases algae are forced to uptake dissolved bicarbonate with δ13C values between
7 to 10permil higher than those of the atmospheric CO2 dissolved in water (Sun et al
2016 Torres et al 2012 Galman et al 2009)
Stable isotope analyses from lake sediments are thus useful tools to
reconstruct shifts in lake-watershed dynamics caused by changes in limnological
parameters and LUCC Our knowledge of the current processes that can affect
stable isotope signals in a watershed-lake system is limited however as monitoring
studies are scarce Besides in order to use stable isotope signatures to reconstruct
past environmental changes we require a multiproxy approach to understand the
role of the different variables in controlling these values Hence in this study we
carried out a detailed survey of current N dynamics in a coastal central Chilean lake
(Lago Vichuqueacuten) and a multiproxy study of a sediment record spanning the last
600 years The characterization of the recent changes in the watershed since 1970s
is based on satellite images to compare recent changes in the lake and assess how
these are related with climate variability and an ever increasing human footprint
(Lara et al 2012 Gallardo et al 2015 Steffen et al 2015) Our main goal is to
investigate how stable isotope values from lake sediment reflect changes in the lake
77
ndash watershed system during periods of high watershed disruption (eg Spanish
Conquest late XIX century Great Acceleration) and recent climate change (eg
Little Ice Age and current global warming)
2 Study Site
Lago Vichuqueacuten (34ordm49`S 72ordm04W 4 m asl 30 m of depth Fig 1) is a
mesotrophic to eutrophic lake with dominant anoxic bottom conditions that is
stratified year around (Frugone-Aacutelvarez et al 2017) The main inlets are the
Vichuqueacuten and Huintildee rivers and the only outlet is the Llico estuary that drains into
the Pacific Ocean High tides can sporadically shift the flow direction of the Llico
estuary which increases the marine influence in the lake Dune accretion gradually
limited ocean-lake connectivity until the estuary was almost completely closed off
by ca 10 ky BP and the current lake regime formed (Frugone-Aacutelvarez et al 2017)
The area is characterized by a mediterranean climate with cold-wet winters and
hot-dry summers and an annual precipitation of ~650 mm and a mean annual
temperature of 15ordmC During the austral winter months (June - August) precipitation
is modulated by north-west shifts of the South Pacific Anticyclone (SPA) followed by
an increased frequency of storm fronts stemming off the South Westerly Winds
(SWW) A strengthened SPA during austral summers (December - March) which
are typically dry and warm blocks the northward migration of storm tracks stemming
off the SWW (Fig1 Frugone-Aacutelvarez et al 2017)
78
Figure 1 a) The location of the Lago Vichuqueacuten in central Chile b) Map of land
uses and cover in 2016 c) Ombrothermic diagram mediterranean climates are
characterized by cold-wet winters with surplus moisture from June to August and
hot-dry summers d) Lake bathymetry showing location of cores and water sampling
sites used in this study
Although major land cover changes in the area have occurred since 1975 to the
present as the native forests were replaced by tree (Monterey pine and eucalyptus)
plantations the region was settled before the Spanish conquest (Frugone-Alvarez
et al 2018) Historical documents indicate that Vichuqueacuten was occupied by a
Mitimaes colony as part of the Inka empire (Odone 1998) Unlike other Andean
areas during the Inka period the introduction of corn and quinoa in the Vichuqueacuten
watershed do not seem to have intensified land use The Spanish colonial period in
Chile lasted from 1542 CE to the independence in 1810 CE The first historical
document (1550 CE) shows that the areas around Vichuqueacuten were settled by the
Spanish early on as the Vichuqueacuten watershed was given as an ldquoencomiendardquo
system to Juan Cuevas The ldquoencomiendardquo system provided the owners with land
79
and indigenous people to work but also the introduction of wheat wine cattle
grazing and logging of native forests for lumber extraction and increasing land for
agricultural activities (Odone et al 1998 Armesto et al 2010) During the 19th
century (the Republic) the export of wheat to Australia and Canada generated
intensive changes in land cover use The town of Vichuqueacuten became the regional
capital and plans to build a maritime port in the bay of Vichuqueacuten were drawn
However the fall of international markets in 1880 paralyzed these plans During the
20th century the plantation of trees (Pinus radiata and Eucalyptus globulus) in areas
cleared of native forests was subsidized by two national laws DFL nordm265 (1931) and
DFL nordm 701 (1974) both of which provided funds for such plantations During the last
decades the urbanization with summer vacation homes along the shorelines of
Lago Vichuqueacuten has greatly increased Extensive lake eutrophication is currently a
large environmental problem (EULA 2008)
3 Methods
Coring campaigns were organized from 2011 to 2015 (Fig 1d) We recovered
12 short cores and 4 long cores (up to 14 m long) at several sites using a hammer-
modified UWITEC gravity corer Sediment cores (VIC11-2A 170 cm VIC13-2B 170
cm VIC13-2D 1200 cm and VIC15-1A 119 cm) were split photographed sub-
sampled and stored in the Pyrenean Institute of Ecology (IPE-CSIC Spain) Core
VIC13-2B was selected for detailed multiproxy analyses (including elemental
geochemistry C and N isotopes XRF and 14C dating) Stable isotope analyses
(δ13C δ15N) were performed at the Laboratory of Biogeochemistry and Applied
Stable Isotopes (LABASI-PUC Chile) Sediment samples for δ13Corg were pre-
treated with 50 ml of HCl and 50 ml of deionized water and dried at 60ordmC for 4 hr to
remove carbonates (Harris et al 2011) Isotope analyses were conducted using a
80
Delta V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via
a Conflo IV interface Isotope results are expressed in standard delta notation (δ)
and expressed in per mil (permil) relative to the Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Total carbon (TC)
Total Organic Carbon (TOC) Total Inorganic Carbon (TIC) and Total Sulfur (TS)
were measured every 2 cm with a LECO SC 144 DR at the IPE-CSIC
An AVAATECH X-Ray Fluorescence II core scanner with a Rh X-ray tube from
the University of Barcelona was used to obtain XRF logs every 4 mm of resolution
Results are expressed as element intensities in counts per second (cps) Tube
voltage was operated at 30 kV and 10 kV to obtain the abundances of 18 elements
(Pb Al Si S Cl K Ca Ti V Mn Fe Co Zn Br Rb Sr Y Zr) with an average of
at least of 1800 cps (less for Pb=437 and Zn=934 cps) Pb was included due to
similar behavior with Co and Fe Element ratios were calculated to describe changes
in redox conditions (MnFe) endogenic productivity (BrTi) carbonate formation
(SrCa and CaTi) and sediment input (ZrTi) (Barreiro-Lostres et al 2014
Carrevedo et al 2015 Frugone-Aacutelvarez et al 2017 Kylander et al 2011 Moreno
et al 2007a)
Several campaigns were carried out to sample the POM from the water column
two per hydrologic year from November 2015 to August 2018 A liter of water was
recovered in three sites through to the lake two are from the shallower areas (with
samples taken at 2 and 5 m depth at each site) and one in the deeper central portion
(at 2 5 and 20 m depth) of the lake (Fig 1d) The samples were filtered using glass
fiber filters (25 m) and then frozen to obtain the stable carbon and nitrogen isotope
signal of lacustrine POM Additionally soil and vegetation samples from the
following communities native species meadow hydrophytic vegetation and
81
Monterrey pine (Pinus radiata) were taken for stable isotope analyses (see in
supplementary material)
The age model for the complete Lago Vichuqueacuten sedimentary sequence is
based on Frugone-Aacutelvarez et al (2017) with the last 1000 years based on
210Pb137Cs dating techniques in VIC15-1A and 14C AMS dates on bulk sediment
samples (Supplementary Table S1) The 14C measurements of lake water DIC show
a moderate reservoir effect of 180 plusmn 25 14C yr) The revised age-depth model used
here includes three more 14C AMS dates performed with the program Bacon to
establish the deposition rates along the core (Blaauw and Christen 2011 Fig 2)
The age-depth model indicates that average resolution between 0 to 87 cm is lt2
cm per year and from 88 to 170 cm it is lt47 cm per year
82
Figure 2 Chronological age-depth model for the Lago Vichuqueacuten sedimentary
sequence spanning the last 1000 yr (see also Frugone-Alvarez et al 2017)
To estimate land use changes in the watershed we use Landsat MSS images
for 1975 and Landsat TM for 1989 and 2014 all taken in either summer or autumn
(Table 1) We performed supervised classification of land uses (maximum likelihood
83
algorithm) for each year (1975 1989 and 2014) and results were mapped using
ArcGIS 102
Table 1 Images using for LUCC reconstruction
Source of LUCC
Acquisition
Date Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat TM 19991226 30 m
CONAF 2009 30 m
Land cover Chile 2014 30 m
CONAF 2016 30 m
Previous Work on Lago Vichuqueacuten sedimentary sequence
The sediments are organic-poor dark brown to brown laminated silt with some
intercalated thin coarser clastic layers Lacustrine facies have been classified
according to elemental composition (TOC TS TIC and TN) grain size and
sedimentary textures (Frugone-Aacutelvarez et al 2017) Four main types of lacustrine
facies were identified in this short core Facies L1 is a laminated (1cm) black to dark
brown diatom-poor silt with relatively higher TOC percentages (TOC about 185)
TS (about of 18) and TOCTN (about 92) Facies L2 is characterized by a
homogeneous black organic-poor silty clay with low TOC (mean= 13) TS (mean=
13) TOCTN (mean= 88) values Facies L3 is a laminated (1cm) black organic-
poor (TOC mean 14) silts with higher TS (mean= 21) and TOCTN ratios
(mean= 88) Facies L3 is interpreted as ldquobaselinerdquo deposition on the distal areas
84
of Vichuqueacuten lake Mineralogy is dominated by mica (muscovite) clay minerals
(chlorite) and plagioclase (albite) Quartz and calcite occur at the top and pyrite
occurs in the lower part of the sequence Facies T is composed by massive banded
sediments 4 cm thick and coarse grain size It is interpreted as an instantaneous
depositional event (for more detail see Frugone-Aacutelvarez et al 2017) In this work
we identified four subunits based on geochemical and stable isotope signals
4 Results
41 Geochemistry and PCA analysis
High positive correlations exist between Al Si K and Ti (r = 078 ndash 096
supplementary Fig S1) and between Zn Zr Rb and Y (r = 072 - 096) which reflect
the silicate content of the watershed-derived sediments (Marzecoacuteva et al 2011) Zr
is commonly associated with minerals more abundant in coarser deposits Thus the
ZrTi profile could reflect grain size (Cuven et al 2010) showing higher variability
in the upper part of the Lake Vichuqueacuten sequence and in the alternation between
facies L1 and L2 Another group of elements made up of Co Fe and Pb displayed
positive correlations (r = 067ndash 097) and represents the input of heavy metals Br
Cl and Ca show a positive correlation (r = 049 ndash 06 p-value = 00001) BrTi ratio
is interpreted as a productivity indicator due to Br having a strong affinity with humic
and organic compounds (Marzecoacuteva et al 2011 Frugone-Aacutelvarez et al 2017) In
our Lago Vichuqueacuten sequence BrTi values are high from 170 to 132 cm and from
36 cm to the top of core where it shows several sharps peaks (Fig 3) The MnFe
ratio is indicative of lake bottom oxygenation (Naecher et al 2013) as under
reducing conditions Mn tends to become more mobile than Fe leading to a
decreased MnFe (Marzecoacuteva et al 2011) Several sharp peaks in MnFe occurred
85
from 127 to 120 cm and from 57 to 30 cm (MgFegt03) The silicate group and the
Br Cl Ca Mn group are negatively correlated (r= -012 and -066)
Principal Component Analysis (PCA) was undertaken on the XRF
geochemical data to investigate the main factors controlling sediment deposition in
Lago Vichuqueacuten (Fig 3) The four main eigenvectors explain 801 of the variance
(supplementary material Table S2) The principal component (PC1) explain 437
of the total of variance and grouped elements are associated with terrigenous input
to the lake Positive values of the biplot have been attributed to higher heavy metals
deposition (Pb Co and Fe) and negative values to higher silicate input (Al K Ti and
Si) in a high energy catchment environment (Zr) The PC2 explains 198 of the
total of variance and highlights the endogenic productivity in the lake The positive
loading is compound by Br Cl Ca and Sr and to a lesser extent by Y Zn Rb and
Mn The PC2 may explain both the precipitation of carbonates (Ca Sr) and biological
production (Br)
86
Figure 3 Principal Component Analysis of XRF geochemical measurements in
VIC13-2B Lago Vichuqueacuten lake sediments
42 Sedimentary units
Based on geochemical and stable isotope analysis we identified four
lithological subunits in the short core sedimentary sequence Our PCA analyses and
Pearson correlations pointed out which variables were better for characterizing the
subunits (Fig 4) in terms of endogenic productivity heavy metals and terrestrial
input The unit 2c (170-131 cm) is characterized by the occurrence of some clastic
layers (L2 from 160 and 150 cm) high δsup1⁵N values (mean= +59 plusmn 04permil) with
Magnetic Susceptibility (MS) BrTi ZrTi and SrCa values increase towards the top
Also it has relatively low δsup1sup3C values (mean= -265 plusmn 11permil) and CNmolar ratios
(mean=78 plusmn 04) The ZrTi show upwards increasing values reaching the highest
values of the sequence at the top of this unit suggesting a coarsening upward trend
and relatively higher depositional energy The MS trend also indicates higher
erosion in the watershed and enhanced delivery of ferromagnetic minerals likely
from soil particles particularly from 150 to 160 cm (Frugone-Aacutelvarez at al 2017)
The subunit 2b (130-118 cm) is also composed of black silts but it has the
lowest MS values of the whole sequence and its onset is marked by a sharp
decrease in ZrTi This subunit is marked by sharp peaks of MnFe (at 126 and 120
cmgt027) SrCa (at 125 and 120 cmgt033) CaTi (at 124 and 120 cmgt199) TOC
(about 26 at 130 124 122 and 118 cm) and δsup1⁵N (gt+67permil at 126 and 120 cm)
BrTi oscillates around low values (about 01) whereas the δsup1sup3C values range
between -262 and -282permil
87
The unit 2a (58-117 cm) shows increasing and then decreasing MS values
and displayed sharps peaks of δsup1⁵N (gt64 at 102 to 99 87 and 75 cm) and CN
(about 10 at 86 74 66 and 60 cm) SrCa (mean = 04 plusmn 013) BrTi (mean = 008
plusmn 002) MnFe (mean = 002 plusmn 001) and δsup1sup3C (mean = -260 plusmn 07permil) oscillate in
low values The upper part of this unit (72 ndash 57 cm) shows an increase of SrCa
(from 03 to 05)
The upper unit 1a (0 - 57 cm) starts with a fine clastic layer (T at 57 to 54
cm) interpreted as deposition during a high-energy event It is characterized by
lower BrTi (mean = 003 plusmn 002) TOC (mean = 12 plusmn 03) and δsup1sup3C (mean= -
266 plusmn 06permil) higher δsup1⁵N (mean = +55 plusmn 08 permil) This unit is made of alternating
fine (lt 1cm) black and dark brown laminae with carbonate (L1 facies) Recently
deposited sediments show decreasing MS values increasing TOC (mean= 15 plusmn
04 ) sharp peaks of BrTi (gt015 at 33 21 5 and 1 cm) and the lowest δsup1⁵N values
of the entire record (mean= 45 plusmn 06 permil) and δsup1sup3C values (mean= -277 plusmn 09 permil)
Heavy metal content increases abruptly in the upper section of the core (2 - 20 cm)
(peaks of FeTi CoTi and PbTi)
88
Figure 4 Sedimentary units of Lago Vichuqueacuten sedimentary sequence Selected
variables illustrate watershed input (Magnetic Susceptibility ZrTi CoTi and MnFe)
endogenic productivity (SrCa CaTi TS) organic matter source (BrTi TOC
CNmolar and stable isotope records (δ13Corg and δ15Nbulk)
43 Recent seasonal changes of particulate organic matter on water column
The average of δ15NPOM from the three sites across to the lake was +86 plusmn 58
permil oscillating widely from -14 to +193permil (Fig 5) Important seasonal differences
occur at 2 and 5 m depth with more negative values during summer (~53 plusmn 52 permil)
than winter (~122 plusmn 40permil) At 20 m depth δ15NPOM values exhibit small seasonal
ranges (mean = +10permil for winter and +87permil for summer) The δ13CPOM average was
-296 plusmn 33permil with slightly seasonal and water column depth differences However
more positive δ13CPOM values occurred during winter (mean=-290 plusmn 41permil) than in
summer (mean= -301 plusmn 19 permil) Like the δ15NPOM values CNPOM (mean= 83 plusmn 17)
displayed important seasonal and water depth differences Lower CNPOM ratios
89
occurred during summer at 2 and 5 m depth (mean= 95 plusmn 17) and higher and more
constant values during winter (mean = 73 plusmn 09) at the same depth At 20 m CNPOM
shows similar values in both winter (70) and summer (74)
Figure 5 Seasonal POM averages from samples taken of the Lago Vichuqueacuten
water column Samples were taken from 2015 to 2018 at 2 (n=16) 5 (n=14) and 20
(n=8) meters depth
44 Stable isotope values across the Lake Vichuqueacuten watershed
Figure 6 shows modern vegetation soil and sediment isotope values found for
the watershed Huintildee tributary and Vichuqueacuten lake The δsup1⁵NFoliar values from
meadow plantations and macrophytes have similar range values with a mean of
+89 plusmn50 permil (n=18) +74 plusmn 75 permil (n=5) +82 plusmn 36 permil (n=6) respectively Native
vegetation has overall lower δsup1⁵NFoliar values with a mean of 14 plusmn 21 permil (n=28 see
Table 2 in supplementary material) The δsup1sup3CFoliar from terrestrial samples exhibit
similar values across the different plant communities (tree plantation mean=-274 plusmn
13permil meadow mean = -275 plusmn 23 permil native species mean=-272 plusmn 14permil) whereas
macrophytes display slightly more negative values with a mean of -287 plusmn 23permil
Macrophytes substrate displayed the highest δsup1⁵N values with a mean of +60 plusmn
14permil (n=3) Similar and relatively high δsup1⁵N values for soil of plantation (mean= +54
plusmn 13permil n=4) meadow (mean= +49 plusmn 04permil n=8) and surface river sediment
90
(mean= +52 plusmn 17permil n=10) Native forest soils oscillate around slightly more
negative values with a mean of +45 plusmn 12permil (n=4) Slightly more positive soil δsup1sup3C
values occur both underneath native forests and in tree plantations with means of -
284 plusmn 23permil and -298 plusmn 13permil respectively compared to either meadow soils
(mean= -302 plusmn 11permil) aquatic sediments underneath macrophytes (-311 plusmn 10permil)
or from surface river sediments (mean= -312 plusmn 10permil)
Figure 6 Stable isotope content of modern soil river sediment and foliar vegetation
used as end members in the sedimentary sequence of Lago Vichuqueacuten a)
Scatterplot of foliar δsup1⁵N and δsup13C values for vegetation from Lago Vichuqueacuten
watershed (plantation meadow and native species) and macrophytes on Lake
Vichuqueacuten See supplementary material for more detail of vegetation types b)
Scatterplot of soil δsup1⁵N and δsup13C values across the watershed the sediments of the
Huintildee and Vichuqueacuten rivers and the fluvial sediment corresponding to the
macrophyte vegetation
45 Land use and cover change from 1975 to 2014
Major land use changes between 1975 CE and 2016 CE in the Lago
Vichuqueacuten watershed are summarized in Figure 7 The watershed has a surface
area of 535329 km2 of which native vegetation (26) and shrublands (53)
represent 79 of the total surface in 1975 Meadows are confined to the valley and
91
represent 17 of watershed surface Tree plantations initially occupied 1 of the
watershed and were first located along the lake periphery By 1989 the areas of
native forests shrublands and meadows had decreased to 22 31 and 14
respectively whereas tree plantations had expanded to 30 These trends
continued almost invariably until 2016 when shrublands and meadows reached 17
and 5 of the total areas while tree plantations increased to 66 Native forests
had practically disappeared by 1989 and then increased up to 7 of the total area
in 2016 (Fig 6) preferentially occupying the southeastern sector of the watershed
Figure 7 Changes in land use and cover between 1975 and 2016 in the Lago
Vichuqueacuten watershed as measured from satellite images The major change is
represented by the replacement of native forest shrubland and meadows by
plantations of Monterrey pine (Pinus radiata)
Figure 8 shows correlations between lake sediment stable isotope values and
changes in the soil cover from 1975 to 2013 Positive relationships occurred
between sediment δsup1⁵N and δsup13C values from core VIC13-2B (unit 1) and the
92
percentage of native forest (r=089 for δsup1⁵N 082 for δsup13C) shrublands (r = 071 for
δsup1⁵N 057 for δsup13C) and meadow cover (r = 088 for δsup1⁵N 081 for δsup13C) All of these
correlations are significant (p value lt 0001) In contrast significant negative
correlations (p lt0001) occurred between tree plantation cover and lake sediment
stable isotope values (δsup1⁵N r = -082 and δsup13C r = -067) native forest (r = -097)
meadows (r = -086) and shrubland (r =-093)
Figure 8 Correlation plots of land use and cover change versus lake sediment
stable isotope values The δsup1⁵N values are positively correlated with native forests
agricultural fields and meadow cover across the watershed Total Plantation area
increases are negatively correlated with native forest meadow and shrubland total
area Significance levels are indicated by the symbols p-values (0 0001 001
005 01 1) lt=gt symbols ( )
93
5 Discussion
51 Seasonal variability of POM in the water column
The stable isotope values of POM can vary during the annual cycle due to
climate and biologic controls namely temperature and length of the photoperiod
which affect phytoplankton growth rates and isotope fractionation in the water
column (Gu et al 2006 Leng et al 2006) POM from Lago Vichuqueacuten surface
samples (2 and 5 m depth) displayed lower δ13CPOM values during the summer than
in winter During C uptake phytoplankton preferentially utilize 12C leaving the
DICpool enriched in 13C Therefore as temperature increases during the summer
phytoplankton growth generates OM enriched in 12C until this becomes depleted
and then the biomas come to enriched u At the onset of winter the DICpool is now
enriched in 13C and despite an overall decrease in phytoplankton production the
OM produced will also be enriched in 13C In contrast δ13CPOM values at 20 m depth
did not reflect these seasonal differences probably due to water-column
stratification that maintains similar temperatures and biological activity throughout
the year
Lake N availability depends on N sources including inputs from the
watershed and the atmosphere (ie deposition of N compounds and fixation of
atmospheric N2) which varies during the hydrologic year The fixation of atmospheric
N2 is an important natural source of N to the lake occurring mainly during the
summer season associated with higher temperature and light (Gu et al 2006)
Because the δ15NPOM of N2 fixers is close to 0permil with almost no overall isotope
fractionation (Leng et al 2006) when N2 is the major N source δ15NPOM values are
typically low However when DIN concentrations are high or alternatively when little
94
N2 fixation occurs δ15NPOM values will be higher (Gu et al 2006) δ15NPOM values
from summer Lago Vichuqueacuten samples were lower than those from winter with large
differences between the seasons (~5permil Fig 5 and 9) Furthermore δ15NPOM values
were high when monthly average temperature was low and monthly precipitation
was high (Fig 9) This pattern is consistent with the dominance of high N2 fixation
by cyanobacteria associated with increased summer temperatures This correlation
of δ15NPOM values with temperature further suggests a functional group shift i e
from N fixers to phytoplankton that uptake DIN The correlation between wetter
months and higher δ15NPOM values could be caused by increased N input from the
watershed due to increased runoff during the winter season The lack of data of the
δ15NPOM between the 30 mm and the 160 mm of precipitation is due to the
mediterranean-type climate that concentrates precipitations in the winter months
Finally Lago Vichuqueacuten is characterized by pronounced seasonality leading to
higher phytoplankton biomass in summer characterized by low δ15NPOM In winter
low biomass production and increased input from watershed is associated to high
δ15NPOM
95
Figure 9 Seasonal changes can drive δ15NPOM values in Lago Vichuqueacuten Data
correspond to average monthly temperature and total monthly precipitation for the
months when the water samples were taken (years 2015 - 2018) P-valuelt005
52 Stable isotope signatures in the Lake Vichuqueacuten watershed
The natural abundance of 15N14N isotopes of soil and vegetation samples
from the Lago Vichuqueacuten watershed appear to result from a combination of factors
isotope fractionation different N sources for plants and soil microorganisms (eg N2
fixation Nitrates) depth in the soil where N uptake by vegetation occurs and N loss
mechanisms (ie denitrification leaching and ammonia volatilization Hogberg
1997) The lowest δsup1⁵Nfoliar values are associated with native species and are
probably due to the contribution of N-fixing species (eg Sophora) (Fig6 and for
more detail see Table S3 in supplementary material) The number native N-fixers
species present in the Chilean mediterranean vegetation are not well known
however so there could be other N-fixing species in this group Alternatively δsup1⁵Nfoliar
values reflect soil N uptake (Kahmen et al 2008) In environments limited by N
plants have δsup1⁵N values similar to the soil δsup1⁵N values however the denitrification
and volatilization of ammonia can lead to the remain N of soil to come enriched in
15N (Cifuentes et al 1989 Craine et al 2009 Bruland and Mckenzie 2010) Soil N
isotope samples from native species communities tends to display relatively high
δsup1⁵N values respect to foliar samples due to loss of N-soil
The higher foliar and soil δsup1⁵N values obtained from samples of meadows
aquatic macrophytes and tree plantations can be attributed to the presence of
greater amounts of nitrate available for plant uptake (Fig 6) Houmlgberg (1997)
suggests that the availability of different N sources in soils (ie nitrates versus
96
ammonia) with different residence times can also explain these δsup1⁵NFoliar values
Indeed Feigin et al (1974) described differences of up to 20permil between ammonia
and nitrates sources Denitrification and nitrification discriminate much more against
15N than N2 fixation does (Craine et al 2009) leading to soils (and plants after
uptake) enriched in 14N
In general multiple processes that affect the isotopic signal result in similar
δsup1⁵N values between the soil of the watershed and the sediments of the river
However POM isotope fluctuations allow to say that more negative δsup1⁵N values are
associated to lake productivity while more positive δsup1⁵N values are associated with
N input from the watershed
δ13C values in Lago Vichuqueacuten watershed reflect CO2 gas exchange between
C3 plants and algae with the atmosphere During photosynthesis plants
discriminate against the heavier stable isotope of carbon (13C) in favor of the lighter
isotope 12C which results in δ13CFoliar values between -220permil and -320permil (Browman
and Cook 2002 Liu et al 2007) Likewise the δ13CFoliar values from Lago Vichuqueacuten
oscillate around -27permil (Fig 6) Plants transform atmospheric CO2 into soil organic
carbon (C) which in turn reflects this initial discrimination against 13C during C
uptake and post photosynthetic fractionation (Farquhar et al 1982 1989 Badeck
et al 2005 Pausch and Kuzyakov 2017) Slightly more positive δ13CFoliar values
(about 15permil) were measured in comparison with their δ13CSoil values This may be
reflecting the C transference from plants to the soil but also a soil-atmosphere
interchange The preferential assimilation of the light isotopes (12C) during soil
respiration carried by the roots and the microbial biomass that is associated with the
decomposition of litter roots and soil organic matter explain this differential
(Berhardt et al 2006 Blagodatskaya et al 2010 Midwood and Millard 2011)
97
In general the δ13CSoil values from the Lago Vichuqueacuten watershed oscillated
around -290permil and did not vary with our plant classification types Here we use
these values as terrestrial-end members to track changes in source OM (Fig 6)
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from the terrestrial watershed By the other hand more positive δ13C
values most likely reflect an increased aquatic OM component as indicated by POM
isotope fluctuations (Fig 9)
53 Recently land use and cover change and its influences on N inputs to the lake
Tree plantations in the Lago Vichuqueacuten watershed have increased greatly in
the last few decades (from 1 in 1975 to 66 in 2016 Fig 7) replacing previous
native forests (from 26 to 7) meadows (17 to 5) and shrublands (53 to
17) In 1975 tree plantations were confined to the lake perimeter with discrete
patches distributed throughout the watershed The ldquoForest Lawrdquo (DFL 701- passed
in 1974) allocated state funding to afforestation efforts and management of tree
plantations which greatly favored the replacement native forests by introduced trees
This increase is marked by a sharp and steady decrease in lake sediment δ15N and
δ13C values because tree plantations function as a nutrient sink whereas other land
uses such as meadows favor nutrient loss from the watershed (Fig8) Bruland and
Mackenzie (2014) noted a decrease in wetland δ15N values when watershed
forested cover increased and concluded that N inputs to the wetlands are lower from
the forested areas as they generally do not export as much N as agricultural lands
A positive correlation between native vegetation and δ15Ncore values can be
explained by the relatively scarce arboreal cover in the watershed in 1975 when
native forest occupied just 26 of the watershed surface whereas shrublands and
98
meadows occupied more than the 70 of the surface of the watershed with the
concomitant elevated loss of N (Fig 7 and Fig 8)
54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years
Sedimentological compositional and geochemical indicators all show changes in
the N cycle associated to LUCC lake dynamics and climate changes (Fig 10) From
the pre-Columbian indigenous settlement including the Spanish colonial period up
to the start of the Republic (1300 - 1800 CE) the introduction of crops such as
quinoa and wheat but also the clearing of land for extensive agriculture would have
favored the entry of N into the lake Conversely major changes observed during the
last century were characterized by a sharp decrease of N input that were coeval
with the increase of tree plantations
From 1365 until 1550 CE (unit 2c 2b and half of 2a Fig 3 and Fig 10) pre-
Columbian agricultural societies planted mostly crops of corn and quinoa (Ramirez
and Vidal 1985) This period is characterized by large peaks seen in the δsup1⁵N record
(eg 1403 1452 1476 and 1505) with overall high δsup1⁵N values (up to 5permil) indicating
that N input from watershed was elevated and oscillating to the beat of the NT These
positive δsup1⁵N peaks could be due to several causes including a) the clearing of land
for farming b) N loss via denitrification which would be generally augmented in
anoxic environments (Diebel et al 2009) such as Lago Vichuqueacuten (low MnFe
values about 0001 Fig 4) or c) climate especially cold-wet winters or hot-dry
summers can also exert control on the δsup1⁵N record Indeed tree-ring records and
summer temperature reconstructions show overall wetcold conditions during this
period (Christie et al 2009 Von Gunten et al 2009 Fig 10) Increased
precipitation would bring more sediment (and nutrients) from the watershed into the
99
lake and increase lake productivity which is also detected by the geochemical
proxies for productivity (peaks of BrTi SrCa and TOC Fig 4 and Fig 10) (see also
Frugone-Alvarez et al 2017)
Figure 10 Changes in the N availability during the last six centuries in Lago
Vichuqueacuten a) lake sediment δsup1⁵N values displayed more positive values during the
prehistoric period Spanish Colony and the starting 19th century which is associated
with enhanced N input from the watershed by extensive clearing and crop
plantations The inset shows this relationship between sediment δsup1⁵N and
100
percentage of meadow cover over the last 30 years b) Summer temperature
reconstruction from central Chile (von Gunten et al 2009) showing a
correspondence with cold phases and high δsup1⁵N values at Vichuqueacuten except for the
last 100 years c) Palmer Drought Severity Index (PDSI) is an interannual moisture
variability reconstruction for late springndashearly summer during the last six centuries
(Christie et al 2009) Grey shadow indicating higher precipitation periods
From 1550 and 1810 CE (unit 2a) and during the Colonial period several peaks
of δ15N could be further related not only to high precipitation (Fig 10 grey shadow)
but also pulses of enhanced N input from the watershed linked to human land use
In 1550 CE Juan Cuevas was granted lands and indigenous workers under the
encomienda system for agricultural and mining development of the Vichuqueacuten
village (Ramiacuterez and Vidal 1985) Historical documents show that around 1579 CE
the Vichuqueacuten watershed was occupied by indigenous communities dedicated to
wheat plantations and vineyards wood extraction and gold mining (Odone 1998)
The introduction of the Spanish agricultural system implied not just a change in the
types of crops used (from quinoa to vineyards and wheat) but also a clearing of
native species for the continuous increase of agricultural surface and wood
extraction During the Colonial period wheat from Vichuqueacuten was exported to Peru
(Ramiacuterez and Vidal 1985) Armesto et al (2009) indicate that during the XVIII and
XIX centuries the extraction of wood for mining operations was important enough to
cause extensive loss of native forests The independence and instauration of the
Chilean Republic did not change this prevailing system Increases in the
contributions of N to the lake during the second half of the XIX century (peaks in
δsup1⁵N ca 1870 CE) could be due to expanding farm land dedicated for wheat
101
production and increased commercial trade with California and Canada (Ramiacuterez
and Vidal 1985)
In contrast LUCC in the last century are clearly related to the development of
large-scale tree plantations (Fig 7) The δsup1⁵N record reaches the lowest values of
the entire sequence in the last few decades (Fig 10) A marked increase in lake
productivity NT concentration and decreasing sediment input is synchronous (unit
1 Fig 4) with trees replacing meadows shrublands and areas with native forests
(Fig 7 and 8) The implementation of the lsquoForest Lawrsquo in 1974 had a stronger impact
on the landscape and lake ecosystem dynamics than the impacts of ongoing climate
change in the region which is much more recent (Garreaud et al 2018) although
the prevalence of hot dry summers seen over the last decade would also be
associated with low δsup1⁵N values and increase of NT (Fig 10) Heavy metals ratios
(FeTi CoTi and PbTi) show notable increases in lake sediments from 1992 to 2011
CE (Fig 4) Although this could be related to mining in the El Maule region the
closest mines are 60 Km away (Pencahue and Romeral) so local factors related to
shoreline urbanization for the summer homes and an increase in tourist activity
could also be a major factor
6 Conclusions
The N isotope signal in the watershed depends on the rates of exchange
between vegetation and soil cover (Fig 11) Plants preferentially uptake 14N and the
underlying soils become enriched in 15N especially when the terrestrial ecosystem
is N-limited andor significant N loss occurs (ie denitrification andor ammonia
volatilization) LUCC in the Lago Vichuqueacuten watershed have heavily modified the
links between terrestrial and aquatic ecosystems with agriculture practices
102
contributing more N to the lake than tree plantations or native forests In situ lake
processes can also fractionate N isotopes An increase of N-fixing species results
in OM depleted in 15N which results in POM with lower δsup1⁵N values during these
periods During winter phytoplankton is typically enriched in 15N due to the
decreased abundance of N-fixing species and increased N input from the
watershed This in turn generates the highest δsup1⁵NPOM values in Lago Vichuqueacuten
Periods of elevated productivity tend to deplete the dissolved inorganic N of 14N
resulting in even higher δ15N values
Our study illustrates the usefulness of δsup1⁵N as a tool for reconstructing the past
influence of LUCC on N availability in lake ecosystems To constrain the relative
roles of the diverse forcing mechanisms that can alter N cycling in mediterranean
ecosystems all main components of the N cycle should be monitored seasonally
(or monthly) including the measurements of δ15N values in land samples
(vegetation-soil) as well as POM
103
Figure 11 Summary of human and environmental factors controlling the δ15N
values of lake sediments Particulate organic matter(POM) δ15N values in
mediterranean lakes are driven by N input from the watershed that in turn depend
on land use and cover changes (ie forest plantation agriculture) andor seasonal
changes in N sources andor lake ecosystem processes (ie bioproductivity redox
condition denitrification) Arrows indicate the direction of N fluxes (inputoutput from
the N cycle) N cycle processes that deplete lake sediments of 15N are shown in
blue whereas those that enrich sediments in 15N are shown in red
104
Supplementary material
Figure S1 Pearson correlate coefficient between geochemical variables in core
VIC13-2B Positive and large correlations are in blue whereas negative and small
correlations are in red (p valuelt0001)
Figure S2 Principal Component Analysis of geochemical elements from core
VIC13-2B
105
Table S1 Lago Vichuqueacuten radiocarbon samples
RADIOCARBON
LAB CODE
SAMPLE
CODE
DEPTH
(m)
MATERIAL
DATED
14C AGE ERROR
D-AMS 029287
VIC13-2B-
1 043 Bulk 1520 24
D-AMS 029285
VIC13-2B-
2 085 Bulk 1700 22
D-AMS 029286
VIC13-2B-
2 124 Bulk 1100 29
Poz-63883 Chill-2D-1 191 Bulk 945 30
D-AMS 001133
VIC11-2A-
2 201 Bulk 1150 44
Poz-63884
Chill-2D-
1U 299 Bulk 1935 30
Poz-64089
VIC13-2D-
2U 463 Bulk 1845 30
Poz-64090
VIC13-2A-
3U 469 Bulk 1830 35
D-AMS 010068
VIC13-2D-
4U 667 Bulk 2831 25
Poz-63886
VIC13-2D-
4U 719 Bulk 3375 35
106
D-AMS 010069
VIC13-2D-
5U 775 Bulk 3143 27
Poz-64088
VIC13-2D-
5U 807 Bulk 3835 35
D-AMS-010066
VIC13-2D-
7U 1075 Bulk 6174 31
Poz-63885
VIC13-2D-
7U 1197 Bulk 6440 40
Poz-5782 VIC13-15 DIC 180 25
Table S2 Loadings of the trace chemical elements used in the PCA
Elementos PC1 PC2 PC3 PC4
Zr 0922 0025 -0108 -0007
Zn 0913 -0124 -0212 0001
Rb 0898 -0057 -0228 0016
K 0843 0459 0108 0113
Ti 0827 0497 0060 -0029
Al 0806 0467 0080 0107
Si 0803 0474 0133 0136
Y 0784 -0293 -0174 0262
V 0766 0455 0090 -0057
Br 0422 -0716 -0045 0226
Ca 0316 -0429 0577 0489
Sr 0164 -0420 0342 -0182
Cl 0151 -0781 -0397 0162
107
Mn -0121 -0091 0859 0095
S -0174 -0179 -0051 0714
Pb -0349 0414 -0282 0500
Fe -0700 0584 -0023 0280
Co -0704 0564 -0107 0250
Table S3 Stable Isotope values from vegetation of Lago Vichuqueacutenacutes watershed
Taxa Classification δsup1⁵N δsup1sup3C CN
molar
Poaceae Meadow 1216 -2589 3602
Juncacea Meadow 1404 -2450 3855
Cyperaceae Meadow 1031 -2596 1711
Taraxacum
officinale Meadow 836 -2400 2035
Poaceae Meadow 660 -2779 1583
Poaceae Meadow 453 -2813 1401
Poaceae Meadow 966 -2908 4010
Juncus Meadow 1247 -2418 3892
Poaceae Meadow 747 -3177 6992
Poaceae Meadow 942 -2764 3147
Poaceae Meadow 1479 -2634 2895
Poaceae Meadow 1113 -2776 1795
Poaceae Meadow 2215 -2737 7971
Poaceae Meadow 1121 -2944 2934
Poaceae Meadow 638 -3206 1529
108
Macrophytes Macrophytes 886 -3044 2286
Macrophytes Macrophytes 1056 -2720 2673
Macrophytes Macrophytes 769 -3297 1249
Macrophytes Macrophytes 967 -2763 1442
Macrophytes Macrophytes 959 -2670 2105
Macrophytes Macrophytes 334 -2728 1038
Acacia dealbata
Introduced
species 656 -2696 1296
Acacia dealbata
Introduced
species 487 -2941 1782
Acacia dealbata
Introduced
species 220 -2611 3888
Luma apiculata Native species 433 -2542 4135
Luma apiculata Native species 171 -2664 7634
Luma apiculata Native species -001 -2736 6283
Luma apiculata Native species 029 -2764 6425
Azara sp Native species 159 -2868 8408
Azara sp Native species 101 -2606 2885
Baccharis concava Native species 104 -2699 5779
Baccharis concava Native species 265 -2488 4325
Baccharis concava Native species 287 -2562 7802
Baccharis concava Native species 427 -2781 5204
Baccharis linearis Native species 190 -2610 4414
Baccharis linearis Native species 023 -2825 5647
109
Peumus boldus Native species 042 -2969 6327
Peumus boldus Native species 205 -2746 4110
Peumus boldus Native species 183 -2743 6293
Chusquea quila Native species 482 -2801 4275
Poaceae meadow 217 -2629 7214
Lobelia sp Native species 224 -2645 3963
Lobelia sp Native species -091 -2565 4538
Aristotelia chilensis Native species -035 -2785 5247
Aristotelia chilensis Native species -305 -2889 2305
Aristotelia chilensis Native species 093 -2836 5457
Chusquea quila Native species 173 -2754 3534
Chusquea quila Native species 045 -2950 6739
Quillaja saponaria Native species 223 -2838 9385
Scirpus meadow 018 -2820 7115
Sophora sp Native species -184 -2481 2094
Sophora sp Native species -181 -2717 1721
Pinus radiata
Introduced
trees 1581 -2602 3679
Pinus radiata
Introduced
trees 1431 -2784 4852
Pinus radiata
Introduced
trees -091 -2708 9760
Pinus radiata
Introduced
trees 153 -2568 3470
110
Salix sp
Introduced
trees 632 -2878 1921
LITERATURE CITED
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A historical
framework for land cover change in southwestern South America in the past 15000
years Land use policy 27 148ndash160
httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the next
carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014
httpsdoiorg101002eft2235
Blaauw M Christeny JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474
httpsdoiorg10121411-BA618
Bowman DMJS Cook GD 2002 Can stable carbon isotopes (δ13C) in soil
carbon be used to describe the dynamics of Eucalyptus savanna-rainforest
boundaries in the Australian monsoon tropics Austral Ecol
httpsdoiorg101046j1442-9993200201158x
Brahney J Ballantyne AP Turner BL Spaulding SA Otu M Neff JC 2014
Separating the influences of diagenesis productivity and anthropogenic nitrogen
deposition on sedimentary δ15N variations Org Geochem 75 140ndash150
httpsdoiorg101016jorggeochem201407003
111
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409
httpsdoiorg102134jeq20090005
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego R
Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and
environmental change from a high Andean lake Laguna del Maule central Chile
(36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the
Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from
tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Craine JM Elmore AJ Aidar MPM Dawson TE Hobbie EA Kahmen A
Mack MC Kendra K Michelsen A Nardoto GB Pardo LH Pentildeuelas J
Reich B Schuur EAG Stock WD Templer PH Virginia RA Welker JM
Wright IJ 2009 Global patterns of foliar nitrogen isotopes and their relationships
with cliamte mycorrhizal fungi foliar nutrient concentrations and nitrogen
availability New Phytol 183 980ndash992 httpsdoiorg101111j1469-
8137200902917x
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty
Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-
010-9453-1
Diebel M Vander Zanden MJ 2012 Nitrogen stable isotopes in streams stable
isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19
1127ndash1134 httpsdoiorg10189008-03271
112
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in freshwater
wetlands record long-term changes in watershed nitrogen source and land use SO
- Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash
2916
Fariacuteas L Castro-Gonzales M Cornejo M Charpentier J Faundez J
Boontanon N Yoshida N 2009 Denitrification and nitrous oxide cycling within the
upper oxycline of the oxygen minimum zone off the eastern tropical South Pacific
Limnol Oceanogr 54 132ndash144
Farquhar GD Ehleringer JR Hubick KT 1989 Carbon Isotope Discrimination
and Photosynthesis Annu Rev Plant Physiol Plant Mol Biol
httpsdoiorg101146annurevpp40060189002443
Farquhar GD OrsquoLeary MH Berry JA 1982 On the relationship between
carbon isotope discrimination and the intercellular carbon dioxide concentration in
leaves Aust J Plant Physiol httpsdoiorg101071PP9820121
Fogel ML Cifuentes LA 1993 Isotope fractionation during primary production
Org Geochem httpsdoiorg101007978-1-4615-2890-6_3
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A
Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-
resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS)
implications for past sea level and environmental variability J Quat Sci 32 830ndash
844 httpsdoiorg101002jqs2936
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924
httpsdoiorg104319lo20095430917
113
Gerhart LM McLauchlan KK 2014 Reconstructing terrestrial nutrient cycling
using stable nitrogen isotopes in wood Biogeochemistry 120 1ndash21
httpsdoiorg101007s10533-014-9988-8
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and oxygen
isotope fractionation during dissimilatory nitrate reduction by denitrifying bacteria 53
2533ndash2545 httpsdoiorg10230740058342
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater
eutrophic lake Limnol Oceanogr 51 2837ndash2848
httpsdoiorg104319lo20065162837
Harris D Horwaacuteth WR van Kessel C 2001 Acid fumigation of soils to remove
carbonates prior to total organic carbon or CARBON-13 isotopic analysis Soil Sci
Soc Am J 65 1853 httpsdoiorg102136sssaj20011853
Houmlgberg P 1997 Tansley review no 95 natural abundance in soil-plant systems
New Phytol httpsdoiorg101046j1469-8137199700808x
Kylander ME Ampel L Wohlfarth B Veres D 2011 High-resolution X-ray
fluorescence core scanning analysis of Les Echets (France) sedimentary sequence
New insights from chemical proxies J Quat Sci 26 109ndash117
httpsdoiorg101002jqs1438
Lara A Solari ME Prieto MDR Pentildea MP 2012 Reconstruccioacuten de la
cobertura de la vegetacioacuten y uso del suelo hacia 1550 y sus cambios a 2007 en la
ecorregioacuten de los bosques valdivianos lluviosos de Chile (35o - 43o 30acute S) Bosque
(Valdivia) 33 03ndash04 httpsdoiorg104067s0717-92002012000100002
Lehmann M Bernasconi S Barbieri A McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during
114
simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66
3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007
Liu G Li M An L 2007 The Environmental Significance Of Stable Carbon
Isotope Composition Of Modern Plant Leaves In The Northern Tibetan Plateau
China Arctic Antarct Alp Res httpsdoiorg1016571523-0430(07-505)[liu-
g]20co2
Marzecovaacute A Mikomaumlgi A Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54
httpsdoiorg103176eco2011105
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash
1643 httpsdoiorg1011770959683613496289
Meyers PA 2003 Application of organic geochemistry to paleolimnological
reconstruction a summary of examples from the Laurention Great Lakes Org
Geochem 34 261ndash289 httpsdoiorg101016S0146-6380(02)00168-7
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland
Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Odone C 1998 El Pueblo de Indios de Vichuqueacuten siglos XVI y XVII Rev Hist
Indiacutegena 3 19ndash67
Pausch J Kuzyakov Y 2018 Carbon input by roots into the soil Quantification of
rhizodeposition from root to ecosystem scale Glob Chang Biol
httpsdoiorg101111gcb13850
115
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98
httpsdoiorg1011772053019614564785
Sun W Zhang E Jones RT Liu E Shen J 2016 Biogeochemical processes
and response to climate change recorded in the isotopes of lacustrine organic
matter southeastern Qinghai-Tibetan Plateau China Palaeogeogr Palaeoclimatol
Palaeoecol httpsdoiorg101016jpalaeo201604013
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of
different trophic status J Paleolimnol 47 693ndash706
httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl
httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M 2009 High-resolution quantitative climate
reconstruction over the past 1000 years and pollution history derived from lake
sediments in Central Chile Philos Fak PhD 246
Woodward C Chang J Zawadzki A Shulmeister J Haworth R Collecutt S
Jacobsen G 2011 Evidence against early nineteenth century major European
induced environmental impacts by illegal settlers in the New England Tablelands
south eastern Australia Quat Sci Rev 30 3743ndash3747
httpsdoiorg101016jquascirev201110014
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K Yeager
KM 2016 Different responses of sedimentary δ15N to climatic changes and
116
anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau
J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
117
DISCUSION GENERAL
El Nitroacutegeno (N) es un elemento clave que controla la dinaacutemica y
funcionamiento de ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al
1997) Pese a que el N (N2) es muy abundante en la atmoacutesfera (78) es escaso
en los ecosistemas pues la mayoriacutea de la biota no puede acceder a su forma
molecular (Galloway et al 1995 Zaehle et al 2013) En este sentido la entrada
natural de N en los ecosistemas es viacutea fijacioacuten bioloacutegica del N2 por bacterias que lo
convierten en compuestos bioloacutegicamente disponibles (Battye et al 2017) Debido
a que la fijacioacuten es un proceso energeacuteticamente costoso los ecosistemas
comuacutenmente estaacuten limitados por N (Vitousek and Howarth 1991) Los LUCC
contribuyen al incremento del N disponible y son una de las principales causas de
eutroficacioacuten y contaminacioacuten de los cuerpos de agua (Woodward et al 2012)
En Chile central los LUCC principalmente relacionados con las actividades
agriacutecolas y forestales han causado grandes cambios en el ciclo del N debido al
118
reemplazo de la cobertura natural por un monocultivo y al uso de fertilizantes que
modifican los aportes de MO y N a los cuerpos de agua El programa de
estabilizacioacuten de laderas impulsada por el gobierno (Albert 1900) y la ley forestal
de 1974 han fomentado la forestacioacuten con plantaciones de Pinus radiata y
Eucalyptus globulus y en el caso de Lago Vichuqueacuten estas actualmente cubren maacutes
del 60 de su cuenca (Cap2 Fig 5) Ademaacutes en Laguna Matanzas la
sobreexplotacioacuten del agua en conjunto con el cambio climaacutetico actual el que ha
conducido a una de las megasequiacuteas maacutes severas en las uacuteltimas deacutecadas
(Garreaud et al 2017) generoacute una combinacioacuten letal que secoacute el lago en tan solo
10 antildeos (Cap1 Fig1) Las evidencias obtenidas a partir de estos dos lagos
permiten identificar las huellas del Antropoceno en Chile central basadas en el
registro sedimentario lacustre
La transferencia de N desde los ecosistemas terrestres a acuaacuteticos es un
proceso natural con controles bioacuteticos climaacuteticos y geoloacutegicos El enlace
hidroloacutegico entre los ecosistemas terrestres y acuaacuteticos hace que el estado troacutefico
de los lagos esteacute vinculado al de su cuenca (McLauchlan et al 2013) En Chile
central se sabe poco del impacto de los LUCC sobre el ciclo de nutrientes en los
ecosistemas lacustres aunque al menos desde la colonizacioacuten espantildeola se tienen
registros de influencia humana en las cuencas Durante la colonia espantildeola
Laguna Matanzas fue una hacienda ganadera que exportaba productos caacuternicos al
Peruacute (Contreras-Lopez et al 2014) A su vez en el Lago Vichuqueacuten se realizaban
extensas actividades agriacutecolas hasta comienzos del s XX como por ejemplo
cultivos de vid y trigo ademaacutes de extraccioacuten de madera y mineriacutea de oro (Odone
1997) En las uacuteltimas deacutecadas los LUCC han estado vinculados principalmente con
el desarrollo forestal al compaacutes de una estrategia gubernamental de fomentar con
119
incentivos financieros (Ley de Bosques DFL nordm265 de 1931 y DFL 701 de 1974)
esta actividad El incremento de la superficie forestal es especialmente fuerte en
ambos lagos a partir de la deacutecada de 1980 y hasta 2016 (Laguna de Matanzas 0-
17 Lago Vichuqueacuten 1-66) Este aumento habriacutea sido en detrimento del bosque
nativo y los pastizales (Cap 1 Fig 5 y Cap 2 Fig 8) El incremento de la superficie
forestal generariacutea una disminucioacuten de los aportes de sedimentos y nutrientes al lago
y en este sentido un cambio de estado en los flujos de N (e g tipping points) que
a su vez ha quedado registrado en la sentildeal isotoacutepica (δ15N) y la acumulacioacuten de
MO en los sedimentos lacustres
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central
Los sedimentos lacustres son verdaderos ldquosumiderosrdquo que pueden llegar a
registrar lo que ocurre en cuanto a la historia ambiental de su cuenca En esta tesis
se utilizan los isoacutetopos estables de N (15N y 14N) en sedimentos lacustres para
reconstruir los cambios en la disponibilidad de N en el tiempo y establecer la
magnitud de impacto generado por actividades humanas El fraccionamiento
cineacutetico en los lagos es mediado por procesos bioloacutegicos que mediante la
asimilacioacuten de N genera un producto (eg fitoplancton) maacutes ldquoligerordquo (valores maacutes
bajos de 15N) que su remanente (eg DIN) (Fogel y Cifuentes 1993) Sin embargo
en periacuteodos en que los lagos estaacuten limitados por N el fitoplancton discrimina menos
y la MO resultante queda enriquecida en 15N (Fig 1 a y b) Ademaacutes la
desnitrificacioacuten acentuada en lagos de fondo anoacutexicos (eg Laguna Matanzas
entre 1800 y 1940 Cap1 Fig 6) conduce a un enriquecimiento en 15N de los
sedimentos y de la columna de agua Esta informacioacuten queda reflejada en la MO
120
de los sedimentos lacustres lo que permite conocer la disponibilidad del N en el
tiempo a partir de las variaciones de 15N
En esta tesis analizamos la MO de los sedimentos lacustres para reconstruir
la influencia de los LUCC en los aportes de sedimentos y N al lago En teacuterminos de
asimilacioacuten de N se puede distinguir entre dos grupos principales de productores
primarios que componen el POM (Fig1)
1 Los que asimilan el DIN compuesto por amonio nitratos y nitritos Aquiacute el
δ15N resultante fluctuaraacute en torno a valores maacutes positivos en la medida que
la productividad se incrementa o no hay reposicioacuten del DIN (Fig 1)
2 Los fijadores de N Dado que es un proceso energeacuteticamente costoso en
ambientes que no estaacuten limitados por N muchas veces son excluiacutedas
competitivamente por el resto del fitoplancton Si el DIN queda agotado por
el incremento de la productividad (o bien si el ambiente estaacute limitado ya sea
por N o por otros nutrientes) los fijadores de N pueden florecer y la MO que
se genera a partir de ello tendraacute un δ15N con valores en torno a 0permil
De este modo la MO en los sedimentos lacustres dependeraacute de la
composicioacuten de productores primarios (ie fijadores vs asimiladores del DIN)
ademaacutes de los aportes desde la cuenca tanto de MO como del N de la cuenca que
pasa a formar parte del DIN (eg fertilizantes estieacutercol) (Fig 1)
La MO de los lagos estudiados en esta tesis ha sido analizada a partir de
variables geoquiacutemicas isotoacutepicas frotis y estaacute compuesta principalmente por
diatomeas y MO amorfa A partir de los datos obtenidos en esta tesis los valores
de δ15N aumentan cuando hay una mayor cobertura agriacutecola o pastizales A su vez
tiende a disminuir cuando aumenta la cobertura arboacuterea independiente si esta es
por plantaciones forestales o por bosque nativo
121
Por otra parte la dinaacutemica del lago tambieacuten imprime caracteriacutesticas
especiales en el POM observaacutendose variaciones estacionales en los valores
δ15NPOM Durante la estacioacuten caacutelida y seca se observan valores maacutes bajos que
durante la estacioacuten friacutea y huacutemeda posiblemente relacionado con el incremento de
la contribucioacuten de especies fijadoras de N (Lago Vichuqueacuten Fig 9 Cap 2) durante
el verano En la estacioacuten friacutea y huacutemeda el DIN seriacutea la principal fuente de N Las
mayores entradas de MO y N terrestre debidos a un incremento del lavado de la
cuenca como consecuencia de las precipitaciones y la mineralizacioacuten de la MO
podriacutean estimular la produccioacuten de MO lacustre por crecimiento del fitoplancton
Como consecuencia se observan tendencias decrecientes de los valores de
δ15N en la MO sedimentaria cuando la productividad en el lago estaacute relacionada
con especies fijadoras de N mientras que el δ15N muestra aumentos cuando la
productividad del lago estaacute asociada principalmente al consumo del DIN pero
tambieacuten cuando predominan procesos de perdida de N (eg desnitrificacioacuten Fig
1)
Hemos identificado tres fases en la dinaacutemica del δ15N para ambos lagos
Entre 1800 y 1940 CE la cuenca de laguna de Matanzas estaba dominada por
actividades agriacutecolas y ganaderas (δ15Nsuelo gt 4permil Cap 1 Fig 4 y 6) Las entradas
de sedimentos y MO desde la cuenca son altas (δ13C lt -209permil ZrTi ~029 AlTi
~013) y coincide con un clima maacutes huacutemedo y friacuteo (Von Gunten et al 2009
Christiansen et al 2011) El lago teniacutea un nivel de agua maacutes profundo dominado
por condiciones anoacutexicas (MnFe ~0013) que contribuiriacutean a mayores valores de
δ15N en la columna de agua y por ende de los sedimentos acumulados en el fondo
debido a la peacuterdida diferencial de 14N viacutea desnitrificacioacuten (Fig 7 Cap1) La baja
122
produccioacuten de MO lacustre (BrTi ~002 TOC~17) coincide con un bajo contenido
de NT (~478 μg) y valores maacutes positivos de δ15N (gt 4permil)
La actividad agriacutecola tambieacuten dominaba en la cuenca del Lago Vichuqueacuten
durante este periacuteodo (δ15Nsuelo gt4permil Cap2 Fig 6) El aporte sedimentario desde la
cuenca es alto (AlTi ~029 ZrTi ~032) y fue favorecido por mayores
precipitaciones a las actuales (Christiansen et al 2011) El Lago Vichuqueacuten es un
lago estratificado anoacutexico (MnFe ~002) y profundo (~30 m) por lo que la
desnitrificacioacuten juega un rol importante en el enriquecimiento de la MO
sedimentaria La baja produccioacuten de MO (BrTi ~007 TOC ~12) de origen
lacustre (δ13C ~ -268permil) coincide con un bajo contenido de NT (~309μg +6) y
valores maacutes positivos de δ15N (56permil +03)
Durante esta fase en ambos lagos los aportes de N de la cuenca parecen
ser los principales responsables en las fluctuaciones del δ15N Esta sentildeal podriacutea
estar amplificada por bajas temperaturas (que afectan la productividad lacustre) y
altas precipitaciones que favorecen el lavado de la cuenca y aumentan el aporte de
sedimentos y MO desde la cuenca predominantemente agriacutecola
Entre 1940 y 1980 CE en Laguna Matanzas se observa un incremento en
la acumulacioacuten de MO algal (TOC ~2 +1 δ13C ~-230+09permil) El ambiente
deposicional es ligeramente menos turbulento que el anterior (AlTi ~011+001
ZrTi ~03+001) y el lago estaacute en una fase de transicioacuten hacia condiciones maacutes
oacutexicas (MnFe ~002+0004) y maacutes productivas (BrTi ~004+0007) A su vez δ15N
tiende hacia valores maacutes negativos (δ15N ~30permil+03) mientras que el NT aumenta
oscilando en antifase con el δ15N
En Lago Vichuqueacuten en cambio se observa un ligero incremento en la
acumulacioacuten de MO (TOC ~13 +02) de origen terrestre (δ13C ~-27permil +04) La
123
productividad es similar al periacuteodo anterior (BrTi ~007+003) y el ambiente
deposicional es menos turbulento (AlTi ~02 +009 Zrti ~033 +007) El δ15N y el
NT oscilan en torno a valores ligeramente maacutes bajos (δ15N ~51permil +04 NT ~292μg
+6) Este periacuteodo se caracteriza por ser maacutes caacutelido que el anterior lo que
posibilitariacutea un incremento en la productividad lacustre en Laguna Matanzas pero
que no es observada en el Lago Vichuqueacuten
Entre 1980 y la actualidad en Laguna Matanzas habriacutea un incremento de la
acumulacioacuten de MO (TOC ~59 + 3) asociada a un incremento en la productividad
del lago (BrTi ~009+005) y a los aportes de MO terrestres (δ13C ~-24 + 2permil) El
lago se vuelve maacutes oacutexico (Mn ~0003 +0006) y las entradas de sedimento
disminuyen (AlTi~ 009 +002) El δ15N tiende a valores maacutes negativos (δ15N ~13permil
+1) y el NT aumenta de 20 a 55 μg (NT ~346 + 9 μg) tomando valores sin
precedentes en todo el registro En los sedimentos lacustres del Lago Vichuqueacuten
tambieacuten se observa un incremento de la acumulacioacuten de MO (TOC ~17 + 05)
asociada a un incremento en la productividad del lago (BrTi ~01 + 003) y las
entradas de sedimento desde la cuenca disminuyen (AlTi ~005 + 005) El δ15N
(~43 + 05permil) tiende a valores maacutes negativos y el NT incrementa de 20 a 55 μg (NT
~346 + 9 μg) oscilando en antifase
Durante esta fase en ambos lagos se observa un aumento en la
acumulacioacuten de MO sedimentaria un incremento en el NT y valores maacutes negativos
de δ15N que coincide con el incremento de la superficie forestal de las cuencas
(Laguna Matanzas gt17 Lago Vichuqueacuten gt60)
124
Figura 2 Comparacioacuten de los cambios del ciclo del N entre Laguna Matanzas y
Lago Vichuqueacuten con los procesos biogeoquiacutemicos contrastantes en rojo En L
Matanzas las condiciones oacutexicas podriacutean estar favoreciendo la nitrificacioacuten del
amonio En Vichuqueacuten las condiciones anoacutexicas favorecen un aumento en δ15N de
la MO en los sedimentos por desnitrificacioacuten y mineralizacioacuten
Los ambientes mediterraacuteneos en el que los lagos del presente estudio se
encuentran insertos estaacuten caracterizados por una estacioacuten seca prolongada y las
precipitaciones ocurren en eventos puntuales alcanzando altos montos
pluviomeacutetricos en poco tiempo Las lluvias favorecen un lavado de la cuenca la
perdida de N y otros nutrientes Por lo que es esperable que los sedimentos del
lago reflejen mejor los valores isotoacutepicos de los suelos de la cuenca durante los
periacuteodos con altos montos pluviomeacutetricos En el Lago Vichuqueacuten por ejemplo el
125
POM superficial (2 y 5 m de profundidad) despliega valores isotoacutepicos maacutes
positivos en invierno presumiblemente como resultado de mayores aportes de MO
y N de su cuenca (Fig 1b) En ambos lagos los valores maacutes positivos en los
sedimentos lacustres ocurren durante periacuteodos con mayores montos pluviomeacutetricos
(Cap1 Fig 6 y Cap 2 Fig 12)
Ademaacutes del efecto de la precipitacioacuten en el aporte de nutrientes al lago en
esta tesis hemos encontrado que los mayores aportes de N desde la cuenca se dan
cuando la cobertura vegetal del suelo es menor En Laguna Matanzas el desarrollo
de la actividad ganadera desde la llegada de los espantildeoles a la cuenca habriacutea
incrementado los aportes de N al lago Los valores de δ15N en los sedimentos
lacustres depositados durante este periacuteodo son los maacutes positivos de todo el registro
(~4permil) A su vez en Lago Vichuqueacuten los valores isotoacutepicos maacutes positivos se
registran entre 1300 y 1900 CE que coinciden con el mayor desarrollo de
actividades agriacutecola y de extraccioacuten de maderas de la cuenca (Fig 10 Cap 2)
Los recientes LUCC (a partir de 1975) en las cuencas de Laguna Matanzas
y Lago Vichuqueacuten estaacuten caracterizados por un incremento de la superficie forestal
y agriacutecola en reemplazo de la cobertura de pastizal (Laguna Matanzas) y bosque
nativo (Lago Vichuqueacuten) Los valores de δ15N en los testigos sedimentarios fueron
maacutes positivos cuanto mayor la superficie agriacutecola de la cuenca y maacutes negativos
cuanto mayor la superficie forestal Aunque por los alcances de esta tesis no
podemos ser concluyentes respecto del origen del NT (aloacutectonoautoacutectono) parece
ser que esta maacutes ligado a los aportes de la cuenca pese a la disminucioacuten del aporte
sedimentario observado en ambos lagos Las plantaciones forestales a diferencia
del bosque nativo son fertilizadas con una mezcla de N P y K (Toral et al 2005)
Por lo que pese a la disminucioacuten del aporte sedimentario la concentracioacuten de
126
nutrientes en el aporte al lago puede verse incrementada por el tipo de manejo
forestal con respecto al bosque nativo
Los resultados del primer capiacutetulo demuestran que 1) las plantaciones
forestales yo bosque nativo retiene maacutes sedimentos y MO mientras que el uso de
suelo agriacutecola y los pastizales destinados al desarrollo de la actividad ganadera lo
libera 2) Al parecer la desnitrificacioacuten es el proceso natural maacutes importante de
perdida de N en lagos costeros y favorece el enriquecimiento de 15N tanto en la
columna de agua (ie 15N-DIN) como en los sedimentos una vez enterrados y 3) la
desnitrificacioacuten depende de los cambios en las condiciones REDOX en el lago La
oxigenacioacuten en lagos poco profundos -como Laguna Matanzas- es altamente
fluctuante a escalas decadal-centenal depende de la profundidad de la laacutemina de
agua y de la variabilidad de la precipitacioacuten En este sentido en Laguna Matanzas
habriacutea condiciones anoacutexicas en ambientes cuando el lago alcanza niveles maacutes
altos Estas condiciones se observan entre 1820 y 1940 CE y son coetaacuteneas con
episodios maacutes huacutemedos y friacuteos (von Gunten et al 2009 Christie et al 2009) pero
tambieacuten con una fuerte actividad ganadera en la cuenca
Las fuentes variables de N y su fluctuacioacuten en el tiempo hacen necesario
contar con un anaacutelogo moderno que permite usar al δ15N de los sedimentos
lacustres como un indicador indirecto de los cambios en la disponibilidad de N en
el tiempo Por ello en el segundo capiacutetulo integramos muestras de suelo-
vegetacioacuten y un monitoreo de POM de la columna de agua en verano e invierno La
composicioacuten isotoacutepica de POM estariacutea reflejando cambios de la fuente de N (fijacioacuten
vs consumo del DIN) que variacutea durante el antildeo hidroloacutegico Durante el verano la
mayor temperatura y horas-luz favoreceriacutean las entradas de N al lago viacutea fijacioacuten
bioloacutegica de N2 atmosfeacuterico resultando en un incremento de la MO con un valor
127
isotoacutepico bajo (POM verano~5permil Fig 1b) El N2 (δ15N = 0permil) es fijado praacutecticamente
sin fraccionamiento isotoacutepico generando un δ15N de la OM cercano a 0permil (Leng et
al 2006) En invierno las precipitaciones pueden estar favoreciendo un incremento
en la entrada de nutrientes al lago El DIN de la columna de agua llega a contener
valores de δ15N maacutes altos reflejando estos mayores aportes de la cuenca y el POM
del Lago Vichuqueacuten queda enriquecido en 15N (δ15N ~11permil Cap 2 Fig 9) Estas
variaciones estacionales pueden estar evidenciando un cambio en la composicioacuten
de especies de POM desde especies fijadoras a especies que consumen el N de la
columna de agua Sin embargo para corroborar esta informacioacuten seriacutea deseable
contar con el anaacutelisis de la composicioacuten de especies de las muestras de agua
extraidas
Entre los resultados maacutes importantes estaacuten los valores isotoacutepicos del suelo
y la biomasa representativa de la cuenca que incluye un listado de las especies
nativas de la zona que hasta ahora no se contaba (Cap 2 Tabla en material
suplementario) Las especies colectadas despliegan valores de δ15Nfoliar maacutes
positivos (plantaciones forestales y pastizal ~89permil) que el suelo (35permil) salvo por
las especies nativas las que oscilan en torno a valores cercanos al atmosfeacuterico
(~14permil) La razoacuten isotoacutepica 15N14N refleja la dinaacutemica de intercambio de N entre la
vegetacioacuten y el suelo (Koba et al 2002) La discriminacioacuten es positiva en la mayoriacutea
de los sistemas bioloacutegicos por lo que las plantas tienen valores de δ15N maacutes bajos
que el suelo (Evans et al 2001) como sucede con las especies nativas en Lago
Vichuqueacuten En consecuencia las plantas quedan empobrecidas en 15N y conducen
a un enriquecimiento en 15N del suelo sobretodo si no hay reposicioacuten de N por otras
viacuteas (eg fijacioacuten de N) Por lo tanto los valores maacutes negativos de δsup1⁵Nfoliar de las
especies nativas pueden estar relacionados con el consumo preferencial de 14N del
128
suelo donde la discriminacioacuten isotoacutepica de la vegetacioacuten contra el 15N conduce a
valores del δsup1⁵Nsuelo maacutes altos (Houmlgberg 1997) Por el contrario los valores maacutes
positivos de δsup1⁵Nfoliar de las plantaciones forestales y pastos con respecto al suelo
puede deberse por una parte que el suelo no cuenta con mecanismos naturales de
reposicioacuten de N salvo por la descomposicioacuten de la hojarasca que puede ser maacutes
lenta en Pinus radiata que en bosque nativo (Lusk et al 2001) y por otra por el alto
impacto de los aportes de N (y otros nutrientes) derivado de las actividades
humanas (eg uso de fertilizantes) en el suelo
El alcance maacutes significativo de esta tesis se relaciona con un cambio en la
tendencia maacutes negativa de δsup1⁵N y el incremento del NT a partir de 1940 y a partir
de 1970 CE en ambos lagos La causa maacutes probable de esta tendencia es el
reemplazo de la cobertura natural o preexistente a 1940 CE por plantaciones
forestales
En la figura 2 se observa una siacutentesis de los principales procesos que
afectan la sentildeal isotoacutepica de la MO en los sedimentos lacustres de L Matanzas y
L Vichuqueacuten La entrada de N y MO desde la cuenca depende del tipo de cobertura
Las plantaciones forestales y el bosque nativo retienen maacutes nutrientes y sedimentos
en la cuenca mientras que la agricultura los favorece Sin embargo las praacutecticas
de manejo que incluyen la aplicacioacuten de N (y otros nutrientes) pueden aportar maacutes
nutrientes al lago que la cobertra de bosque nativo Cuando las actividades
forestales cubren una mayor superficie de la cuenca (1940 en adelante) δsup1⁵N oscila
en valores maacutes negativos y el NT tiende a incrementarse en los sedimentos
lacustres Cabe destacar que los valores de δsup1⁵N observados en los testigos
sedimentarios a fines del s XX no tienen precedente durante la conquista y colonia
espantildeola o durante el resto del periodo de la Repuacuteblica
129
Figura 2 Modelo de transferencia de N entre los ecosistemas terrestres y
acuaacuteticos El δ15N de los sedimentos lacustres depende de la interaccioacuten entre los
aportes de la cuenca y porcesos ldquoin-lakerdquo Flechas indican la direccioacuten del flujo de
N En azul las salidasentradas de 14N en rojo las salidasentradas de 15N δ15N de
la interaccioacuten plantas-suelo depende del tipo de cobertura de suelo
130
CONCLUSIONES GENERALES
La transferencia de N entre cuencas y lagos es un factor de control del ciclo
del N en los lagos y queda reflejado en la sentildeal isotoacutepica de los sedimentos
lacustres (Leng et al 2006 McLauchlan et al 2007) En Laguna Matanzas el
suelo de las especies nativas y las plantaciones forestales despliegan valores de
δ15N maacutes bajos (~1permil) que los de Lago Vichuqueacuten (~35permil) Coincidentemente los
sedimentos lacustres de Laguna Matanzas oscilan con valores de δ15N maacutes bajos
(-15 a +45permil Fig 1) que los de Lago Vichuqueacuten (+3 a +8permil)
Por otra parte el estudio de la MO de los sedimentos lacustres ha permitido
reconstruir los cambios en los aportes de MO y N desde la cuenca Aunque no es
posible afirmar queacute tipo de N estaacute entrando al lago (eg Nitroacutegeno orgaacutenico e
inorganico) si podemos decir que los cambios en las fluctuaciones de δ15N son
coincentes con el crecimiento de la actividad forestal (1980 - hasta 2016 L
Matanzas 0-17 y L Vichuqueacuten 1-66) El δ15N fluctuacutea hacia valores maacutes
negativos Ademaacutes se observa un incremento del NT en los sedimentos lacustres
cuanto mayor es la superficie forestal
Entre 1800-1940 las cuencas eran fundamentalmente agriacutecolas y
ganaderas (Cap1 Fig 7 y Cap 2 Fig 12) el δ15N de los sedimentos lacustres
oscila con valores maacutes positivos (L Matanzas +40 plusmn 05permil y L Vichuqueacuten 56 plusmn
033permil Fig 1) hay una baja bioproductividad en el lago (BrTi ~002 TOC~17)
lo que coincide con un bajo contenido de NT (~478 μg) Este es un periacuteodo de altas
precipitaciones (Christie et al 2009) que tienden a favorecer el lavado de la cuenca
131
y con ello el aporte de MO sedimentos y nutrientes desde la cuenca tiende a verse
favorecido Aunque las principales actividades humanas en estas cuencas son
diferentes (ganaderiacutea en Laguna Matanzas Contreras-Lopez et al 2014
agricultura en Lago Vichuqueacuten Odone 1997) en ambos casos hubo un reemplazo
de la cobertura natural de bosques nativo que favorecioacute mayores aportes de N y
sedimentos desde la cuenca en un efecto sumado con el aumento de las
precipitaciones
A partir de 1980 CE en ambos lagos se observa una disminucioacuten en los
valores de δ15N y un incremento del NT que no tiene precedentes en todo el registro
y pese a que ambos lagos son limnologicamente muy diferentes En Lago
Vichuqueacuten el δ15N tiende a oscilar en antifase con el NT (Fig 1) En el caso de
Laguna Matanzas se observa una tendencia hacia valores maacutes negativos y a partir
de 1990s a valores maacutes positivos mientras que el NT tiende a incrementarse de
manera sostenida Ello podriacutea estar relacionado con el crecimiento de la actividad
forestal habriacutea una mayor retencioacuten de N y sedimentos en la cuenca ligado al
incremento de la superficie forestal Ademaacutes en el caso de L Matanzas el
incremento de la actividad agriacutecola (sumado al de las plantaciones forestales)
podriacutea expicar el cambio en la tendencia en la deacutecada de los 1990s
En el contexto de Antropoceno esta tesis nos permite identificar un gran
impacto de las actividades humanas en Chile central a partir de la deacutecada de 1940
y una Gran Aceleracoacuten a partir de la deacutecada de 1970s En el registro sedimentario
de los lagos en estudio se observa una gran acumulacioacuten de MO el δ15N oscila
hacia valores maacutes negativos y un gran incremento del NT y el crecimiento de la
actividad forestal parece ser la principal responsable Ademaacutes la sobreexplotacioacuten
132
del agua junto al cambio climaacutetico global ha generado una combinacioacuten letal para
los lagos costeros de Chile central
Figura 1 Comparacioacuten de las fluctuaciones de δ15N durante los uacuteltimos 300
antildeos en Laguna Matanzas y Lago Vichuqueacuten
133
Referencias
Battye W Aneja VP Schlesinger WH 2017 Earthrsquos Future Is nitrogen the next carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia
UC de V 2014 Elementos de la historia natural del An Mus Hist Natulas Vaplaraiso 27 51ndash67
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Evans RD Evans RD 2001 Physiological mechanisms influencing
plant nitrogen isotope composition 6 121ndash126 Fogel ML Cifuentes LA 1993 Isotope fractionation during primary
production Org Geochem 73ndash98 httpsdoiorg101007978-1-4615-2890-6_3 Galloway JN Schlesinger WH Ii HL Schnoor L Tg N 1995
Nitrogen fixation Anthropogenic enhancement-environmental response reactive N Global Biogeochem Cycles 9 235ndash252
Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J
Christie D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Koba K Koyama LA Kohzu A Nadelhoffer KJ 2003 Natural 15 N
Abundance of Plants Coniferous Forest httpsdoiorg101007s10021-002-0132-6
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos
Transactions American Geophysical Union httpsdoiorg1010292007EO070007 McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE
2007 Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
134
Phytologist N Oct N 2006 Tansley Review No 95 15N Natural Abundance in Soil-Plant Systems Peter Hogberg 137 179ndash203
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Turner BL Kasperson RE Matson PA McCarthy JJ Corell RW
Christensen L Eckley N Kasperson JX Luers A Martello ML Polsky C Pulsipher A Schiller A 2003 A framework for vulnerability analysis in sustainability science Proc Natl Acad Sci U S A httpsdoiorg101073pnas1231335100
Vitousek PM Aber JD Howarth RW Likens GE Matson PA
Schindler DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
Vitousek PM Howarth RW 1991 Nitrogen limitation on land and in the
sea How can it occur Biogeochemistry httpsdoiorg101007BF00002772 von Gunten L Grosjean M Rein B Urrutia R Appleby P 2009 A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 Holocene 19 873ndash881 httpsdoiorg1011770959683609336573
Xu R Tian H Pan S Dangal SRS Chen J Chang J Lu Y Maria
Skiba U Tubiello FN Zhang B 2019 Increased nitrogen enrichment and shifted patterns in the worldrsquos grassland 1860-2016 Earth Syst Sci Data httpsdoiorg105194essd-11-175-2019
Zaehle S 2013 Terrestrial nitrogen-carbon cycle interactions at the global
scale Philos Trans R Soc B Biol Sci 368 httpsdoiorg101098rstb20130125
2
LA DEFENSA FINAL DE LA TESIS DOCTORAL TITULADA
ldquoINFLUENCIA DE LOS CAMBIOS DE USOCOBERTURA DEL SUELO Y EL CLIMA EN EL CICLO DEL NITROacuteGENO EN DOS LAGOS COSTEROS DE CHILE CENTRAL A
PARTIR DE LA COLONIZACIOacuteN ESPANtildeOLArdquo
Presentada por la Candidata a Doctor en Ciencias Bioloacutegicas Mencioacuten Ecologiacutea de la Pontificia Universidad Catoacutelica de Chile
SRA MARIacuteA MAGDALENA FUENTEALBA LANDEROS
Ha sido aprobada por el Tribunal Examinador constituido por los profesores
abajo firmantes calificaacutendose el trabajo realizado el manuscrito sometido
y la defensa oral con nota _________ (_______________________________)
DR JOSEacute MIGUEL FARINtildeA R Coordinador Comiteacute de Tesis
Facultad de Ciencias Bioloacutegicas-UC
DR JUAN A CORREA M Decano
Facultad de Ciencias Bioloacutegicas-UC
DR BLAS VALERO G Co-Director de Tesis
Consejo Superior de Investigaciones Cientiacuteficas
DR CLAUDIO LATORRE H Director de Tesis
Facultad de Ciencias Bioloacutegicas-UC
DR JUAN ARMESTO Z Miembro del Comiteacute de Tesis
Facultad de Ciencias Bioloacutegicas-UC
DR RICARDO DE POL H Profesor Invitado
Universidad de Magallanes
Santiago de Chile 30 de septiembre de 2019-
3
TABLA DE CONTENIDO
RESUMEN 9
ABSTRACT 11
INTRODUCCIOacuteN 13
Los ecosistemas mediterraacuteneos y el ciclo del N 15
Los lagos como sensores ambientales 16
El ciclo del N en lagos 18
Reconstruyendo el ciclo del N a partir de variaciones en δ15N 20
Referencias 25
CAPIacuteTULO 1 A COMBINED APPROACH TO ESTABLISHING THE TIMING AND MAGNITUDE OF ANTHROPOGENIC NUTRIENT ALTERATION IN A MEDITERRANEAN COASTAL LAKE- WATERSHED SYSTEM 29
Abstract 30 1 INTRODUCTION 31 2 STUDY SITE 35 3 RESULTS 38
31 Age Model 38 32 The sediment sequence 39 33 Sedimentary units 41 34 Isotopic signatures 42 35 Recent land use changes in the Laguna Matanzas watershed 44
4 DISCUSSION 45 41 N and C dynamics in Laguna Matanzas 45 42 Recent evolution of the Laguna Matanzas watershed 48
5 CONCLUSIONS 54 6 METHODS 55
References 58
CAPIacuteTULO 2 STABLE ISOTOPES TRACK LAND USE AND COVER CHANGES IN A MEDITERRANEAN LAKE IN CENTRAL CHILE OVER THE LAST 600 YEARS 71
Abstract 73 1 Introduction 73
4
2 Study Site 77 3 Methods 79 4 Results 84
41 Geochemistry and PCA analysis 84 42 Sedimentary units 86 43 Recent seasonal changes of particulate organic matter on water column 88 44 Stable isotope values across the Lake Vichuqueacuten watershed 89 45 Land use and cover change from 1975 to 2014 90
5 Discussion 93 51 Seasonal variability of POM in the water column 93 52 Stable isotope signatures in the Lake Vichuqueacuten watershed 95 53 Recently land use and cover change and its influences on N inputs to the lake 97 54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years 98
6 Conclusions 101
LITERATURE CITED 110
DISCUSION GENERAL 117
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central 119
CONCLUSIONES GENERALES 130
Referencias 133
5
A mis padres Arturo y Malena
A mis hijos Xavi y Panchito
6
AGRADECIMIENTOS
Quiero agradecer a mi tutor y mentor Dr Claudio Latorre por brindarme su
apoyo sin el cual no habriacutea logrado concluir esta tesis de doctorado Claudio tu
apoyo constante incentivo y el fijarme metas que a veces me pareciacutean imposibles
de alcanzar no solo han dado forma a esta tesis sino tambieacuten me ha hecho maacutes
exigente como cientiacutefica Claudio destacas no solo por ser un gran cientiacutefico si no
tambieacuten por tu gran calidad humana eres un gran ejemplo
Quiero agradecer Dr Blas Valero-Garceacutes por nuestras numerosas
conversaciones viacutea Skype que incluiacutean vacaciones y fines de semana para discutir
los resultados de la tesis y que han dado forma a esta investigacioacuten principalmente
al primer capiacutetulo Ademaacutes por haberme acogido como un miembro maacutes en el
laboratorio de Paleoambientes Cuaternarios durante las estancias que he realizado
en el transcurso de estos antildeos Blas eres un ejemplo para miacute conjugas ciencia de
calidad calidez y dedicacioacuten por tus estudiantes
A quienes han financiado mi doctorado la Comisioacuten Nacional de
Investigacioacuten Cientiacutefica y Tecnoloacutegica (CONICYT) con sus becas de manutencioacuten
doctoral (2013) gastos operacionales pasantiacutea (2016) postnatal (2017) y termino
de tesis doctoralrdquo (2013) A FONDECYT a traveacutes del proyecto 1160744 de C
Santoro Al Departamento de InvestigacioacutenAl Instituto de Ecologiacutea y Biodiversidad
(IEB) a traveacutes de del PIA financiamiento basal 170008 la Pontificia Universidad
Catoacutelica de Chile por la beca incentivo para tesis interdisciplinaria para doctorandos
(2015)
Agradezco a mis compantildeeros del laboratorio de Paleoecologiacutea y
Paleoclimatologiacutea Karla Matias Dani Carolina Mauricio y Pancho que han hecho
grato mi tiempo en el laboratorio Agradecimientos especiales a Carolina Matiacuteas y
Leo por acompantildearme a terreno
7
Agradezco a los miembros del laboratorio de Paleoambientes Cuaternarios
(IPE) en especial a Raquel Elena Fernando Ana Penelope Graciela Miguel
Sevilla Mariacutea y Miguel Bartolomeacute
Agradezco a los miembros de la comisioacuten Dr Joseacute Miguel Farintildea Dr Juan
Armesto y Dr Ricardo De Pol-Holz por la gran ayuda durante el desarrollo de mi
doctorado en especial por las correcciones finales de la tesis
Al departamento de Ecologiacutea en especial a Rosario Galaz Edita Luengo
Daniela Mora y Valeria Cavallero por su apoyo
A mis compantildeeros de doctorado Beleacuten Gallardo Pancho Diacuteaz Bryan Bularz
Bian Latorre Renato Borras Juan Carlos Hernaacutendez y Paulina Troncoso con
quienes estudieacute aprendiacute y compartimos muy buenos momentos durante los
primeros antildeos del doctorado
A Pablo Sarricolea mi compantildeero de vida Gracias por tu constante e
incondicional apoyo Sin ti durante este proceso todo habriacutea sido cuesta arriba
A mis padres Arturo y Malena gracias por su carintildeo apoyo y compromiso
Papaacute aunque no esteacutes a mi lado siempre te recuerdo con mucho carintildeo A mi
madre que ha sido un constante apoyo sobre todo en el cuidado de mis hijos xavi
y panchito
A mis hermanos Rodrigo y David por estar presentes durante toda esta
etapa Siempre con carintildeo y hermandad
A Rodrigo David Matiacuteas Jany Paola y Julio por cuidar a mis hijos con carintildeo
siempre que estuve ausente por el doctorado
8
ABREVIATURAS
N Nitroacutegeno (Nitrogen)
DIN Nitroacutegeno Inorgaacutenico Disuelto (Dissolved Inorganic Nitrogen)
C Carbono (Carbon)
TOC Carbono Inorgaacutenico Total (Total Organic Carbon)
TIC Carbono Inorgaacutenico Total (Total inorganic Carbon)
TC Carbono Total (Total Carbon)
TS Azufre Total (Total Sulfur)
LUCC Cambios de UsoCobertura del Suelo (Land Use and Cover Change)
OM Materia Orgaacutenica (Organic Matter)
POM Particulate Organic Matter (materia orgaacutenica particulada)
CE Common Era
BCE Before Common Era
Cal BP Calibrado en antildeos radiocarbono antes de 1950
ie id est (esto es)
e g Exempli gratia (por ejemplo)
9
RESUMEN
El ldquoAntropocenordquo se caracteriza por pertubaciones de origen humano que
conllevan impactos globales Por ejemplo los cambios de usocobertura del suelo
(LUCC) son conocidos por perturbar el ciclo del N a partir de la Revolucioacuten Industrial
pero especialmente desde la Gran Aceleracioacuten (1950 CE) en adelante Sin
embargo existen incertezas asociadas a la magnitud del impacto y su efecto
acoplado con los cambios climaacuteticos actuales (ie megasequiacutea disminucioacuten de las
precipitaciones) y acoplamientos con otros ciclos biogeoquiacutemicos (eg ciclo del
Carbono) Este impacto ha cambiado la disponibilidad de N en los ecosistemas
terrestres y acuaacuteticos y sus enlaces Los sedimentos lacustres contienen
informacioacuten de las condiciones paleoambientales del lago y su cuenca en el
momento en que se depositaron y el anaacutelisis de los isoacutetopos estables de N (δ15N)
en los sedimentos permiten reconstruir los cambios en la disponibilidad de N a
traveacutes del tiempo En este trabajo usamos una aproximacioacuten multiproxy que incluye
10
anaacutelisis sedimentoloacutegicos geoquiacutemicos e isotoacutepicos (δ15N δ13C) de los sedimentos
lacustres columna de agua y suelo-vegetacioacuten de la cuenca y la reconstruccioacuten de
los LUCC entre 1975 y 2016 a partir de imaacutegenes satelitales El objetivo de esta
tesis fue evaluar el rol de LUCC como principal control del ciclo del N en el sistema
cuenca-lago durante las uacuteltimas centurias en Chile central Nuestros principales
resultados muestran que valores maacutes positivos de δ15N en los sedimentos lacustres
estaacuten relacionados con grandes aportes de N de la cuenca que a su vez son
mayores cuanto mayor la superficie agriacutecola yo los pastizales mientras que tanto
las plantaciones forestales como los bosques favorecen la retencioacuten de nutrientes
en la cuenca (lo que se ve reflejado en un δ15N maacutes negativo) Esta tesis plantea
un cambio de estado en la dinaacutemica del N asociado a la introduccioacuten y expansioacuten
de las plantaciones forestales o bosques los que retienen el Nitroacutegeno en las
cuencas mientras que el clima juega un rol secundario
11
ABSTRACT
The Anthropocene is characterized by human disturbances at the global
scale For example changes in land use are known to disturb the N cycle since the
industrial revolution but especially since the Great Acceleration (1950 CE) onwards
This impact has changed N availability in both terrestrial and aquatic ecosystems
However there are some important uncertainties associated with the extent of this
impact and how it is coupled to ongoing climate change (ie megadroughts rainfall
variability and shifting stationarity) or to other nutrient cycles (e g carbon cycle)
Lake sediments contain paleoenvironmental information regarding the conditions of
the watershed and associated lakes and which the respective sediments are
deposited Nitrogen stable isotope analysis (δ15N) on lake sediments allows us to
reconstruct the changes in N availability through time Here we used a multiproxy
approach that uses sedimentological geochemical and isotopic analyses on
lacustrine sediments water column and soilvegetation from the watershed as well
12
as land use and cover change (LUCC) from 1975 to 2016 obtained from satellite
images The goal of this thesis is to evaluate the role of LUCC as the main driver for
N cycling in a coastal watershed system of central Chile over the last centuries Our
main results show that more positive δ15N values in lake sediments are related to
higher N contributions from the watershed which in turn increase with increased
agricultural andor pasture cover whereas either forest plantations or native forests
can favor nutrient retention in the watershed (δ15N more negative) This thesis
proposes that N dynamics are mainly driven by the introduction and expansion of
forest or tree plantations that retain nitrogen in the watershed whereas climate plays
a secondary role
13
INTRODUCCIOacuteN
El N es un elemento esencial para la vida y limita la productividad en
ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al 1997) Las actividades
humanas han tenido un profundo impacto sobre el ciclo del N global principalmente
a partir la revolucioacuten industrial (IPCC 2013 2007) Las concentraciones de N se
han incrementado por la fijacioacuten industrial de N atmosfeacuterico (viacutea el proceso Haber-
Bosch Erisman et al 2008) por el uso de fertilizantes sobre la base de N para
mejorar el rendimiento de los cultivos la produccioacuten ganadera y ademaacutes de los
cambios en el uso del suelo (abreviado en ingleacutes- LUCC) (Robertson y Vitousek
2009 Galloway et al 2008 Battye et al 2017 Xu et al 2019) Estas actividades
contribuyen significativamente al incremento de la disponibilidad bioloacutegica de N
cuyas consecuencias para los ecosistemas incluye la perdida de diversidad
modificacioacuten de la disponibilidad de nutrientes y acidificacioacuten de los suelos entre
otras Sin embargo se desconoce la cantidad destino y el alcance que ha tenido
14
el N movilizado entre los ecosistemas generado por la influencia de las actividades
humanas (Vitousek et al 1997)
La entrada natural de N en los ecosistemas terrestres y acuaacuteticos es viacutea
fijacioacuten atmosfeacuterica la que es llevada a cabo por microorganismos muchos de ellos
en relaciones simbioacuteticas (eg Rhizobium en leguminosas) y las algas (Vitousek et
al 1997 Battye et al 2017) Las principales salidas estaacuten relacionadas con la
desnitrificacioacuten volatilizacioacuten del amoniaco y por escurrimiento superficial y
subterraacuteneo (Galloway et al 2008 Diacuteaz et al 2016) En el caso de los sistemas
lacustres ademaacutes estaacuten la resuspencioacuten de N del fondo del lago los aportes desde
la cuenca (entradas de N) La depositacioacuten de MO en el fondo del lago que marca
la salida de N de la columna de agua Estas relaciones de intercambio de N tienen
un control geograacutefico (eg latitud exposicioacuten relieve etc) climaacutetico y biotico
(McLauchlan et al 2013) La alteracioacuten provocada por los LUCC no solo genera
las peacuterdidas de nutrientes del suelo por erosioacuten y escurrimiento superficial sino que
tambieacuten modifica la disponibilidad y transferencia de N entre los ecosistemas
terrestres y acuaacuteticos (Elbert et al 2012 McLauchlan et al 2013) Por ejemplo el
reemplazo de la vegetacioacuten nativa por la agricultura yo plantaciones forestales
altera el ciclo del N debido al despeje de los terrenos viacutea quema de la vegetacioacuten
pero tambieacuten debido al reemplazo de la vegetacioacuten preexistente por un
monocultivo (eg trigo pino) y el uso ineficiente de fertilizantes para mejorar el
rendimiento agriacutecola Estas actividades favorecen un incremento de los aportes de
N (y otros nutrientes) viacutea sistemas de drenajes a lagos los que actuacutean como
sumideros El incremento del N derivado de las actividades humanas tanto en los
ecosistemas terrestres como acuaacuteticos genera incertezas relacionados con la
trayectoria y la magnitud de la disponibilidad de N en ambos ecosistemas (Batye et
15
al 2017 Mclauchlan et al 2017) Por lo tanto se requiere de una liacutenea base de
largo plazo para saber coacutemo estas perturbaciones impactan la disponibilidad de N
en las uacuteltimas deacutecadassiglos en los ecosistemas lacustres y por lo tanto el alcance
real que los LUCC han tenido en el ciclo del N
Los ecosistemas mediterraacuteneos y el ciclo del N
Los ecosistemas mediterraacuteneos son especialmente sensibles a los LUCC
pues estos se encuentran sometidos a un estreacutes hiacutedrico durante largas temporadas
estivales y las precipitaciones se concentran en eventos puntuales y a veces con
altos montos pluviomeacutetricos Las precipitaciones tienen un alto potencial de arrastre
de sedimentos y nutrientes los que actuacutean como fertilizantes al llegar a los
ecosistemas lacustres A su vez un incremento en la disponibilidad de N puede
generar un desbalance con otros nutrientes (eg Fosforo) con efectos sobre la
productividad y diversidad de los lagos (Pentildeuelas et al 2012 Sardans et al 2012
McLauchlan 2013) En este sentido en Chile central se observa una disminucioacuten
de las precipitaciones especialmente fuerte en las uacuteltimas deacutecadas por lo que se ha
denoacuteminado como Megasequiacutea (Garreaud et al 2017) y el efecto de las
precipitaciones esporaacutedicas pueden generar un arrastre de nutrientes que no ha
sido evaluado
Los ecosistemas mediterraacuteneos concentran el 20 de la biodiversidad global
(Myers et al 2000) pero existe una escasez de conocimiento respecto a los
efectos del incremento de N en los cuerpos de agua como consecuencia de las
actividades humanas en el pasado histoacuterico (Ochoa-Hueso et al 2011) y coacutemo la
disponibilidad de N ha variado en el tiempo Los LUCC han aumentado el flujo de
N en las cuencas hidrograacuteficas viacutea escurrimiento superficial y lixiviacioacuten
16
favoreciendo la peacuterdida de sedimentos y nutrientes del suelo en el mundo entero
(Elser et al 2011 McLauchlan et al 2013 Xu et al 2017 y 2019) Ello ha
contribuido a la acidificacioacuten y eutrofizacioacuten generalizada de riacuteos y lagos
(McLauchlan et al 2013 Schindler et al 2008)
El ecosistema mediterraacuteneo de Chile central es una regioacuten ampliamente
intervenida desde al menos la colonizacioacuten espantildeola (1600-1800 CE) Los LUCC
han tenido efectos negativos en la disponibilidad de agua especialmente
observados con el desarrollo de las actividades minera (Prieto et al 2019) Aunque
se sabe de los impactos en los ecosistemas de bosque en teacuterminos de cobertura
debidos al desarrollo silvo-agropecuarias (Armesto et al 2010) se desconoce el
impacto de estas actividades sobre el ciclo del N en los cuerpos de agua o en queacute
momento cobran fuerza suficiente para alterar el flujo de N hacia los lagos de Chile
Central Por lo tanto en esta tesis nos centramos en entender coacutemo los LUCC han
afectado la disponibilidad de N en los lagos costeros de Laguna Matanzas y Lago
Vichuqueacuten desde la llegada de los espantildeoles a Chile hasta el presente
Los lagos como sensores ambientales
Los sedimentos lacustres son buenos sensores de cambios en los aportes
de nutrientes contaminacioacuten y eutroficacioacuten de los cuerpos de agua ya que son
capaces de preservar sentildeales quiacutemicas y fiacutesicas al momento de su formacioacuten y
ademaacutes con alta resolucioacuten y a largo plazo (Leng et al 2006) Por lo tanto
constituyen una herramienta uacutetil para reconstruir el flujo de N entre ecosistemas
terrestres y acuaacuteticos en el pasado sus consecuencias (eg incremento de la
productividad lacustre) y el rol del clima y los LUCC en el pasado (McLauchlan et
al 2013 von Gunten et al 2009) Por ejemplo el cultivo de maiacutez por parte de los
17
nativos iroqueses en Canadaacute provocoacute la eutrofizacioacuten del Lago Crawford (43degN)
durante los siglos XII y XV Entre las consecuencias registradas se documentoacute un
claro incremento de la productividad primaria y cambios en la estructura comunitaria
de diatomeas foacutesiles (incremento de Stephanodiscus complex y disminucioacuten de
Cyclotella bodanica en Ekdahl et al 2007) Otro ejemplo demuestra como las
actividades agriacutecolas en la cuenca del riacuteo Mississippi modificaron los patrones de
sedimentacioacuten y aporte de nutrientes al lago Pepin EE UU (44degN) a partir del
asentamiento europeo en 1840 CE y principalmente desde 1940 CE (Engstrom et
al 2009) Para Chile von Gunten et al (2009) a partir de indicadores
limnogeoloacutegicos evidencioacute la contaminacioacuten y eutroficacioacuten de cinco lagos cercanos
a la ciudad de Santiago (33ordmS) como consecuencia de la deposicioacuten atmosfeacuterica
de metales pesados (especialmente Cu) quema de combustible foacutesil y flujo de
nutrientes durante los uacuteltimos 200 antildeos
Caracteriacutesticas limnoloacutegicas de los lagos
Los procesos limnoloacutegicos afectan la distribucioacuten y abundancia de los
organismos en los lagos Estaacuten influenciados por forzamientos externos por
ejemplo la precipitacioacuten o la penetracioacuten de la luz y la morfometriacutea del lago En este
sentido el nivel del lago dependeraacute del balance entre las entradas y salidas de agua
(precipitaciones escurrimiento superficial y subterraacuteneo y la evaporacioacuten) y la forma
de la cuenca (profundidad pendiente aacuterea del espejo de agua)
En funcioacuten de la profundidad de penetracioacuten de la luz es posible diferenciar
dos aacutereas que determinan la distribucioacuten de los organismos la zona foacutetica (donde
penetra la luz y permite la fotosiacutentesis) y la zona afoacutetica (subyacente a la zona
foacutetica) Al depender de la radiacioacuten la zona foacutetica variacutea estacionalmente Ademaacutes
18
puede variar por el grado de turbidez la morfometriacutea del lago y la produccioacuten de
materia orgaacutenica en la columna de agua
Otro factor que influye en la productividad es el reacutegimen de mezcla de la
columna de agua pues favorece la oxigenacioacuten y el reciclado de nutrientes La
mezcla ocurre en condiciones de gran inestabilidad usualmente relacionado con el
reacutegimen de viento Por el contrario un lago estratificado resulta de grandes
diferencias en temperatura o salinidad entre la superficie (epilimnion) y el fondo del
lago (hipolimnion) que separa las masas de agua superficial y de fondo por una
termoclina (si la estratificacioacuten estaacute relacionada con diferencias de temperatura de
las masas de agua) o haloclina (diferencias de salinidad) En funcioacuten del reacutegimen
de mezcla los lagos se pueden clasificar en (Lewis 1983)
1 Amiacutecticos no hay mezcla vertical de la columna de agua
2 Monomiacutecticos friacuteos y caacutelidos se mezcla una vez al antildeo
3 Dimiacutecticos la columna de agua se mezcla dos veces al antildeo
4 Oligomiacutecticos Lagos estratificados la mayor parte del tiempo pero se mezclan a
intervalos irregulares mayores a 1 antildeo
5 Polimiacutecticos friacuteos y caacutelidos la columna de agua se mezcla varias veces al antildeo
El ciclo del N en lagos
Al estar presente en aacutecidos nucleicos aminoaacutecidos y proteiacutenas el N es un
nutriente esencial de los organismos Por lo tanto su disponibilidad en la columna
de agua es clave para la produccioacuten composicioacuten y acumulacioacuten de la MO a traveacutes
del tiempo (Olson 1998 Talbot 2002) Aunque el principal reservorio de N estaacute en
19
la atmosfera (N2) solo unos pocos organismos (eg cianobacterias) pueden fijarlo
directamente Asiacute el Nitroacutegeno Inorgaacutenico Disuelto (DIN) representa la principal
fuente de N disponible para la produccioacuten primaria en los ecosistemas acuaacuteticos
(Talbot et al 2002) El DIN estaacute compuesto por nitrato (NO3-) nitrito (N02
-) y amonio
(NH4+) y los cambios en su disponibilidad condicionan el tipo de produccioacuten primaria
(eg fijadores vs asimiladores de N ver Scott y Grant 2013 Scott 2008)
La Figura 1 resume los principales componentes en lagos del ciclo del N y
sus interacciones La principal entrada natural de N es viacutea fijacioacuten de N atmosfeacuterico
y es llevado a cabo principalmente por bacterias y algas verde-azules capaces de
romper el triple enlace entre los dos aacutetomos de N a un alto costo energeacutetico (Torres
et al 2012) El N fijado es reducido a NH3 Tras la muerte de los organismos el N
es liberado de sus tejidos por hongos y bacterias implicados en la descomposicioacuten
de la MO Una parte se deposita en el fondo del lago y la otra queda disponible para
ser asimilada por el fitoplancton como amonio mediante el proceso de
amonificacioacuten (Talbot 2002) La amonificacioacuten resulta de la reduccioacuten bacteriana
del amoniaco y se da preferentemente en condiciones anoacutexicas La volatizacioacuten del
amoniaco es un proceso por medio este se incorpora a la atmoacutesfera Este proceso
se da preferentemente en lagos aacutecidos (pH gt 85) El nitrato es otra forma de N
bioloacutegicamente disponible para el fitoplancton En los lagos el NO3 del DIN estaacute
compuesto por los aportes desde la cuenca hidrograacutefica en la que se circunscriben
por la resuspensioacuten desde el fondo del lago favorecido por procesos de mezcla
(descrito en seccioacuten anterior) y por los nitratos de la columna de agua Estos uacuteltimos
son el resultado de la nitrificacioacuten proceso realizado por bacterias aeroacutebicas
mediante el cual el amonio se oxida para formar nitrito y nitrato El ciclo se completa
20
con la desnitrificacioacuten que es la reduccioacuten bacteriana de nitrato al N atmosfeacuterico
Este proceso se da preferentemente en condiciones anoacutexicas
Figura 1 Materia Orgaacutenica sedimentaria y su relacioacuten con el ciclo del N y las
variaciones de δ15N (elaborado a partir de Talbot et al 2002) En la figura se
representan los factores clave en la acumulacioacuten de la MO sedimentaria y su
relacioacuten con el ciclo del N El δ15N de la MO es resultado de los aportes de MO
desde la cuenca MO de macroacutefitas y de la productividad bioloacutegica La productividad
en ecosistemas lacustres estaacute condicionada por DIN y la fijacioacuten de N atmosfeacuterico
El δ15N del pool del DIN disponible se va enriqueciendo por la asimilacioacuten
preferencial de 14N por parte del fitoplancton Ademaacutes el DIN remanente se va
enriqueciendo por accioacuten microbiana (amonificacioacuten nitrificacioacuten desnitrificacioacuten)
Reconstruyendo el ciclo del N a partir de variaciones en δ15N
La sentildeal isotoacutepica de N (δ15N) en los sedimentos lacustres puede ser usada
para reconstruir los cambios pasados del ciclo N la transferencia de N entre
ecosistemas terrestres y acuaacuteticos y el estado troacutefico del lago (Bruland y Mackenzie
2015 McLauchlan et al 2013 Woodward et al 2012 von Gunten et al 2009
Leng et al 2006 Meyers y Teranes 2001) La Figura 1 resume los principales
procesos y fuentes que afectan la sentildeal isotoacutepica δ15N (total o ldquobulkrdquo) en la MO de
21
los sedimentos lacustres Entre los que destacan las fuentes de MO (aloacutectona vs
autoacutectona) la bioproductividad lacustre y las entradas de N (ie fijacioacuten bioloacutegica
de N2) (Torres et al 2012) Estos procesos conducen a un fraccionamiento
isotoacutepico cineacutetico que favorece la disociacioacuten entre sus isotopos ldquopesadosrdquo (15N) y
ldquoligerosrdquo (14N) Asiacute por ejemplo los valores de δ15N de la MO se enriquecen en 15N
en la medida que los sedimentos lacustres estaacuten sometidos a perdidas de 14N viacutea
desnitrificacioacuten (Bruland y Mackenzie 2015 Diaz et al 2016 Talbot et al 2002)
Por otra parte el fitoplancton en la medida que aumenta su biomasa (eg
durante la estacioacuten caacutelida) puede llegar a asimilar el DIN hasta agotarse En este
caso el N remanente se va empobreciendo en 14N si no hay reposicioacuten de N (eg
aporte de NO3 de la cuenca) y la MO queda enriquecida en 15N Esta disminucioacuten
induce a una limitacioacuten de fitoplancton por N y los organismos diztroacuteficos pueden
verse estimulados por lo que se incrementa la entrada de N viacutea fijacioacuten de N2 (Scott
y Grantsz 2013) como resultado la MO oscila en valores maacutes bajos (en torno a 0permil)
La cantidad de MO que se deposita en el fondo del lago depende del
predominio de contribuciones provenientes desde la cuenca (aloacutectona) versus las
producidas dentro del mismo lago (autoacutectona) (Torres et al 2012) Aunque en
general los lagos reciben permanentemente aportes de sedimentos y MO desde
su cuenca los lagos pequentildeos tienden a reflejar maacutes los procesos que ocurren
solamente en su cuenca directa y reflejan los valores isotoacutepicos de la misma (Gu et
al 2006) Woodward et al (2012) realizaron un estudio en 50 lagos en Irlanda que
les permitioacute correlacionar diferentes patrones de LUCC (bosques de coniacuteferas
agricultura ganaderiacutea entre otros) con los valores de δ15N (y δ13C) en los
sedimentos superficiales de los lagos Encontraron que los valores de δ15N son maacutes
negativos en cuencas poco impactadas y maacutes positivos en aquellas con alto
22
impacto humano (eg usos de suelo agriacutecola y ganadero) McLauchlan et al (2007)
encontraron valores maacutes positivos de δ15N en los sedimentos del Mirror Lake New
Hampshire (29ordm N) a partir de la desforestacioacuten de su cuenca en 1790 CE (inicio
del asentamiento europeo en el lago) para el desarrollo de la actividad agriacutecola
Estos valores se volvieron maacutes negativos hacia valores similares al pre-
asentamiento europeo despueacutes del cese de la agricultura en la cuenca y la
recuperacioacuten del bosque a partir de 1929 CE
El anaacutelisis de δ15N en los sedimentos lacustres es una herramienta casi sin
explorar en Chile central Proponemos reconstruir la dinaacutemica de transferencia de
N cuenca-lago como consecuencia de los LUCC desde la colonizacioacuten espantildeola en
los lagos costeros de Laguna Matanzas (34ordmS) y Lago Vichuqueacuten (36ordmS) Como son
muacuteltiples los procesos que pueden afectar la sentildeal isotoacutepica de N (medido como
δ15N) en los sedimentos lacustres existen muchos problemas para su
interpretacioacuten paleoambiental (Leng et al 2006) Por ello en esta tesis utilizamos
un enfoque multiproxy que consiste en integrar el anaacutelisis limnoloacutegico y geoquiacutemico
de los sedimentos lacustres con los valores isotoacutepicos modernos de la columna de
agua (mediante monitoreo del POM) la relacioacuten vegetacioacuten-suelo de la cuenca y la
reconstruccioacuten de los principales usos de suelo de las cuencas desde 1975 CE
mediante el uso de imaacutegenes satelitales Considerando estas muacuteltiples liacuteneas de
evidencia nos planteamos utilizar las variaciones del δ15N para reconstruir los
cambios en la disponibilidad de N en los lagos costeros de Chile central y establecer
coacutemo y desde cuaacutendo las actividades humanas en la cuenca son lo suficientemente
importantes como para modificar el ciclo del N en los lagos desde la colonizacioacuten
espantildeola (siglo XVII)
23
Como hipoacutetesis planteamos que la dinaacutemica de transferencia de sedimentos
y nutrientes entre la cuenca y el lago tiene controles geoloacutegicos climaacuteticos y
bioloacutegicos que interactuacutean en el tiempo registraacutendose en la acumulacioacuten de MO de
los sedimentos lacustres (Fig1) Bajo este marco los LUCC modifican esta
dinaacutemica alterando la transferencia de N a los lagos ya que las actividades agriacutecolas
y ganaderas tienden a aportar maacutes N (aumentando δ15N) que la actividad forestal
(la que disminuye δ15N)
En el primer capiacutetulo titulado ldquoA combined approach to establishing the timing
and magnitude of anthropogenic nutrient alteration in a mediterranean coastal lake-
watershed systemrdquo se evaluoacute el rol de los LUCC en el control de la descarga de N
y sedimentos a Laguna Matanzas en un sistema cuenca-lago desde el siglo XVII
Para ello se realizoacute un anaacutelisis de la dinaacutemica sedimentaria lacustre mediante el
anaacutelisis de muacuteltiples proxys sedimentoloacutegicos (ie minerales y textura)
geoquiacutemicos (eg geoquiacutemica orgaacutenica e isotoacutepica) y bioloacutegicos (ie diatomeas) de
Laguna Matanzas que cubren los uacuteltimos 200 antildeos Ademaacutes se realizoacute una
reconstruccioacuten de los usos de suelo entre 1975 y 2016 a partir de imaacutegenes de
sateacutelites y se colectaron muestras de suelo de las principales coberturas de la
cuenca a los cuales se midioacute el δ15N
Entre los principales resultados obtenidos se destaca la influencia de la
ganaderiacutea y la denitrificacioacuten lo que explica los altos valores de δ15N evidenciados
por el registro lacustre durante la colonizacioacuten espantildeola e inicios de la repuacuteblica A
partir de la Gran Aceleracioacuten (1950 CE) la desforestacioacuten y el reemplazo de la
ganaderiacutea por plantaciones forestales tienen un correlato en el registro
sedimentario con valores de δ15N maacutes bajos Estos resultados demuestran que los
LUCC son el factor de primer orden para explicar los cambios observados en
24
nuestro registro de δ15N y a su vez implica un rol secundario del clima como posible
control de la disponibilidad de N en Laguna Matanzas Este capiacutetulo fue sometido
a la revista Scientific Reports y estaacute en revisioacuten desde el 26 de marzo de 2019 En
la elaboracioacuten del mauscrito han colaborado Claudio Latorre Blas Valero-Garceacutes
Matiacuteas Frugone Pablo Sarricolea Santiago Giralt Manuel Contreras-Lopez
Ricardo Prego y Patricia Bernardez
El segundo capiacutetulo se titula ldquoStable isotopes track land use and cover
changes in a mediterranean lake in central Chile over the last 600 yearsrdquo Aquiacute
evaluamos la dinaacutemica del N actual en el sistema cuenca-lago considerando los
valores isotoacutepicos modernos tanto de la vegetacioacuten como del suelo ademaacutes de los
cambios estacionales del POM de la columna de agua Esta informacioacuten se utiliza
como un anaacutelogo moderno para conocer el rol de los LUCC en la disponibilidad de
N en los uacuteltimos 600 antildeos Para ello se colectaron muestras filtradas de la columna
de agua a 2 5 y 20 m dos veces por antildeo entre noviembre de 2015 y julio de 2018
y se obtuvo el δsup1⁵N del POM Ademaacutes se colectoacute muestras de vegetacioacuten y suelo
de la cuenca diferenciando entre especies nativas plantaciones forestales y
vegetacioacuten herbaacutecea Esta informacioacuten se analizoacute en conjunto con la reconstruccioacuten
de los LUCC entre 1975 y 2014 y el anaacutelisis limnoloacutegico del lago Ello nos permitioacute
evaluar coacutemo el δ15N de los sedimentos lacustres refleja los valores δ15N de la
cuenca en el Lago Vichuqueacuten y el control que ejerce el clima en la sentildeal isotoacutepica
de POM Este capiacutetulo seraacute sometido a la revista Science of total environmet
proximamente En la elaboracioacuten del mauscrito han colaborado Claudio Latorre
Blas Valero-Garceacutes Matiacuteas Frugone Pablo Sarricolea Leonardo Villacis y M Laura
Carrevedo
25
Entre los principales resultados encontramos que el δ15N en los sedimentos
lacustres es maacutes positivo con presencia de agricultura en la cuenca y maacutes negativo
cuando esta es reemplazada por cobertura arboacuterea bien sea plantaciones
forestales bien sea bosque nativo A su vez durante la estacioacuten seca y caacutelida la
mayor entrada al lago es viacutea fijacioacuten de N atmosfeacuterico (bajos valores de δ15NPOM)
Durante la estacioacuten huacutemeda en cambio prevalecen los aportes de la cuenca con
altos valores de δ15NPOM Ello estariacutea evidenciando un cambio estacional en la
composicioacuten de especies en verano ligado a especies fijadoras de N y en invierno
las algas y microorganismos que consumen el DIN de la columna de agua
Referencias
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea AM Marquet PA 2010 From the Holocene to the Anthropocene A historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the
next carbon Earth rsquo s Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409 httpsdoiorg102134jeq20090005
Diacuteaz FP Frugone M Gutieacuterrez RA Latorre C 2016 Nitrogen cycling in an
extreme hyperarid environment inferred from δ15 N analyses of plants soils and herbivore diet Sci Rep 6 1ndash11 httpsdoiorg101038srep22226
Ekdahl EJ Teranes JL Wittkop CA Stoermer EF Reavie ED Smol JP
2007 Diatom assemblage response to Iroquoian and Euro-Canadian eutrophication of Crawford Lake Ontario Canada J Paleolimnol 37 233ndash246 httpsdoiorg101007s10933-006-9016-7
Elbert W Weber B Burrows S Steinkamp J Buumldel B Andreae MO
Poumlschl U 2012 Contribution of cryptogamic covers to the global cycles of carbon and nitrogen Nat Geosci 5 459ndash462
26
httpsdoiorg101038ngeo1486 Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506
httpsdoiorg101126science1215567 ARTICLE Engstrom DR Almendinger JE Wolin JA 2009 Historical changes in
sediment and phosphorus loading to the upper Mississippi River Mass-balance reconstructions from the sediments of Lake Pepin J Paleolimnol 41 563ndash588 httpsdoiorg101007s10933-008-9292-5
Erisman J W Sutton M A Galloway J Klimont Z amp Winiwarter W (2008)
How a century of ammonia synthesis changed the world Nature Geoscience 1(10) 636
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892
httpsdoiorg101126science1136674 Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J Christie
D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592 Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
IPCC 2013 Climate Change 2013 The Physical Science BasisContribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change Cambridge University Press Cambridge United Kingdom and New York NY USA
httpsdoiorg101017CBO9781107415324 Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007 Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32iumliquestfrac12S 3050iumliquestfrac12miumliquestfrac12asl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470
27
httpsdoiorg101073pnas0701779104 McLauchlan KK Gerhart LM Battles JJ Craine JM Elmore AJ Higuera
PE Mack MC McNeil BE Nelson DM Pederson N Perakis SS 2017 Centennial-scale reductions in nitrogen availability in temperate forests of the United States Sci Rep 7 1ndash7 httpsdoiorg101038s41598-017-08170-z
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Myers N Mittermeler RA Mittermeler CG Da Fonseca GAB Kent J
2000 Biodiversity hotspots for conservation priorities Nature httpsdoiorg10103835002501
Pentildeuelas J Poulter B Sardans J Ciais P Van Der Velde M Bopp L
Boucher O Godderis Y Hinsinger P Llusia J Nardin E Vicca S Obersteiner M Janssens IA 2013 Human-induced nitrogen-phosphorus imbalances alter natural and managed ecosystems across the globe Nat Commun 4 httpsdoiorg101038ncomms3934
Prieto M Salazar D Valenzuela MJ 2019 The dispossession of the San
Pedro de Inacaliri river Political Ecology extractivism and archaeology Extr Ind Soc httpsdoiorg101016jexis201902004
Robertson GP Vitousek P 2010 Nitrogen in Agriculture Balancing the Cost of
an Essential Resource Ssrn 34 97ndash125 httpsdoiorg101146annurevenviron032108105046
Sardans J Rivas-Ubach A Pentildeuelas J 2012 The CNP stoichiometry of
organisms and ecosystems in a changing world A review and perspectives Perspect Plant Ecol Evol Syst 14 33ndash47 httpsdoiorg101016jppees201108002
Schindler DW Howarth RW Vitousek PM Tilman DG Schlesinger WH
Matson PA Likens GE Aber JD 2007 Human Alteration of the Global Nitrogen Cycle Sources and Consequences Ecol Appl 7 737ndash750 httpsdoiorg1018901051-0761(1997)007[0737haotgn]20co2
Scott E and EM Grantzs 2013 N2 fixation exceeds internal nitrogen loading as
a phytoplankton nutrient source in perpetually nitrogen-limited reservoirs Freshwater Science 2013 32(3)849ndash861 10189912-1901
28
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Talbot M R (2002) Nitrogen isotopes in palaeolimnology In Tracking
environmental change using lake sediments (pp 401-439) Springer Dordrecht
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable
isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K
Yeager KM 2016 Different responses of sedimentary δ15N to climatic changes and anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
29
CAPIacuteTULO 1 A COMBINED APPROACH TO ESTABLISHING THE TIMING
AND MAGNITUDE OF ANTHROPOGENIC NUTRIENT ALTERATION IN A
MEDITERRANEAN COASTAL LAKE- WATERSHED SYSTEM
30
A combined approach to establishing the timing and magnitude of anthropogenic
nutrient alteration in a mediterranean coastal lake- watershed system
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugoneabc Pablo
Sarricolead Santiago Giralte Manuel Contreras-Lopezf Ricardo Prego g Patricia
Bernaacuterdez g Blas Valero-Garceacutesch
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Institute of Earth Sciences Jaume Almera (ICTJA-CSIC) CLluis Solegrave Sabaris sn Barcelona E-
08028 Spain
f Facultad de Ingenieriacutea y Centro de Estudios Avanzados Universidad de Playa Ancha Traslavintildea
450 Vintildea del Mar Chile
g Instituto de Investigaciones Marinas (CSIC) CEduardo Cabello 6 36208 Vigo Spain
h Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding author
E-mail address
clatorrebiopuccl magdalenafuentealbagmailcom
Abstract
Since the industrial revolution and especially during the Great Acceleration (1950
CE) human activities have profoundly altered the global nutrient cycle through land
use and cover changes (LUCC) However the timing and intensity of recent N
variability together with the extent of its impact in terrestrial and aquatic ecosystems
and coupled effects of regional LUCC and climate are not well understood Here
we used a multiproxy approach (sedimentological geochemical and isotopic
31
analyses historical records climate data and satellite images) to evaluate the role
of LUCC as the main control for N cycling in a coastal watershed system of central
Chile during the last few centuries The largest changes in N dynamics occurred in
the mid-1970s associated with the replacement of native forests and grasslands for
livestock grazing by plantations of introduced Monterey pine (Pinus radiata) and
eucalyptus (Eucalyptus globulus) Lake productivity increased and is evidenced by
an increase trend in δ15N values Our study shows that anthropogenic land
usecover changes are key in controlling nutrient supply and N availability in
Mediterranean watershed ndash lake systems and that large-scale forestry
developments during the mid-1970s likely caused the largest changes in central
Chile
Keywords Anthropocene Organic geochemistry watershedndashlake system Stable
Isotope Analyses Land usecover change Nitrogen cycle Mediterranean
ecosystems central Chile
1 INTRODUCTION
Human activities have become the most important driver of the nutrient cycles in
terrestrial and aquatic ecosystems since the industrial revolution (Gruber and
Galloway 2008 Galloway et al 2008 Holtgrieve et al 2011 Fowler et al 2013
Goyette et al 2016) Among these N is a common nutrient that limits productivity
in terrestrial and aquatic ecosystems (Vitousek and Howarth 1991 McLauchlan et
al 2013) With the advent of the Haber-Bosch industrial N fixation process in the
early 20th century total N fluxes have surpassed previous planetary boundaries
32
(Galloway et al 2008 Elser 2011) reaching unprecedented values (ie tipping
points) in the Earth system especially during what is now termed the Great
Acceleration (which began in the 1950s) which intensified in the 1970s (Howarth
2004 Steffen et al 2015) Yet dramatic changes in LUCC have occurred in the last
few centuries mainly linked to forestry and agro-pastoral activities (Poraj-Goacuterska et
al 2017 Ge et al 2019 Cheng et al 2019) These have had large impacts on N
(and Carbon -C-) availability in terrestrial and aquatic ecosystems but the synergic
effect with climate change and global N dynamics has not been established
(Vitousek et al 1997 Mclauchlan et al 2007 2013 Pongratz et al 2010
Woodward et al 2012 Mclauchlan et al 2017)
The onset of the Anthropocene poses significant challenges in mediterranean
regions that have a strong seasonality of hydrological regimes and an annual water
deficit (Stocker et al 2013) Mediterranean climates occur in all continents
(California central Chile Australia South Africa circum-Mediterranean regions)
providing a unique opportunity to investigate global change processes during the
Anthropocene in similar climate settings but with variable geographic and cultural
contexts The effects of global change in mediterranean watersheds have been
analyzed from different perspectives hydrology (Giorgi 2006 Arnell and Gosling
2013 Prudhomme et al 2014) vegetation dynamics (Lenihan et al 2003 Vicente-
Serrano et al 2013 Matesanz and Valladares 2014) sediment dynamics (Garciacutea-
Ruiz et al 2013 Garciacutea-Ruiz 2010 Syvitski et al 2005) changes in
biogeochemical cycles (Thomas et al 2013 Fowler et al 2013 Frank et al 2015)
carbon storage (Muntildeoz-Rojas et al 2015) and biodiversity (Hooper et al 2012) A
recent review showed an extraordinarily high variability of erosion rates in
mediterranean watersheds positive relationships with slope and annual
33
precipitation and the paramount effect of land use (Garciacutea-Ruiz et al 2015)
However the temporal context and effect of LUCC on nutrient supply to
mediterranean lakes has not been analyzed in much detail
Major LUCC in central Chile occurred during the Spanish Colonial period
(17th-18th centuries) (Armesto 1994 Gurnell et al 2001 Figueroa et al 2004
Martiacuten-Foreacutes et al 2012 Contreras-Loacutepez et al 2014) with the onset of
industrialization and mostly during the mid to late 20th century (von Gunten et al
2009 Gayo et al 2012) Recent copper pollution caused by 20th century mining
and industrial smelters has been documented in cores throughout the Andes
(Laguna el Ocho and Laguna Ensuentildeo von Gunten et al 2009) and also from our
surveys in the central valley (Batuco wetland) coastal range (Cordillera de Name)
and along the coast itself (Bucalemito and Colejuda) (Valero-Garceacutes et al 2010
unpublished data)
Paleolimnological studies have shown how these systems respond to
climate LUCC and anthropogenic impacts during the last millennia (Jenny et al
2002 2003 Villa Martiacutenez et al 2002 Frugone-Aacutelvarez et al 2017 Contreras et
al 2018) Furthermore changes in sediment and nutrient cycles have also been
identified in associated terrestrial ecosystems dating as far back as the Spanish
Conquest and related to fire clearance and wood extraction practices of the native
forests (Armesto 1994 Contreras-Loacutepez et al 2014) Nevertheless pollen and
limnological evidence argue for a more recent timing of the largest anthropogenic
impacts in central Chile For example paleo records show that during the mid-20th
century increased soil erosion followed replacement of native forest by Pinus
radiata and Eucalyptus globulus plantations at Laguna Matanzas Aculeo and
34
Vichuqueacuten lakes (Jenny et al 2002 Villa-Martiacutenez et al 2002 2003 Frugone-
Aacutelvarez et al 2017)
Lakes are a central component of the global carbon cycle Lakes act as a
sink of the carbon cycle both by mineralizing terrestrially derived organic matter and
by storing substantial amounts of organic carbon (OC) in their sediments (Anderson
et al 2009) Paleolimnological studies have shown a large increase in OC burial
rates during the last century (Heathcote et al 2015) however the rates and
controls on OC burial by lakes remain uncertain as do the possible effects of future
global change and the coupled effect with the N cycle LUCC intensification of
agriculture and associated nutrient loading together with atmospheric N-deposition
are expected to enhance OC sequestration by lakes Climate change has been
mainly responsible for the increased algal productivity since the end of the 19th
century and during the late 20th century in lakes from both the northern (Ruumlhland et
al2015) and southern hemispheres (Michelutti et al 2015 Carrevedo et al 2015)
but many studies suggest a complex interaction of global warming and
anthropogenic influences and it remains to be proven if climate is indeed the only
factor controlling these transitions (Catalaacuten et al 2013) Alternative causes for
recent N (Galloway et al 2008) increases in high altitude lakes such as catchment
mediated processes cannot be ruled out (Camarero and Catalan 2012 Fritz and
Anderson 2013) Few lake-watershed systems have robust enough chronologies of
recent changes to compare variations in C and N with regional and local processes
and even fewer of these are from the southern hemisphere (McLauchlan et al
2007 Holtgrieve et al 2011)
In this paper we present a multiproxy lake-watershed study including N and
C stable isotope analyses on a series of short cores from Laguna Matanzas in
35
central Chile focused in the last 200 years We complemented our record with land
use surveys satellite and aerial photograph studies Our major objectives are 1) to
reconstruct the dynamics among climate human activities and changes in the N
cycle over the last two centuries 2) to evaluate how human activities have altered
the N cycle during the Great Acceleration (since the mid-20th century)
2 STUDY SITE
Laguna Matanzas (33ordm45rsquoS 71ordm 40acuteW 7 m asl Fig 1) is a coastal lake located
in central Chile near to a large populated area (Santiago gt6106 inhabitants) The
lake has a surface area of 15 km2 with a max depth of 3 m and a watershed of 30
km2 The lake basin is emplaced over Pleistocene-Holocene aeolian and alluvial fan
deposits (SERNAGEOMIN 2003) A recent phase of dune activity occurred from the
mid to late Holocene which mostly sealed off the basin from the ocean (Villa-
Martinez 2002) Climate in Laguna Matanzas is characterized by cool-wet winters
and hot-dry summers with annual precipitation of ~510 mm and a mean annual
temperature of 12ordmC Central Chile is in the transition zone between the southern
hemisphere mid-latitude westerlies belt and the South Pacific Anticyclone (or SPA)
(Garreaud et al 2009) In winter precipitation is modulated by the north-west
displacement of the SPA the northward shift of the westerlies wind belt and an
increased frequency of storm fronts stemming off the Southern Hemisphere
Westerly Winds (SWW) (Frugone-Aacutelvarez et al 2017) Austral summers are
typically dry and warm as a strong SPA blocks the northward migration of storm
tracks stemming off the SWW
36
Historic land cover changes started after the Spanish conquest with a Jesuit
settlement in 1627 CE near El Convento village and the development of a livestock
ranch that included the Matanzas watershed After the Jesuits were expelled from
South America in 1778 CE the farm was bought by Pedro Balmaceda and had more
than 40000 head of cattle around 1800 CE (Contreras-Lopez et al 2014) The first
Pinus radiata and Eucalyptus globulus trees were planted during the second half of
the 19th century and were mostly used for dune stabilization (Albert 1900 Gibson
1972) However the main plantation phase occurred 60 years ago (Villa-Martinez
2002) as a response to the application of Chilean Forestry Laws promulgated in
1931 and 1974 and associated state subsidies
Major land cover changes occurred recently from 1975 to 2008 as shrublands
were replaced by more intensive land uses practices such as farmland and tree
plantations (Schulz et al 2010) Laguna Matanzas is part of the Reserva Nacional
Humedal El Yali protected area (Ramsar Ndeg 878) Despite this protected status the
lake and its watershed have been heavily affected by intense agricultural and
farming activities during the last decades The main inlet ldquoLas Rosasrdquo has been
diverted for crop irrigation causing a significant loss of water input to the lake
Consequently the flooded area of the lake has greatly decreased in the last couple
of decades (Fig 1b) Exotic tree species cover a large surface area of the
watershed Recently other activities such as farms for intensive chicken production
have been emplaced in the watershed
37
Figure 1 Laguna Matanzas a) Digital Elevation Model showing the watershed and
the surface hydrologic connection with Colejuda and Cabildo lakes b) Climograph
depicting the warm dry season in austral summer c) Annual precipitation from
1965-2015 (arrow shows start of the most recent ldquomega-droughtrdquo- see Garreaud et
al 2017) d) A decline of lake area occurs between 2007 and 2018 Lake surface
area decreased first along the western sector (in 2007) followed by more inland
areas (in 2018)
38
3 RESULTS
31 Age Model
The age model for the Matanzas sequence was developed using Bacon software
to establish the deposition rates and age uncertainty (Blaauw and Christen 2011)
It is based on 210Pb and 137Cs dates and two 14C dates (Table 2) According to this
age model the lake sequence spans the last 1000 years (Fig 2) A major breccia
layer (unit 3b) was deposited during the early 18th century which agrees with
historic documents indicating that a tsunami impacted Laguna Matanzas and its
watershed in 1730 CE (Contreras-Loacutepez et al 2014) Here we focus in the last 200
years were the most important changes occurred in terms of LUCC (after the
sedimentary hiatus caused by the tsunami) The Spanish colonial period (17th -18th
century) brought new forms of territorial management along with an intensification
of watershed use which remained relatively unchanged until the 1900s
39
Figure 2 a) Age-depth model obtained for the Laguna de Matanzas sedimentary
sequence A hiatus occurs at 80 cm depth (unit 3b) The section of core used for our
analysis is highlighted in a red rectangle b) Close up of the age model used for
analysis of recent anthropogenic influences on the N cycle c) Information regarding
the 14C dates used to construct age model
Lab code Sample ID
Depth (cm) Material Fraction of modern C
Radiocarbon age
Pmc Error BP Error
D-AMS 021579
MAT11-6A 104-105 Bulk Sediment
8843 041 988 37
D-AMS 001132
MAT11-6A 1345-1355
Bulk Sediment
8482 024 1268 21
POZ-57285
MAT13-12 DIC Water column 10454 035 Modern
Table 2 Laguna Matanzas radiocarbon dates
32 The sediment sequence
Laguna Matanzas sediments consist of massive to banded mud with some silt
intercalations They are composed of silicate minerals (plagioclase quartz and clay
minerals) with relatively high TOC content (Fig 3) Pyrite is a common mineral
indicating dominant anoxic conditions in the lake sediments whereas aragonite
occurs only in the uppermost section Mineralogical analyses visual descriptions
texture and geochemical composition were used to characterize five main facies
(Fig 3) F1 (organic-rich mud) represents baseline sedimentation in a shallow well-
mixed brackish highly productive lake and F1acute (dark orange) is a less organic facies
than F1 (more details see table in the supplementary material) F2 (massive to
banded silty mud) indicates periods of higher clastic input into the lake but finer
(mostly clay minerals) likely from suspension deposition associated with flooding
40
events Aragonite (up to 15 ) occurs in both facies but only in samples from the
uppermost 30 cm and it is interpreted as endogenic linked to higher MgFe waters
and elevated biologic productivity
Figure 3 Sedimentary facies and units mineralogy grain size elemental geochemical
and isotopic composition of Laguna Matanzas core Values highlighted in gray indicate
that these are above average
The banded to laminated fining upward silty clay layers (F3) reflect
deposition by high energy turbidity currents The presence of aragonite suggests
that littoral sediments were incorporated by these currents Non-graded laminated
coarse silt layers (F4) do not have aragonite indicating a dominant watershed
41
sediment source Both facies are interpreted as more energetic flood deposits but
with different sediment sources A unique breccia layer with coarse silt matrix and
cm-long soft clasts (F5) is interpreted as a high-energy event (ie a tsunami)
capable of eroding the littoral zone and depositing coarse clastic material in the
distal zone of the lake Similar coarse breccia layers have been found at several
coastal sites in Chile and interpreted as tsunami-related deposits (Vargas et al
2005 Le Roux et al 2008)
33 Sedimentary units
Three main units and six subunits have been defined (Fig 3) based on
sedimentary facies and sediment composition We use ZrTi as an indicator of the
mineral fraction transported from the watershed (Marzecoacuteva et al 2011) as higher
ZrTi (F3 and F4) is commonly associated with coarser sediments (Cuven et al
2010) A high correlation among Br BrTi and TOC (r = 046-087 p value=0)
supports the use of BrTi as an indicator of lake productivity (Marzecoacuteva et al 2011
Frugone-Aacutelvarez et al 2017) The MnFe ratio is indicative of lake bottom
oxygenation (Naecher et al 2013) as under reducing conditions Mn mobilizes more
than Fe leading to a decreased MnFe ratio (Marzecoacuteva et al 2011) SrTi indicates
periods of increased aragonite formation as Sr is preferentially included in the
aragonite mineral structure (Veizer et al 1971) (See supplementary material)
The basal unit (3c 129 to 99 cm) is relatively organic-rich (TOC mean=26
BioSi mean=5) and composed by F1 (without aragonite) with some coarser F4
flood layers (ZrTi mean=025) Unit 3b (98 to 80 cm) is interpreted as a tsunami or
storm surge deposit (breccia F5 grading into massive to banded silt F2) Unit 3a
(79 to 73 cm) is characterized by relatively low productivity (TOC=2 BrTi=002
42
BioSi=4) under anoxic conditions (MnFe=001) Unit 2a (72 to 31cm) has
relatively less organic content and more intercalated clastic facies F3 and F4 The
top of this unit (43-30 cm) has elevated TS values The Subunit 1b (30-20 cm)
shows increasing TOC BioSi and BrTi values (TOC mean = 29 BioSi mean =
54 BrTi mean=004) and the upper subunit 1a (19-0 cm) has the highest TOC
(mean = 64) and BioSi (mean = 56) high BrTi mean = 010 and the presence
of aragonite More frequent anoxic conditions (MnFe lower than 001) during units
3 and 2 shifted towards more oxic episodes (MnFe = 003) in unit 1 (Fig 3)
34 Isotopic signatures
Figure 4 shows the isotopic signature from soil samples of the major land
usescover present in the Laguna Matanzas used as an end member in comparison
with the lacustrine sedimentary units δ15N from cropland samples exhibit the
highest values whereas grassland and soil samples from lake shore areas have
intermediate values (Fig 4) Tree plantations and native forests have similarly low
δ15N values (+11 permil SD=24) All samples (except those from the lake shore)
exhibit low δ13C values (from ndash285permil to ndash298permil) CNmolar from agriculture land
lakeshore area and non-vegetation areas samples display the lowest values (about
18) CNmolar from tree plantations and native forest have the highest values (383
and 267 respectively)
43
Figure 4 C-N stable isotope plot showing a comparison of lake sediments grouped
by sedimentary units (MAT11-6A) with the soil end members of present-day (lake
shore and land usecover) from Laguna Matanzas
The δ15N values from sediment samples (MAT11-6A) range from ndash15 and
+53permil (mean= +35 SD=05) δ13C values range from ndash266permil to ndash202permil (mean=
ndash240 SD=14) In unit 3c δ15N and δ13C show relatively high values (mean=
+41permil and ndash233permil respectively) δ15N and δ13C from unit 3b and 3a fluctuate at
slightly lower values than in 3c (mean δ15N from 3a= +38permil and 3b mean= +39permil
mean δ13C from 3a= ndash242permil and 3b mean= ndash244permil) In unit 2a δ15N values are
relatively high (mean= +38permil) but show a slightly decreasing trend (from +52 to
+24permil at 66 cm and 36 cm respectively) δ13C also decreases (mean= ndash242permil)
reaching minimum values at 45 cm (ndash266permil) with a sharp increase towards the top
of this unit with maximum values ca ndash210permil Unit 1 exhibits the lowest δ15N values
(1a mean= +11permil 1b mean= +28permil) with negative values in the uppermost
44
sediments (ndash04permil at 14 cm) δ13C values show a decreasing trend over most of
subunit 1b and increase only near the very top of this unit
35 Recent land use changes in the Laguna Matanzas watershed
Major LUCC from 1975 (unit 1b) to 2016 CE in the Laguna Matanzasacutes
watershed is summarized in Figure 5 The watershed has a surface area of 30 km2
of which native forest (36) and grassland areas (44) represented 80 of the
total surface in 1975 The area occupied by agriculture was only 02 and tree
plantations were absent Isolated burned areas (33) were located mostly in the
northern part of the watershed By 1989 tree plantations surface area had increased
to 5 burned areas to 17 and agricultural fields to 9 (a 45-fold increase) and
native forest and grassland sectors decreased to 23 and 27 respectively By
2016 agricultural land and tree plantations have increased to 17 of the total area
whereas native forests decreased to 21
45
Figure 5 Land uses and cover changes from 1975 to 2016 in Laguna Matanzas
watershed from natural cover and areas for livestock grazing (grassland) to the
expansion of agriculture and forest plantation
4 DISCUSSION
41 N and C dynamics in Laguna Matanzas
Small lakes with relatively large watersheds such as Laguna Matanzas would
be expected to have relatively high contributions of allochthonous C to the sediment
OM (Gu et al 2006) Terrestrial C3 plants (δ13C =ndash26 to ndash28permil Meyers and Terranes
2001 Ku et al 2007) are dominant in the Laguna Matanzas watershed Likewise
our soil samples ranged across similar although slightly more negative values
46
(δ13C = ndash30 to ndash28permil Fig 4) to those proposed by Meyers and Terranes (2001)
and are used here as terrestrial end members oil samples were taken from the lake
shore (δ13C = ndash22permil) and POM from surface water (δ13C = ndash24permil) were more
positive than the terrestrial end member and are used as lacustrine end members
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from terrestrial vegetation and more positive δ13C values have increased
aquatic OM (Gu et al 2006 Torres et al 2012) Phytoplankton preferentially uptake
12C leaving the DIC pool enriched in 13C (Gu et al 2006) especially when there are
no important external sources of C (eg decreased C input from the watershed)
Therefore during events of elevated primary productivity the phytoplankton uptakes
12C until its depletion and are then obligated to use the heavier isotope resulting in
an increase in δ13C Changes in lake productivity thus greatly affect the C isotope
signal (Torres et al 2012) with high productivity leading to elevated δ13C values
(Torres et al 2012 Gu et al 2006)
In a similar fashion the N isotope signatures in Laguna Matanzas reflect a
combination of factors including different N sources (autochthonousallochthonous)
and lake processes such as productivity isotope fractionation in the water column
and sediment denitrification Elevated δ15N values from a POM sample (+22permil) and
average values from the lake shore (mean=+34permil SD=028) are used as aquatic
end members whereas terrestrial samples have values from +10 +24 (tree species)
to +77permil +35 (agriculture) which we use as terrestrial end members (Fig 4)
Autochthonous OM in aquatic ecosystems typically displays low δ15N values
when the OM comes from N-fixing species Atmospheric fixation of N2 by
cyanobacteria ends in OM with δ15N values close to 0permil (Leng et al 2006)
Phytoplankton preferentially uptake 14N from Dissolved Inorganic Nitrogen (DIN) in
47
the water column and derived OM typically have δ15N values lower than DIN values
When productivity increases the remaining DIN becomes depleted in 14N which in
turn increases the δ15N values of phytoplankton over time especially if the N not
replenished (Torres et al 2012) Thus high POM δ15N values from Laguna
Matanzas reflect elevated phytoplankton productivity with a 14N- depleted DIN In
addition N-watershed inputs also contribute to high δ15N values Heavily impacted
watersheds by human activities are often reflected in isotope values due to land use
changes and associated modified N fluxes For example the input of N runoff
derived from the use of inorganic fertilizers leads to the presence of elevated δ15N
(between ndash4 to +4permil) values in the water bodies (Elliott and Brush 2006 Diebel and
Vander Zanden 2009) Widory et al (2004) reported a direct relationship between
elevated δ15N values and increased nitrate concentration from manure in the
groundwater of Brittany (France) Terranes and Bernasconi (2000) found a good
correlation between augmented nutrient loading and a progressive increase in δ15N
values of sedimentary OM related to agricultural land use
Post-depositional diagenetic processes can further affect C and N isotope
signatures Several studies have shown a decrease in δ13C values of OM in anoxic
environments particularly during the first years of burial related to the selective
preservation of OM depleted in 13C (Hollander and Smith 2001 Lehmann et al
2002 Gӓlman et al 2009 Torres et al 2012) Diagenetic processes can also lead
to post-burial N isotope enrichment of the sediments Indeed 14N is consumed more
rapidly than 15N by denitrifying bacteria which intensifies under anoxic conditions
(Diebel and Vander Zanden 2009) Thus the remaining OM pool will be enriched
in 15N leaving to elevated δ15N values (Granger et al 2008 Menzel et al 2013)
48
In summary the relatively high δ15N values in sediments of Laguna Matanzas
reflect N input from an agriculturegrassland watershed with positive synergetic
effects from increased lake productivity enrichment of DIN in the water column and
most likely denitrification The increase of algal productivity associated with
increased N terrestrial input andor recycling of lake nutrients (and lesser extent
fixing atmospheric N) and denitrification under anoxic conditions can all increase
δ15N values (Fig 3) In addition elevated lake productivity without C replenishing
(eg by terrestrial C input) produces shifts towards positive δ13C values whereas C
input from the watershed generates more negative δ13C values
42 Recent evolution of the Laguna Matanzas watershed
Sedimentological compositional and geochemical indicators show three
depositional phases in the lake evolution under the human influence in the Laguna
Matanzas over the last two hundred years Although the record is longer (around
1000 years) we analyzed the last two centuries (unit 2 and 1) to provide a recent
historical context for the large changes detected during the 20th century
The first phase lasted from the beginning of the 19th century until ca 1940
(unit 2a Fig 3 4) and was characterized by moderate productivity with elevated
sediment input from the watershed as indicated by our geochemical proxies (BrTi
= 004 AlTi gt 01 Fig 6) Higher lake levels and dominant anoxic bottom conditions
(MnFe = 001 Fig3) can be related to increased precipitation (mean= 557 mmyr)
and lower temperatures (summer annual temperature lt19ordmC) During the Spanish
colonial period the Laguna Matanzas watershed was used as a livestock farm
(Jesuits 1627-1780 unit 3 followed by the Pedro Balmaceda era 1780-1810- unit
2a) (Contreras-Loacutepez et al 2014) The main buildings and farm facilities at the El
49
Convento village During this period livestock grazing and lumber extraction for
mining would have involved extensive deforestation and loss of native vegetation
(eg Armesto et al 1994 2010) However the Matanzas pollen record does not
show any significant regional deforestation during this period (Villa Martiacutenez 2002)
suggesting that the impact may have been highly localized
Lake productivity sediment input and elevated precipitation (Fig 6) all
suggest that N availability was related to this increased input from the watershed
The N from cow manure and soil particles would have led to higher δ15N values
(Elliot and Brush 2016) In addition anoxic lake bottom conditions would lead to
even further enrichment of buried sediment N The δ13C values lend further support
to our interpretation of increased sediment input -and N- from the watershed
Decreased δ13C values reach their lowest values in the entire record (ndash270permil) at
ca 1910 CE (Fig 4 6)
During most of the 19th century human activities in Laguna Matanzas were
similar to those during the Spanish Colonial period However the appearance of
Pinus radiata and Eucalyptus globulus pollen ca 1890 CE (Gibson 1972) the dune
stabilization-afforestation program which began ca 1900 CE (Albert 1900) and the
application of the Chilean forestry law of 1931 (DFL nordm265) contributed to an
increased capacity of the surrounding vegetation to retain nutrients and sediments
The law subsidized forest plantations in areas devoid of vegetation and prohibited
the cutting of forest on slopes greater than 45ordm These land use changes were coeval
with decreased sediment inputs (AlTi trend) from the watershed slightly increased
lake productivity (BrTi from 001 to 003) and a decrease in annual precipitation
(Fig 6) N isotope values become more negative during this period although they
remained high (from +49permil to +37permil) whereas the δ13C trend towards more
50
positive values reflects changes in the N source from watershed to in-lake dynamics
(e g increased endogenic productivity)
The second phase started after 1940 and is clearly marked by an abrupt
change in the general trend of δ15N that oscillates around +41permil (SD = 04permil) during
the first phase to +24permil (SD = 14permil) during the second These δ15N trends reflect
the lowest watershed nutrient and sediment inputs (based on the AlTi record)
decreased precipitation (mean = 318 mm year) and a slight increase in lake
productivity (increased BrTI) Depositional dynamics in the lake likely crossed a
threshold as human activity intensified throughout the watershed and lake levels
decreased
During the Great Acceleration δ15N values shifted towards higher values to
ca 3permil with an increase in δ13C values that are not reflected either in lake
productivity or lake level As the sediment input from the watershed increased and
precipitation remained as low as the previous decade δ15N values during this period
are likely related to watershed clearance which would have increased both nutrient
and sediment input into the lake
The δ13C trend to more positive values reaching the peaks in the 1960s (ndash
212permil) at the same time as the peaks in temperature (196 ordmC mean annual) with a
downward trend in precipitation A shift in OM origin from macrophytes and
watershed input influences to increased lake productivity could explain this trend
(Fig 4 1b)
In the 1970s the Laguna Matanzasacute watershed was mostly covered by native
forest (36) and grassland areas were intended for livestock grazing (44 Fig 5)
Both soils have δ15NSoil lt 5permil (Fig 4) Farming fields occupied a very small area and
tree plantations were almost nonexistent The decreasing trend in δ15N values seen
51
in our record is interrupted by several large peaks that occurred between ca 1975
and ca 1989 when the native forest and grassland areas fell by 23 and 27
respectively largely due to fires affecting 17 of the forests Agriculture fields
increased by 4 and forest plantations increased by 9 (unit 1a) Concomitantly
sediment ndash and likely N - inputs from the watershed decreased (as indicated by the
trend in AlTi) the precipitation was still relatively low (Fig 6) These changes are
likely related to the increase of vegetation cover especially of tree plantations (which
have more negative δ15N values) The small increase in productivity in the lake could
have been favored by increased temperature (von Gunten et al 2009) After 1989
the increase in agriculture land (17 in 2016) correlates with increasing δ15N δ13C
and TOC trends in spite of declining rainfall The increase of forest plantations was
mostly in response to the implementation of the Law Decree of Forestry
Development (DL 701 of 1974) that subsidized forest plantation After 1989 the
increase in agricultural land (17 in 2016) is synchronous with increasing δ15N
δ13C TOC and MnFe trends despite the decline in rainfall and overall lower lake
levels as more water is used for irrigation
The third phase started c 1990 CE (unit 1a) when OM accumulation rates
increase and δ13C δ15N decreased reaching their lowest values in the sequence
around 2000 CE Afterward during the 21st century δ13C and δ15N values again
began to increase The onset of unit 1 is marked by increased lake productivity and
decreased sediment input (AlTi lt 02) synchronous with intensive farming replacing
forestry and extensive agriculture (Fig 5 6)
A change in the general trend of δ15N values which decreased until 1990
(unit 1a) as the native forest and grassland area fell to 23 and 27 respectively
is most likely due to deforestation and fires Agriculture surface increased to 4 and
52
forest plantations increased by 9 (Fig 5) Concomitantly sediment ndash and likely N
ndash inputs from the watershed decreased probably related to the low precipitation (Fig
1b) and the increase of vegetation cover in the watershed in particularly by tree
plantations (with more negative δ15N Fig 4)
At present agriculture and tree plantations occupy around 34 of the
watershed surface whereas native forests and grassland cover 21 and 25
respectively Lake productivity as indicated by BrTi (Fig6) is higher and generates
OM with lower δ15N and higher δ13C values (about +2permil and ndash20permil ca 2000 CE
respectively) due to in-lake processes (ie biological N fixation and nutrient
recycling) and driven by changes in the arboreal cover which diminishes nutrient
flux into the lake (Fig6)
53
Figure 6 Anthropogenic and climatic forcing and lake dynamics response
(productivity sediment input N and C cycles) at Matanzas Lake over the last two
54
centuries Mean annual precipitation reconstructed and temperatures (von Gunten
et al 2009) Vertical gray bars indicate mega-droughts
5 CONCLUSIONS
Human activities have been the main factor controlling the N and C cycle in
the Laguna Matanzas during the last two centuries The N isotope signature in the
lake sediments reflects changes in the watershed fluxes to the lake but also in-lake
processes such as productivity and post-depositional changes Denitrification could
have been a dominant process during periods of increased anoxic conditions which
were more frequent prior to Great Acceleration (1950 CE) Higher δ15N and lower
δ13C values are associated with increased nutrient input from the watershed due to
increased livestock grazing and agriculture pressure (phase 1 Fig 7a) whereas
lower isotope values occurred during periods of increased forest plantations (phase
3 Fig 7c) During periods of increased lake productivity - such as in the last few
decades - δ15N values increased significantly
The most important change in C and N dynamics in the lake occurred after the
1950s (Phase 2 and 3) as tree plantations dominated the watershed Recent
changes in N dynamics can be explained by the higher nutrient contribution
associated with intensive agriculture (i e fertilizers) since the 1990s Although the
replacement of livestock activities with forestry and farming seems to have reduced
nutrient and soil export from the watershed to the lake the inefficient use of fertilizer
(by agriculture) can be the ultimate responsible for lake productivity increase during
the last decades
55
Figure 7 Schematic diagrams illustrating the main factors controlling the
isotope N signal in sediment OM of Laguna Matanzas N input from watershed
depends on human activities and land cover type Agriculture practices and cattle
(grassland development) contribute more N to the lake than native forest and
plantations Periods of higher productivity tend to deplete the dissolved inorganic N
in 14N resulting in higher δ15N (OM) The denitrification processes are more effective
in anoxic conditions associated with higher lake levels
6 METHODS
Short sediment cores were recovered from Laguna Matanzas using an Uwitec
gravity piston corer in 2011 and 2013 (Fig 1a) Sediment cores (MAT11-6A 129 cm
MAT13-2A 149 cm MAT13-3A 33 cm MAT13-4A 99 cm) were split
photographed sub-sampled and stored at the Pyrenean Institute of Ecology (IPE-
CSIC Spain) Core MAT11-6A was obtained from the central sector of the lake and
56
was selected for detailed multiproxy analyses (including elemental geochemistry C
and N isotope analyses XRF and 14C dating)
The isotope analyses (δ13C and δ15N) were performed at the Laboratory of
Biogeochemistry and Applied Stable Isotopes (LABASI-PUC Chile) using a Delta
V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via a
Conflo IV interface Isotope results are expressed in standard delta notation (δ) in
per mil (permil) relative to the standards Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Sediment samples
for δ13Corg were pre-treated with 50 ml of HCl and 50 ml of deionized water and
dried at 60ordmC for 4 hr to remove carbonates (Harris et al 2011)
Total Carbon (TC) Total Organic Carbon (TOC) Total Inorganic Carbon (TIC)
and Total Sulphur (TS) were measured every cm with a LECO SC 144 DR at IPE-
CSIC XRF measurements were carried out every 4 mm in MAT11-1A-1G core using
an AVAATECH X-ray Fluorescence II core scanner at the University of Barcelona
(Spain) Results are expressed as element intensities in counts per second (cps)
Tube voltage was operated at 30 kV and 10 kV to obtain the abundances of 15
elements (Al Si S Cl K Ca Ti V Mn Fe Br Rb Sr Y Zr) with an average at
least of 1600 cps (less for Br=1000)
Biogenic silica content mineralogy and grain size were measured every 4
cm Biogenic silica was measured following Mortlock and Froelich (1989) and
Bernaacuterdez et al (2005) using an Auto Analyzer Technicon AAII for dissolved silicate
analysis Mineralogy was analyzed with a Siemens D-500 X-ray diffractometer (Cu
kα 40 kV 30 mA graphite monochromator) at the ICTJA-CSIC (Spain) Grain size
analyses were performed in a Beckmann Coulter LS 13 320 Particle Size Analyzer
57
at the IPE-CSIC The samples were classified according to textural classes as
follows clay (lt2 μm) silt (20-2 microm) and sand (gt2 microm) fractions
The age-depth model for the Laguna Matanzas sedimentary sequence was
constructed using 210Pb and 137Cs dating techniques (MAT13-4A) as well as two 14C
AMS dates on bulk sediment samples (MAT11-6A fig 2) We dated the dissolved
inorganic carbon (DIC) in the water column and no significant reservoir effect is
present in the modern-day water column (10454 + 035 pcmc Table 2) An age-
depth model was obtained with the Bacon R package to estimate the deposition
rates and associated age uncertainties along the core (Blaauw and Christen 2011)
To estimate LUCC of Laguna Matanzasacutes watershed we use satellite images
Landsat MSS for 1975 Landsat TM for 1989 and Landsat OLI for 2016 all taken in
summer or autumn (Table 1) We performed supervised classification of land uses
(maximum likelihood algorithm) for each year (1975 1989 and 2016) and the results
were mapped using software ArcGIS 102 in 2017
Satellite Images Acquisition Date
Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat OLI 20160404 30 m
Table 1 Landsat imagery
Surface water samples were filtered for obtained particulate organic matter In
addition soil samples from the main land usecover present in the Laguna Matanzas
watershed were collected Elemental C N and their corresponding isotopes from
POM and soil were obtained at the LABASI and used here as end members
Daily precipitation at Santo Domingo (33ordm39`S 71ordm3 6`W)- the nearest weather
station to Laguna Matanzasndash was compiled using the redPrec R package (Fig1 d
Serrano-Notivoli et al 2017 Sarricolea et al 2019) To extend our precipitation
58
reconstruction back to 1824 we correlated this dataset with that available for
Santiago The Santiago data was compiled from data published in the Anales of
Universidad of Chile (1850) for the years 1824 to 1849 Almeyda (1940) for the years
1849 to 1864 and the Quinta Normal series from 1866 to the present (Direccioacuten
Meteoroloacutegica de Chile) We generated a linear regression model between the
presentday Santo Domingo station and the compiled Santiago data with a Pearson
coefficient of 087 and p-valuelt 001
Acknowledgments This research was funded by grants CONICYT AFB170008
to the Institute of Ecology and Biodiversity (IEB) and FONDECYT 1150763 (to CL)
Doctoral grant Becas Chile 21150224 MEDLANT (Spanish Ministry of Economy
and Competitiveness grant CGL2016-76215-R) Additional funding was provided
by the Laboratorio Internacional de Cambio Global (LINCGlobal PUC-CSIC) We
thank R Lopez E Royo and M Gallegos for help with sample analyses We thank
the Laboratory of Biogeochemistry and Applied Stable Isotopes (LABASI) of the
Department of Ecology (PUC) for sample analyses
References
Albert F 1900 Las dunas o sean arenas volantes voladeros arenas muertas invasion de las arenas playas I meacutedanos del centro de Chile comprendiendo el litoral desde el liacutemite norte de la provincia de Aconcagua hasta el liacutemite sur de la de Arauco Memorias cientiacuteficas I Lit
Anderson NJ W D Fritz SC 2009 Holocene carbon burial by lakes in SW
Greenland Glob Chang Biol httpsdoiorg101111j1365-2486200901942x
Armesto J Villagraacuten C Donoso C 1994 La historia del bosque templado
chileno Ambient y Desarro 66 66ndash72 httpsdoiorg101007BF00385244
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A
59
historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Arnell NW Gosling SN 2013 The impacts of climate change on river flow
regimes at the global scale J Hydrol 486 351ndash364 httpsdoiorg101016JJHYDROL201302010
Bernaacuterdez P Prego R Franceacutes G Gonzaacutelez-Aacutelvarez R 2005 Opal content in
the Riacutea de Vigo and Galician continental shelf Biogenic silica in the muddy fraction as an accurate paleoproductivity proxy Cont Shelf Res httpsdoiorg101016jcsr200412009
Blaauw M Christen JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474 httpsdoiorg10121411-BA618
Brush GS 2009 Historical land use nitrogen and coastal eutrophication A
paleoecological perspective Estuaries and Coasts 32 18ndash28 httpsdoiorg101007s12237-008-9106-z
Camarero L Catalan J 2012 Atmospheric phosphorus deposition may cause
lakes to revert from phosphorus limitation back to nitrogen limitation Nat Commun httpsdoiorg101038ncomms2125
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego
R Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and environmental change from a high Andean lake Laguna del Maule central Chile (36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Catalan J Pla-Rabeacutes S Wolfe AP Smol JP Ruumlhland KM Anderson NJ
Kopaacuteček J Stuchliacutek E Schmidt R Koinig KA Camarero L Flower RJ Heiri O Kamenik C Korhola A Leavitt PR Psenner R Renberg I 2013 Global change revealed by palaeolimnological records from remote lakes A review J Paleolimnol httpsdoiorg101007s10933-013-9681-2
Chen Y Y Huang W Wang W H Juang J Y Hong J S Kato T amp
Luyssaert S (2019) Reconstructing Taiwanrsquos land cover changes between 1904 and 2015 from historical maps and satellite images Scientific Reports 9(1) 3643 httpsdoiorg101038s41598-019-40063-1
Contreras S Werne J P Araneda A Urrutia R amp Conejero C A (2018)
Organic matter geochemical signatures (TOC TN CN ratio δ 13 C and δ 15 N) of surface sediment from lakes distributed along a climatological gradient on the western side of the southern Andes Science of The Total Environment 630 878-888 httpsdoiorg101016jscitotenv201802225
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia UC
60
de V 2014 ELEMENTOS DE LA HISTORIA NATURAL DEL An Mus Hist Natulas Vaplaraiso 27 51ndash67
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-010-9453-1
Diebel M Vander Zanden MJ 201209 Nitrogen stable isotopes in streams
stable isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19 1127ndash1134 httpsdoiorg10189008-03271
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506 Espizua LE Pitte P 2009 The Little Ice Age glacier advance in the Central
Andes (35degS) Argentina Palaeogeogr Palaeoclimatol Palaeoecol 281 345ndash350 httpsdoiorg101016JPALAEO200810032
Figueroa JA Castro S A Marquet P A Jaksic FM 2004 Exotic plant
invasions to the mediterranean region of Chile causes history and impacts Invasioacuten de plantas exoacuteticas en la regioacuten mediterraacutenea de Chile causas historia e impactos Rev Chil Hist Nat 465ndash483 httpsdoiorg104067S0716-078X2004000300006
Fowler D Coyle M Skiba U Sutton MA Cape JN Reis S Sheppard
LJ Jenkins A Grizzetti B Galloway N Vitousek P Leach A Bouwman AF Butterbach-bahl K Dentener F Stevenson D Amann M Voss M 2013 The global nitrogen cycle in the twenty- first century Philisophical Trans R Soc B httpsdoiorghttpdxdoiorg101098rstb20130164
Frank D Reichstein M Bahn M Thonicke K Frank D Mahecha MD
Smith P van der Velde M Vicca S Babst F Beer C Buchmann N Canadell JG Ciais P Cramer W Ibrom A Miglietta F Poulter B Rammig A Seneviratne SI Walz A Wattenbach M Zavala MA Zscheischler J 2015 Effects of climate extremes on the terrestrial carbon cycle Concepts processes and potential future impacts Glob Chang Biol httpsdoiorg101111gcb12916
Fritz SC Anderson NJ 2013 The relative influences of climate and catchment
processes on Holocene lake development in glaciated regions J Paleolimnol httpsdoiorg101007s10933-013-9684-z
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
61
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS) implications for past sea level and environmental variability J Quat Sci 32 830ndash844 httpsdoiorg101002jqs2936
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892 httpsdoiorg101126science1136674
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924 httpsdoiorg104319lo20095430917
Garciacutea-Ruiz JM 2010 The effects of land uses on soil erosion in Spain A
review Catena httpsdoiorg101016jcatena201001001
Garciacutea-Ruiz JM Loacutepez-Moreno JI Lasanta T Vicente-Serrano SM
Gonzaacutelez-Sampeacuteriz P Valero-Garceacutes BL Sanjuaacuten Y Begueriacutea S Nadal-Romero E Lana-Renault N Goacutemez-Villar A 2015 Los efectos geoecoloacutegicos del cambio global en el Pirineo Central espantildeol una revisioacuten a distintas escalas espaciales y temporales Pirineos httpsdoiorg103989Pirineos2015170005
Garciacutea-Ruiz JM Nadal-Romero E Lana-Renault N Begueriacutea S 2013
Erosion in Mediterranean landscapes Changes and future challenges Geomorphology 198 20ndash36 httpsdoiorg101016jgeomorph201305023
Garreaud RD Vuille M Compagnucci R Marengo J 2009 Present-day
South American climate Palaeogeogr Palaeoclimatol Palaeoecol httpsdoiorg101016jpalaeo200710032
Gayo EM Latorre C Jordan TE Nester PL Estay SA Ojeda KF
Santoro CM 2012 Late Quaternary hydrological and ecological changes in the hyperarid core of the northern Atacama Desert (~21degS) Earth-Science Rev httpsdoiorg101016jearscirev201204003
Ge Y Zhang K amp Yang X (2019) A 110-year pollen record of land use and land
cover changes in an anthropogenic watershed landscape eastern China Understanding past human-environment interactions Science of The Total Environment 650 2906-2918 httpsdoiorg101016jscitotenv201810058
Goyette J Bennett EM Howarth RW Maranger R 2016 Global
Biogeochemical Cycles 1000ndash1014 httpsdoiorg1010022016GB005384Received
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and
oxygen isotope fractionation during dissimilatory nitrate reduction by
62
denitrifying bacteria 53 2533ndash2545 httpsdoiorg10230740058342
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
Gurnell AM Petts GE Hannah DM Smith BPG Edwards PJ Kollmann
J Ward J V Tockner K 2001 Effects of historical land use on sediment yield from a lacustrine watershed in central Chile Earth Surf Process Landforms 26 63ndash76 httpsdoiorg1010021096-9837(200101)261lt63AID-ESP157gt30CO2-J
Heathcote A J et al Large increases in carbon burial in northern lakes during the
Anthropocene Nat Commun 610016 doi 101038ncomms10016 (2015) Hollander DJ Smith MA 2001 Microbially mediated carbon cycling as a
control on the δ13C of sedimentary carbon in eutrophic Lake Mendota (USA) New models for interpreting isotopic excursions in the sedimentary record Geochim Cosmochim Acta httpsdoiorg101016S0016-7037(00)00506-8
Holtgrieve GW Schindler DE Hobbs WO Leavitt PR Ward EJ Bunting
L Chen G Finney BP Gregory-Eaves I Holmgren S Lisac MJ Lisi PJ Nydick K Rogers LA Saros JE Selbie DT Shapley MD Walsh PB Wolfe AP 2011 A coherent signature of anthropogenic nitrogen deposition to remote watersheds of the Northern Hemisphere Science (80) 334 1545ndash1548 httpsdoiorg101126science1212267
Hooper DU Adair EC Cardinale BJ Byrnes JEK Hungate BA Matulich
KL Gonzalez A Duffy JE Gamfeldt L Connor MI 2012 A global synthesis reveals biodiversity loss as a major driver of ecosystem change Nature httpsdoiorg101038nature11118
Horvatinčić N Sironić A Barešić J Sondi I Krajcar Bronić I Borković D
2016 Mineralogical organic and isotopic composition as palaeoenvironmental records in the lake sediments of two lakes the Plitvice Lakes Croatia Quat Int httpsdoiorg101016jquaint201701022
Howarth RW 2004 Human acceleration of the nitrogen cycle Drivers
consequences and steps toward solutions Water Sci Technol httpsdoiorg1010382Fscientificamerican0490-56
Jenny B Valero-Garceacutes BL Urrutia R Kelts K Veit H Appleby PG Geyh
M 2002 Moisture changes and fluctuations of the Westerlies in Mediterranean Central Chile during the last 2000 years The Laguna Aculeo record (33deg50primeS) Quat Int 87 3ndash18 httpsdoiorg101016S1040-
63
6182(01)00058-1
Jenny B Wilhelm D Valero-Garceacutes BL 2003 The Southern Westerlies in
Central Chile Holocene precipitation estimates based on a water balance model for Laguna Aculeo (33deg50primeS) Clim Dyn 20 269ndash280 httpsdoiorg101007s00382-002-0267-3
Leng M J Lamb A L Heaton T H Marshall J D Wolfe B B Jones M D
amp Arrowsmith C 2006 Isotopes in lake sediments In Isotopes in palaeoenvironmental research (pp 147-184) Springer Dordrecht
Le Roux JP Nielsen SN Kemnitz H Henriquez Aacute 2008 A Pliocene mega-
tsunami deposit and associated features in the Ranquil Formation southern Chile Sediment Geol 203 164ndash180 httpsdoiorg101016jsedgeo200712002
Lehmann M Bernasconi S Barbieri a McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66 3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Lenihan JM Drapek R Bachelet D Neilson RP 2003 Climate change
effects on vegetation distribution carbon and fire in California Ecol Appl httpsdoiorg101890025295
McLauchlan K K Gerhart L M Battles J J Craine J M Elmore A J
Higuera P E amp Perakis S S (2017) Centennial-scale reductions in nitrogen availability in temperate forests of the United States Scientific reports 7(1) 7856 httpsdoi101038s41598-017-08170-z
Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32ordmS 3050 masl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
Martiacuten-Foreacutes I Casado MA Castro I Ovalle C Pozo A Del Acosta-Gallo
B Saacutenchez-Jardoacuten L Miguel JM De 2012 Flora of the mediterranean basin in the chilean ESPINALES Evidence of colonisation Pastos 42 137ndash160
Marzecovaacute A Mikomaumlgi a Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54 httpsdoiorg103176eco2011105
Matesanz S Valladares F 2014 Ecological and evolutionary responses of
Mediterranean plants to global change Environ Exp Bot httpsdoiorg101016jenvexpbot201309004
64
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Menzel P Gaye B Wiesner MG Prasad S Stebich M Das BK Anoop A
Riedel N Basavaiah N 2013 Influence of bottom water anoxia on nitrogen isotopic ratios and amino acid contributions of recent sediments from small eutrophic Lonar Lake central India Limnol Oceanogr 58 1061ndash1074 httpsdoiorg104319lo20135831061
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Michelutti N Wolfe AP Cooke CA Hobbs WO Vuille M Smol JP 2015
Climate change forces new ecological states in tropical Andean lakes PLoS One httpsdoiorg101371journalpone0115338
Muntildeoz-Rojas M De la Rosa D Zavala LM Jordaacuten A Anaya-Romero M
2011 Changes in land cover and vegetation carbon stocks in Andalusia Southern Spain (1956-2007) Sci Total Environ httpsdoiorg101016jscitotenv201104009
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Poraj-Goacuterska A I Żarczyński M J Ahrens A Enters D Weisbrodt D amp
Tylmann W (2017) Impact of historical land use changes on lacustrine sedimentation recorded in varved sediments of Lake Jaczno northeastern Poland Catena 153 182-193 httpdxdoiorg101016jcatena201702007
Pongratz J Reick CH Raddatz T Claussen M 2010 Biogeophysical versus
biogeochemical climate response to historical anthropogenic land cover change Geophys Res Lett 37 1ndash5 httpsdoiorg1010292010GL043010
Prudhomme C Giuntoli I Robinson EL Clark DB Arnell NW Dankers R
Fekete BM Franssen W Gerten D Gosling SN Hagemann S Hannah DM Kim H Masaki Y Satoh Y Stacke T Wada Y Wisser D 2014 Hydrological droughts in the 21st century hotspots and uncertainties from a global multimodel ensemble experiment Proc Natl Acad Sci httpsdoiorg101073pnas1222473110
65
Ruumlhland KM Paterson AM Smol JP 2015 Lake diatom responses to
warming reviewing the evidence J Paleolimnol httpsdoiorg101007s10933-
015-9837-3
Sarricolea P Meseguer-Ruiz Oacute Serrano-Notivoli R Soto M V amp Martin-Vide
J (2019) Trends of daily precipitation concentration in Central-Southern Chile Atmospheric research 215 85-98 httpsdoiorg101016jatmosres201809005
Sernageomin 2003 Mapa geologico de chile version digital Publ Geol Digit Serrano-Notivoli R de Luis M Begueriumlia S 2017 An R package for daily
precipitation climate series reconstruction Environ Model Softw httpsdoiorg101016jenvsoft201611005
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Stine S 1994 Extreme and persistent drought in California and Patagonia during
mediaeval time Nature 369 546ndash549 httpsdoiorg101038369546a0
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL
Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Syvitski JPM Voumlroumlsmarty CJ Kettner AJ Green P 2005 Impact of humans
on the flux of terrestrial sediment to the global coastal ocean Science 308(5720) 376-380 httpsdoiorg101126science1109454
Thomas RQ Zaehle S Templer PH Goodale CL 2013 Global patterns of
nitrogen limitation Confronting two global biogeochemical models with observations Glob Chang Biol httpsdoiorg101111gcb12281
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Urbina Carrasco MX Gorigoitiacutea Abbott N Cisternas Vega M 2016 Aportes a
la historia siacutesmica de Chile el caso del gran terremoto de 1730 Anuario de Estudios Americanos httpsdoiorg103989aeamer2016211
Vargas G Ortlieb L Chapron E Valdes J Marquardt C 2005 Paleoseismic
inferences from a high-resolution marine sedimentary record in northern Chile
66
(23degS) Tectonophysics httpsdoiorg101016jtecto200412031 399(1-4) 381-398httpsdoiorg101016jtecto200412031
Valero-Garceacutes BL Latorre C Morelliacuten M Corella P Maldonado A Frugone M Moreno A 2010 Recent Climate variability and human impact from lacustrine cores in central Chile 2nd International LOTRED-South America Symposium Reconstructing Climate Variations in South America and the Antarctic Peninsula over the last 2000 years Valdivia p 76 (httpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-yearshttpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-years
Vicente-Serrano SM Gouveia C Camarero JJ Begueria S Trigo R
Lopez-Moreno JI Azorin-Molina C Pasho E Lorenzo-Lacruz J Revuelto J Moran-Tejeda E Sanchez-Lorenzo A 2013 Response of vegetation to drought time-scales across global land biomes Proc Natl Acad Sci httpsdoiorg101073pnas1207068110
Villa Martiacutenez R P 2002 Historia del clima y la vegetacioacuten de Chile Central
durante el Holoceno Una reconstruccioacuten basada en el anaacutelisis de polen sedimentos microalgas y carboacuten PhD
Villa-Martiacutenez R Villagraacuten C Jenny B 2003 Pollen evidence for late -
Holocene climatic variability at laguna de Aculeo Central Chile (lat 34o S) The Holocene 3 361ndash367
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Human alteration of the global nitrogen cycle sources and consequences Ecological applications 7(3) 737-750 httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Rein B Urrutia R amp Appleby P (2009) A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 The Holocene 19(6) 873-881 httpsdoiorg1011770959683609336573
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Widory D Kloppmann W Chery L Bonnin J Rochdi H Guinamant JL
2004 Nitrate in groundwater An isotopic multi-tracer approach J Contam Hydrol 72 165ndash188 httpsdoiorg101016jjconhyd200310010
67
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
68
Supplementary material
Facie Name Description Depositional Environment
F1 Organic-rich
mud
Massive to banded black
organic - rich (TOC up to 14 )
mud with aragonite in dm - thick
layers Slightly banded intervals
contain less OM (TOClt4) and
aragonite than massive
intervals High MnFe (oxic
bottom conditions) High CaTi
BrTi and BioSi (up to 5)
Distal low energy environment
high productivity well oxygenated
and brackish waters and relative
low lake level
F2 Massive to
banded silty clay
to fine silt
cm-thick layers mostly
composed by silicates
(plagioclase quartz cristobalite
up to 65 TOC mean=23)
Some layers have relatively high
pyrite content (up to 25) No
carbonates CaTi BrTi and
BioSi (mean=48) are lower
than F1 higher ZrTi (coarser
grain size)
Deposition during periods of
higher sediment input from the
watershed
69
F3 Banded to
laminated light
brown silty clay
cm-thick layers mostly
composed of clay minerals
quartz and plagioclase (up to
42) low organic matter
(TOC mean=13) low pyrite
and BioSi content
(mean=46) and some
aragonite
Flooding events reworking
coastal deposits
F4 Laminated
coarse silts
Thin massive layers (lt2mm)
dominated by silicates Low
TOC (mean=214 ) BrTi
(mean=002) MnFe (lt02)
TIC (lt034) BioSi
(mean=46) and TS values
(lt064) and high ZrTi
Rapid flooding events
transporting material mostly
from within the watershed
F5 Breccia with
coarse silt
matrix
A 17 cm thick (80-97 cm
depth) layer composed by
irregular mm to cm-long ldquosoft-
clastsrdquo of silty sediment
fragments in a coarse silt
matrix Low CaTi BrTi and
MnFe ratios and BioSi
Rapid high energy flood
events
70
(mean=43) and high ZrTi
(gt018)
Table Sedimentological and compositional characteristics of Laguna Matanzas
facies
71
CAPIacuteTULO 2 STABLE ISOTOPES TRACK LAND USE AND COVER
CHANGES IN A MEDITERRANEAN LAKE IN CENTRAL CHILE OVER THE
LAST 600 YEARS
72
Stable isotopes track land use and cover changes in a mediterranean lake in
central Chile over the last 600 years
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugone-Aacutelvarezabc Pablo
Sarricolead Leonardo Villaciacutese M Laura Carrevedob Blas Valero-Garceacutescg
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Departamento de Ecologiacutea Universidad de ChileLas Palmeras 3425 Nuntildeoa Santiago Chile
f Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding authors E-mail addresses clatorrebiopuccl magdalenafuentealbagmailcom
Key words Anthropocene lake sediment δsup1⁵N δsup1sup3C Great Acceleration organic
geochemistry watershedndashlake system Stable Isotope Analyses land usecover
change Nitrogen cycle mediterranean ecosystems central Chile
73
Abstract
Nutrient fluxes in many aquatic ecosystems are currently being overridden by
anthropic controls especially since the industrial revolution (mid-1800s) and the
Great Acceleration (mid-1900s) Land use and cover changes (LUCC) alter the
availability and fluxes of nutrients such as nitrogen that are transferred via runoff
and groundwater into lakes By altering lake productivity and trophic status these
changes are often preserved in the sedimentary record Here we use
geolimnological proxies and stable isotope analyses (δsup1⁵N δsup13C) on lake sediments
to reconstruct the lake-watershed nutrient transfer in a central Chilean lake (Lago
Vichuqueacuten) over the past 600 years We compare our multiproxy record from recent
lake sediments to the soilvegetation relationship across the watershed as well as
land usecover changes from 1975 to 2014 derived from satellite images Our results
show that lake sediment δsup1⁵N values increased with meadow cover but decreased
with tree plantations suggesting increased nitrogen retention when trees dominate
the watershed Our δsup1⁵N record from central Chile can thus be interpreted as a proxy
for nutrient availability over the last 600 years mainly derived from land use changes
coupled with climate drivers Although variable sources of organic matter and in situ
fractionation often hinder straightforward environmental interpretations of stable N
isotope signatures our study shows that lake sediment δsup1⁵N can be a useful tool for
assessing the contribution of past human activities in nutrient and nitrogen cycling
1 Introduction
Nitrogen (N) is a limiting nutrient in terrestrial and aquatic ecosystems (Vitousek
et al 1997) Changes in its availability can drive eutrophication and increase
pollution in these ecosystems (McLauchlan et al 2013) Although recent human
74
impacts on the global N cycle have been significant the consequences of increased
anthropogenic N on these ecosystems are unclear (Gerhart and Mclauchlan 2014
Battye et al 2017) Isotopic signatures are key to understand N dynamics in lakes
nevertheless in situ andor diagenetic fractionation along with multiple sources of
organic matter (OM) often hinder straightforward environmental interpretations from
isotopes Monitoring δ15N and δ13C values as components of the N cycle
specifically those related to the link between terrestrial and aquatic ecosystems can
help differentiate between effects from processes versus sources in stable isotope
values (eg from Particulate Organic Matter -POM- soil and vegetation) and
improve how we interpret variations in δ15N (and δ13C) values at longer temporal
scales
The main processes controlling stable N isotopes in bulk lake OM are source
lake productivity diagenesis and denitrification (Talbot 2001 Leng et al 2006
Fariacuteas et al 2009 Torres et al 2012 Brahney et al 2014) N sources depend on
contributions from the watershed (ie soil and biomass) the transfer of atmospheric
N to the lake (ie atmospheric N2 fixation) and nutrient recycling (Leng et al 2006)
Atmospheric N2 (isotopically defined as δ15N =0) is fixed by cyanobacteria with
minimal fractionation resulting in δ15NOM values that fluctuate from -2 to +2permil (Fogel
and Cifuentes 1993 Talbot 2001 Elliot and Brush 2006) Besides N2 fixation by
cyanobacteria phytoplankton species assimilate dissolved inorganic nitrogen (DIN)
and values typically range from +7 to +10permil (Xu et al 2016 Meyers et al 2003) In
addition seasonal changes in POM occur in the lake water column Gu et al (2006)
sampled the water column of Lake Wauberg (Florida USA) during a hydrologic year
and found a higher development of N fixing species during the summer A major
factor behind this increase are human activities in the watershed which control the
75
inputs of the sediment and nutrients to the lake (Woodward et al 2011) Some
studies have shown higher δ15N values in lake sediments from watersheds that are
highly affected by human activities (Bruland and Mackenzie 2010 Woodward et al
2011) Syntenic fertilizers displayed δ15N values between -4 to +4permil cattle manure
around +8 to +22permil and surface and groundwater OM between 0 to +10permil (Elliott
and Brush 2006 Leng et al 2006) Although relatively low δ15N values from
fertilizers constitute major N input to human-altered watersheds the elevated loss
of 14N via volatilization of ammonia and denitrification leaves the remaining total N
input enriched in 15N (Bruland and Mackenzie 2010)
In addition to the different sources and variations in lake productivity early
diagenesis at the sedimentndashwater interface in the sediment can further alter
sediment OM isotope values (Brahney et al 2014 Galman et al 2009) During
diagenetic loss of N in lacustrine systems kinetic fractionation occurs and the
remaining N is enriched in 15N leading to higher δ15N values (Galman et al 2006
Talbot 2001) Denitrification tends to enrich lake sediments in 15N due to the
assimilation of nitrates by denitrifying bacteria (Granger et al 2008) and is more
prevalent in anoxic environments (Brahney et al 2016 Lehman et al 2002)
Carbon isotopes in lake sediments can also provide useful information about
paleoenvironmental changes OM origin and depositional processes (Meyers et al
2003) Allochthonous organic sources (high CN ratios) produce isotope values
similar to values from catchment vegetation Autochthonous organic matter (low CN
ratio) is influenced by fractionation both in the lake and the watershed leading up to
carbon assimilation by lacustrine biota (Woodward et al 2011) Changes in
productivity greatly affect δ13COM values (Torres et al 2012) as during C uptake
plankton preferentially assimilate 12C leaving the dissolved inorganic carbon (DIC)
76
pool enriched in 13C (Gu et al 2006) with phytoplankton averaging ca 20 permil lower
than the respective δ13CDIC values (Leng et al 2006) Under conditions of low to
moderate primary productivity plankton preferentially uptake the lighter 12C
resulting in low δ13C values displayed by the OM (Torres et al 2012) Conversely
during high primary productivity phytoplankton will uptake 12C until its depletion and
is then forced to assimilate the heavier isotope resulting in an increase in δ13C
values Higher productivity in C-limited lakes due to slow water-atmosphere
exchange of CO2 also results in high δ13C values (Galman et al 2009) In these
cases algae are forced to uptake dissolved bicarbonate with δ13C values between
7 to 10permil higher than those of the atmospheric CO2 dissolved in water (Sun et al
2016 Torres et al 2012 Galman et al 2009)
Stable isotope analyses from lake sediments are thus useful tools to
reconstruct shifts in lake-watershed dynamics caused by changes in limnological
parameters and LUCC Our knowledge of the current processes that can affect
stable isotope signals in a watershed-lake system is limited however as monitoring
studies are scarce Besides in order to use stable isotope signatures to reconstruct
past environmental changes we require a multiproxy approach to understand the
role of the different variables in controlling these values Hence in this study we
carried out a detailed survey of current N dynamics in a coastal central Chilean lake
(Lago Vichuqueacuten) and a multiproxy study of a sediment record spanning the last
600 years The characterization of the recent changes in the watershed since 1970s
is based on satellite images to compare recent changes in the lake and assess how
these are related with climate variability and an ever increasing human footprint
(Lara et al 2012 Gallardo et al 2015 Steffen et al 2015) Our main goal is to
investigate how stable isotope values from lake sediment reflect changes in the lake
77
ndash watershed system during periods of high watershed disruption (eg Spanish
Conquest late XIX century Great Acceleration) and recent climate change (eg
Little Ice Age and current global warming)
2 Study Site
Lago Vichuqueacuten (34ordm49`S 72ordm04W 4 m asl 30 m of depth Fig 1) is a
mesotrophic to eutrophic lake with dominant anoxic bottom conditions that is
stratified year around (Frugone-Aacutelvarez et al 2017) The main inlets are the
Vichuqueacuten and Huintildee rivers and the only outlet is the Llico estuary that drains into
the Pacific Ocean High tides can sporadically shift the flow direction of the Llico
estuary which increases the marine influence in the lake Dune accretion gradually
limited ocean-lake connectivity until the estuary was almost completely closed off
by ca 10 ky BP and the current lake regime formed (Frugone-Aacutelvarez et al 2017)
The area is characterized by a mediterranean climate with cold-wet winters and
hot-dry summers and an annual precipitation of ~650 mm and a mean annual
temperature of 15ordmC During the austral winter months (June - August) precipitation
is modulated by north-west shifts of the South Pacific Anticyclone (SPA) followed by
an increased frequency of storm fronts stemming off the South Westerly Winds
(SWW) A strengthened SPA during austral summers (December - March) which
are typically dry and warm blocks the northward migration of storm tracks stemming
off the SWW (Fig1 Frugone-Aacutelvarez et al 2017)
78
Figure 1 a) The location of the Lago Vichuqueacuten in central Chile b) Map of land
uses and cover in 2016 c) Ombrothermic diagram mediterranean climates are
characterized by cold-wet winters with surplus moisture from June to August and
hot-dry summers d) Lake bathymetry showing location of cores and water sampling
sites used in this study
Although major land cover changes in the area have occurred since 1975 to the
present as the native forests were replaced by tree (Monterey pine and eucalyptus)
plantations the region was settled before the Spanish conquest (Frugone-Alvarez
et al 2018) Historical documents indicate that Vichuqueacuten was occupied by a
Mitimaes colony as part of the Inka empire (Odone 1998) Unlike other Andean
areas during the Inka period the introduction of corn and quinoa in the Vichuqueacuten
watershed do not seem to have intensified land use The Spanish colonial period in
Chile lasted from 1542 CE to the independence in 1810 CE The first historical
document (1550 CE) shows that the areas around Vichuqueacuten were settled by the
Spanish early on as the Vichuqueacuten watershed was given as an ldquoencomiendardquo
system to Juan Cuevas The ldquoencomiendardquo system provided the owners with land
79
and indigenous people to work but also the introduction of wheat wine cattle
grazing and logging of native forests for lumber extraction and increasing land for
agricultural activities (Odone et al 1998 Armesto et al 2010) During the 19th
century (the Republic) the export of wheat to Australia and Canada generated
intensive changes in land cover use The town of Vichuqueacuten became the regional
capital and plans to build a maritime port in the bay of Vichuqueacuten were drawn
However the fall of international markets in 1880 paralyzed these plans During the
20th century the plantation of trees (Pinus radiata and Eucalyptus globulus) in areas
cleared of native forests was subsidized by two national laws DFL nordm265 (1931) and
DFL nordm 701 (1974) both of which provided funds for such plantations During the last
decades the urbanization with summer vacation homes along the shorelines of
Lago Vichuqueacuten has greatly increased Extensive lake eutrophication is currently a
large environmental problem (EULA 2008)
3 Methods
Coring campaigns were organized from 2011 to 2015 (Fig 1d) We recovered
12 short cores and 4 long cores (up to 14 m long) at several sites using a hammer-
modified UWITEC gravity corer Sediment cores (VIC11-2A 170 cm VIC13-2B 170
cm VIC13-2D 1200 cm and VIC15-1A 119 cm) were split photographed sub-
sampled and stored in the Pyrenean Institute of Ecology (IPE-CSIC Spain) Core
VIC13-2B was selected for detailed multiproxy analyses (including elemental
geochemistry C and N isotopes XRF and 14C dating) Stable isotope analyses
(δ13C δ15N) were performed at the Laboratory of Biogeochemistry and Applied
Stable Isotopes (LABASI-PUC Chile) Sediment samples for δ13Corg were pre-
treated with 50 ml of HCl and 50 ml of deionized water and dried at 60ordmC for 4 hr to
remove carbonates (Harris et al 2011) Isotope analyses were conducted using a
80
Delta V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via
a Conflo IV interface Isotope results are expressed in standard delta notation (δ)
and expressed in per mil (permil) relative to the Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Total carbon (TC)
Total Organic Carbon (TOC) Total Inorganic Carbon (TIC) and Total Sulfur (TS)
were measured every 2 cm with a LECO SC 144 DR at the IPE-CSIC
An AVAATECH X-Ray Fluorescence II core scanner with a Rh X-ray tube from
the University of Barcelona was used to obtain XRF logs every 4 mm of resolution
Results are expressed as element intensities in counts per second (cps) Tube
voltage was operated at 30 kV and 10 kV to obtain the abundances of 18 elements
(Pb Al Si S Cl K Ca Ti V Mn Fe Co Zn Br Rb Sr Y Zr) with an average of
at least of 1800 cps (less for Pb=437 and Zn=934 cps) Pb was included due to
similar behavior with Co and Fe Element ratios were calculated to describe changes
in redox conditions (MnFe) endogenic productivity (BrTi) carbonate formation
(SrCa and CaTi) and sediment input (ZrTi) (Barreiro-Lostres et al 2014
Carrevedo et al 2015 Frugone-Aacutelvarez et al 2017 Kylander et al 2011 Moreno
et al 2007a)
Several campaigns were carried out to sample the POM from the water column
two per hydrologic year from November 2015 to August 2018 A liter of water was
recovered in three sites through to the lake two are from the shallower areas (with
samples taken at 2 and 5 m depth at each site) and one in the deeper central portion
(at 2 5 and 20 m depth) of the lake (Fig 1d) The samples were filtered using glass
fiber filters (25 m) and then frozen to obtain the stable carbon and nitrogen isotope
signal of lacustrine POM Additionally soil and vegetation samples from the
following communities native species meadow hydrophytic vegetation and
81
Monterrey pine (Pinus radiata) were taken for stable isotope analyses (see in
supplementary material)
The age model for the complete Lago Vichuqueacuten sedimentary sequence is
based on Frugone-Aacutelvarez et al (2017) with the last 1000 years based on
210Pb137Cs dating techniques in VIC15-1A and 14C AMS dates on bulk sediment
samples (Supplementary Table S1) The 14C measurements of lake water DIC show
a moderate reservoir effect of 180 plusmn 25 14C yr) The revised age-depth model used
here includes three more 14C AMS dates performed with the program Bacon to
establish the deposition rates along the core (Blaauw and Christen 2011 Fig 2)
The age-depth model indicates that average resolution between 0 to 87 cm is lt2
cm per year and from 88 to 170 cm it is lt47 cm per year
82
Figure 2 Chronological age-depth model for the Lago Vichuqueacuten sedimentary
sequence spanning the last 1000 yr (see also Frugone-Alvarez et al 2017)
To estimate land use changes in the watershed we use Landsat MSS images
for 1975 and Landsat TM for 1989 and 2014 all taken in either summer or autumn
(Table 1) We performed supervised classification of land uses (maximum likelihood
83
algorithm) for each year (1975 1989 and 2014) and results were mapped using
ArcGIS 102
Table 1 Images using for LUCC reconstruction
Source of LUCC
Acquisition
Date Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat TM 19991226 30 m
CONAF 2009 30 m
Land cover Chile 2014 30 m
CONAF 2016 30 m
Previous Work on Lago Vichuqueacuten sedimentary sequence
The sediments are organic-poor dark brown to brown laminated silt with some
intercalated thin coarser clastic layers Lacustrine facies have been classified
according to elemental composition (TOC TS TIC and TN) grain size and
sedimentary textures (Frugone-Aacutelvarez et al 2017) Four main types of lacustrine
facies were identified in this short core Facies L1 is a laminated (1cm) black to dark
brown diatom-poor silt with relatively higher TOC percentages (TOC about 185)
TS (about of 18) and TOCTN (about 92) Facies L2 is characterized by a
homogeneous black organic-poor silty clay with low TOC (mean= 13) TS (mean=
13) TOCTN (mean= 88) values Facies L3 is a laminated (1cm) black organic-
poor (TOC mean 14) silts with higher TS (mean= 21) and TOCTN ratios
(mean= 88) Facies L3 is interpreted as ldquobaselinerdquo deposition on the distal areas
84
of Vichuqueacuten lake Mineralogy is dominated by mica (muscovite) clay minerals
(chlorite) and plagioclase (albite) Quartz and calcite occur at the top and pyrite
occurs in the lower part of the sequence Facies T is composed by massive banded
sediments 4 cm thick and coarse grain size It is interpreted as an instantaneous
depositional event (for more detail see Frugone-Aacutelvarez et al 2017) In this work
we identified four subunits based on geochemical and stable isotope signals
4 Results
41 Geochemistry and PCA analysis
High positive correlations exist between Al Si K and Ti (r = 078 ndash 096
supplementary Fig S1) and between Zn Zr Rb and Y (r = 072 - 096) which reflect
the silicate content of the watershed-derived sediments (Marzecoacuteva et al 2011) Zr
is commonly associated with minerals more abundant in coarser deposits Thus the
ZrTi profile could reflect grain size (Cuven et al 2010) showing higher variability
in the upper part of the Lake Vichuqueacuten sequence and in the alternation between
facies L1 and L2 Another group of elements made up of Co Fe and Pb displayed
positive correlations (r = 067ndash 097) and represents the input of heavy metals Br
Cl and Ca show a positive correlation (r = 049 ndash 06 p-value = 00001) BrTi ratio
is interpreted as a productivity indicator due to Br having a strong affinity with humic
and organic compounds (Marzecoacuteva et al 2011 Frugone-Aacutelvarez et al 2017) In
our Lago Vichuqueacuten sequence BrTi values are high from 170 to 132 cm and from
36 cm to the top of core where it shows several sharps peaks (Fig 3) The MnFe
ratio is indicative of lake bottom oxygenation (Naecher et al 2013) as under
reducing conditions Mn tends to become more mobile than Fe leading to a
decreased MnFe (Marzecoacuteva et al 2011) Several sharp peaks in MnFe occurred
85
from 127 to 120 cm and from 57 to 30 cm (MgFegt03) The silicate group and the
Br Cl Ca Mn group are negatively correlated (r= -012 and -066)
Principal Component Analysis (PCA) was undertaken on the XRF
geochemical data to investigate the main factors controlling sediment deposition in
Lago Vichuqueacuten (Fig 3) The four main eigenvectors explain 801 of the variance
(supplementary material Table S2) The principal component (PC1) explain 437
of the total of variance and grouped elements are associated with terrigenous input
to the lake Positive values of the biplot have been attributed to higher heavy metals
deposition (Pb Co and Fe) and negative values to higher silicate input (Al K Ti and
Si) in a high energy catchment environment (Zr) The PC2 explains 198 of the
total of variance and highlights the endogenic productivity in the lake The positive
loading is compound by Br Cl Ca and Sr and to a lesser extent by Y Zn Rb and
Mn The PC2 may explain both the precipitation of carbonates (Ca Sr) and biological
production (Br)
86
Figure 3 Principal Component Analysis of XRF geochemical measurements in
VIC13-2B Lago Vichuqueacuten lake sediments
42 Sedimentary units
Based on geochemical and stable isotope analysis we identified four
lithological subunits in the short core sedimentary sequence Our PCA analyses and
Pearson correlations pointed out which variables were better for characterizing the
subunits (Fig 4) in terms of endogenic productivity heavy metals and terrestrial
input The unit 2c (170-131 cm) is characterized by the occurrence of some clastic
layers (L2 from 160 and 150 cm) high δsup1⁵N values (mean= +59 plusmn 04permil) with
Magnetic Susceptibility (MS) BrTi ZrTi and SrCa values increase towards the top
Also it has relatively low δsup1sup3C values (mean= -265 plusmn 11permil) and CNmolar ratios
(mean=78 plusmn 04) The ZrTi show upwards increasing values reaching the highest
values of the sequence at the top of this unit suggesting a coarsening upward trend
and relatively higher depositional energy The MS trend also indicates higher
erosion in the watershed and enhanced delivery of ferromagnetic minerals likely
from soil particles particularly from 150 to 160 cm (Frugone-Aacutelvarez at al 2017)
The subunit 2b (130-118 cm) is also composed of black silts but it has the
lowest MS values of the whole sequence and its onset is marked by a sharp
decrease in ZrTi This subunit is marked by sharp peaks of MnFe (at 126 and 120
cmgt027) SrCa (at 125 and 120 cmgt033) CaTi (at 124 and 120 cmgt199) TOC
(about 26 at 130 124 122 and 118 cm) and δsup1⁵N (gt+67permil at 126 and 120 cm)
BrTi oscillates around low values (about 01) whereas the δsup1sup3C values range
between -262 and -282permil
87
The unit 2a (58-117 cm) shows increasing and then decreasing MS values
and displayed sharps peaks of δsup1⁵N (gt64 at 102 to 99 87 and 75 cm) and CN
(about 10 at 86 74 66 and 60 cm) SrCa (mean = 04 plusmn 013) BrTi (mean = 008
plusmn 002) MnFe (mean = 002 plusmn 001) and δsup1sup3C (mean = -260 plusmn 07permil) oscillate in
low values The upper part of this unit (72 ndash 57 cm) shows an increase of SrCa
(from 03 to 05)
The upper unit 1a (0 - 57 cm) starts with a fine clastic layer (T at 57 to 54
cm) interpreted as deposition during a high-energy event It is characterized by
lower BrTi (mean = 003 plusmn 002) TOC (mean = 12 plusmn 03) and δsup1sup3C (mean= -
266 plusmn 06permil) higher δsup1⁵N (mean = +55 plusmn 08 permil) This unit is made of alternating
fine (lt 1cm) black and dark brown laminae with carbonate (L1 facies) Recently
deposited sediments show decreasing MS values increasing TOC (mean= 15 plusmn
04 ) sharp peaks of BrTi (gt015 at 33 21 5 and 1 cm) and the lowest δsup1⁵N values
of the entire record (mean= 45 plusmn 06 permil) and δsup1sup3C values (mean= -277 plusmn 09 permil)
Heavy metal content increases abruptly in the upper section of the core (2 - 20 cm)
(peaks of FeTi CoTi and PbTi)
88
Figure 4 Sedimentary units of Lago Vichuqueacuten sedimentary sequence Selected
variables illustrate watershed input (Magnetic Susceptibility ZrTi CoTi and MnFe)
endogenic productivity (SrCa CaTi TS) organic matter source (BrTi TOC
CNmolar and stable isotope records (δ13Corg and δ15Nbulk)
43 Recent seasonal changes of particulate organic matter on water column
The average of δ15NPOM from the three sites across to the lake was +86 plusmn 58
permil oscillating widely from -14 to +193permil (Fig 5) Important seasonal differences
occur at 2 and 5 m depth with more negative values during summer (~53 plusmn 52 permil)
than winter (~122 plusmn 40permil) At 20 m depth δ15NPOM values exhibit small seasonal
ranges (mean = +10permil for winter and +87permil for summer) The δ13CPOM average was
-296 plusmn 33permil with slightly seasonal and water column depth differences However
more positive δ13CPOM values occurred during winter (mean=-290 plusmn 41permil) than in
summer (mean= -301 plusmn 19 permil) Like the δ15NPOM values CNPOM (mean= 83 plusmn 17)
displayed important seasonal and water depth differences Lower CNPOM ratios
89
occurred during summer at 2 and 5 m depth (mean= 95 plusmn 17) and higher and more
constant values during winter (mean = 73 plusmn 09) at the same depth At 20 m CNPOM
shows similar values in both winter (70) and summer (74)
Figure 5 Seasonal POM averages from samples taken of the Lago Vichuqueacuten
water column Samples were taken from 2015 to 2018 at 2 (n=16) 5 (n=14) and 20
(n=8) meters depth
44 Stable isotope values across the Lake Vichuqueacuten watershed
Figure 6 shows modern vegetation soil and sediment isotope values found for
the watershed Huintildee tributary and Vichuqueacuten lake The δsup1⁵NFoliar values from
meadow plantations and macrophytes have similar range values with a mean of
+89 plusmn50 permil (n=18) +74 plusmn 75 permil (n=5) +82 plusmn 36 permil (n=6) respectively Native
vegetation has overall lower δsup1⁵NFoliar values with a mean of 14 plusmn 21 permil (n=28 see
Table 2 in supplementary material) The δsup1sup3CFoliar from terrestrial samples exhibit
similar values across the different plant communities (tree plantation mean=-274 plusmn
13permil meadow mean = -275 plusmn 23 permil native species mean=-272 plusmn 14permil) whereas
macrophytes display slightly more negative values with a mean of -287 plusmn 23permil
Macrophytes substrate displayed the highest δsup1⁵N values with a mean of +60 plusmn
14permil (n=3) Similar and relatively high δsup1⁵N values for soil of plantation (mean= +54
plusmn 13permil n=4) meadow (mean= +49 plusmn 04permil n=8) and surface river sediment
90
(mean= +52 plusmn 17permil n=10) Native forest soils oscillate around slightly more
negative values with a mean of +45 plusmn 12permil (n=4) Slightly more positive soil δsup1sup3C
values occur both underneath native forests and in tree plantations with means of -
284 plusmn 23permil and -298 plusmn 13permil respectively compared to either meadow soils
(mean= -302 plusmn 11permil) aquatic sediments underneath macrophytes (-311 plusmn 10permil)
or from surface river sediments (mean= -312 plusmn 10permil)
Figure 6 Stable isotope content of modern soil river sediment and foliar vegetation
used as end members in the sedimentary sequence of Lago Vichuqueacuten a)
Scatterplot of foliar δsup1⁵N and δsup13C values for vegetation from Lago Vichuqueacuten
watershed (plantation meadow and native species) and macrophytes on Lake
Vichuqueacuten See supplementary material for more detail of vegetation types b)
Scatterplot of soil δsup1⁵N and δsup13C values across the watershed the sediments of the
Huintildee and Vichuqueacuten rivers and the fluvial sediment corresponding to the
macrophyte vegetation
45 Land use and cover change from 1975 to 2014
Major land use changes between 1975 CE and 2016 CE in the Lago
Vichuqueacuten watershed are summarized in Figure 7 The watershed has a surface
area of 535329 km2 of which native vegetation (26) and shrublands (53)
represent 79 of the total surface in 1975 Meadows are confined to the valley and
91
represent 17 of watershed surface Tree plantations initially occupied 1 of the
watershed and were first located along the lake periphery By 1989 the areas of
native forests shrublands and meadows had decreased to 22 31 and 14
respectively whereas tree plantations had expanded to 30 These trends
continued almost invariably until 2016 when shrublands and meadows reached 17
and 5 of the total areas while tree plantations increased to 66 Native forests
had practically disappeared by 1989 and then increased up to 7 of the total area
in 2016 (Fig 6) preferentially occupying the southeastern sector of the watershed
Figure 7 Changes in land use and cover between 1975 and 2016 in the Lago
Vichuqueacuten watershed as measured from satellite images The major change is
represented by the replacement of native forest shrubland and meadows by
plantations of Monterrey pine (Pinus radiata)
Figure 8 shows correlations between lake sediment stable isotope values and
changes in the soil cover from 1975 to 2013 Positive relationships occurred
between sediment δsup1⁵N and δsup13C values from core VIC13-2B (unit 1) and the
92
percentage of native forest (r=089 for δsup1⁵N 082 for δsup13C) shrublands (r = 071 for
δsup1⁵N 057 for δsup13C) and meadow cover (r = 088 for δsup1⁵N 081 for δsup13C) All of these
correlations are significant (p value lt 0001) In contrast significant negative
correlations (p lt0001) occurred between tree plantation cover and lake sediment
stable isotope values (δsup1⁵N r = -082 and δsup13C r = -067) native forest (r = -097)
meadows (r = -086) and shrubland (r =-093)
Figure 8 Correlation plots of land use and cover change versus lake sediment
stable isotope values The δsup1⁵N values are positively correlated with native forests
agricultural fields and meadow cover across the watershed Total Plantation area
increases are negatively correlated with native forest meadow and shrubland total
area Significance levels are indicated by the symbols p-values (0 0001 001
005 01 1) lt=gt symbols ( )
93
5 Discussion
51 Seasonal variability of POM in the water column
The stable isotope values of POM can vary during the annual cycle due to
climate and biologic controls namely temperature and length of the photoperiod
which affect phytoplankton growth rates and isotope fractionation in the water
column (Gu et al 2006 Leng et al 2006) POM from Lago Vichuqueacuten surface
samples (2 and 5 m depth) displayed lower δ13CPOM values during the summer than
in winter During C uptake phytoplankton preferentially utilize 12C leaving the
DICpool enriched in 13C Therefore as temperature increases during the summer
phytoplankton growth generates OM enriched in 12C until this becomes depleted
and then the biomas come to enriched u At the onset of winter the DICpool is now
enriched in 13C and despite an overall decrease in phytoplankton production the
OM produced will also be enriched in 13C In contrast δ13CPOM values at 20 m depth
did not reflect these seasonal differences probably due to water-column
stratification that maintains similar temperatures and biological activity throughout
the year
Lake N availability depends on N sources including inputs from the
watershed and the atmosphere (ie deposition of N compounds and fixation of
atmospheric N2) which varies during the hydrologic year The fixation of atmospheric
N2 is an important natural source of N to the lake occurring mainly during the
summer season associated with higher temperature and light (Gu et al 2006)
Because the δ15NPOM of N2 fixers is close to 0permil with almost no overall isotope
fractionation (Leng et al 2006) when N2 is the major N source δ15NPOM values are
typically low However when DIN concentrations are high or alternatively when little
94
N2 fixation occurs δ15NPOM values will be higher (Gu et al 2006) δ15NPOM values
from summer Lago Vichuqueacuten samples were lower than those from winter with large
differences between the seasons (~5permil Fig 5 and 9) Furthermore δ15NPOM values
were high when monthly average temperature was low and monthly precipitation
was high (Fig 9) This pattern is consistent with the dominance of high N2 fixation
by cyanobacteria associated with increased summer temperatures This correlation
of δ15NPOM values with temperature further suggests a functional group shift i e
from N fixers to phytoplankton that uptake DIN The correlation between wetter
months and higher δ15NPOM values could be caused by increased N input from the
watershed due to increased runoff during the winter season The lack of data of the
δ15NPOM between the 30 mm and the 160 mm of precipitation is due to the
mediterranean-type climate that concentrates precipitations in the winter months
Finally Lago Vichuqueacuten is characterized by pronounced seasonality leading to
higher phytoplankton biomass in summer characterized by low δ15NPOM In winter
low biomass production and increased input from watershed is associated to high
δ15NPOM
95
Figure 9 Seasonal changes can drive δ15NPOM values in Lago Vichuqueacuten Data
correspond to average monthly temperature and total monthly precipitation for the
months when the water samples were taken (years 2015 - 2018) P-valuelt005
52 Stable isotope signatures in the Lake Vichuqueacuten watershed
The natural abundance of 15N14N isotopes of soil and vegetation samples
from the Lago Vichuqueacuten watershed appear to result from a combination of factors
isotope fractionation different N sources for plants and soil microorganisms (eg N2
fixation Nitrates) depth in the soil where N uptake by vegetation occurs and N loss
mechanisms (ie denitrification leaching and ammonia volatilization Hogberg
1997) The lowest δsup1⁵Nfoliar values are associated with native species and are
probably due to the contribution of N-fixing species (eg Sophora) (Fig6 and for
more detail see Table S3 in supplementary material) The number native N-fixers
species present in the Chilean mediterranean vegetation are not well known
however so there could be other N-fixing species in this group Alternatively δsup1⁵Nfoliar
values reflect soil N uptake (Kahmen et al 2008) In environments limited by N
plants have δsup1⁵N values similar to the soil δsup1⁵N values however the denitrification
and volatilization of ammonia can lead to the remain N of soil to come enriched in
15N (Cifuentes et al 1989 Craine et al 2009 Bruland and Mckenzie 2010) Soil N
isotope samples from native species communities tends to display relatively high
δsup1⁵N values respect to foliar samples due to loss of N-soil
The higher foliar and soil δsup1⁵N values obtained from samples of meadows
aquatic macrophytes and tree plantations can be attributed to the presence of
greater amounts of nitrate available for plant uptake (Fig 6) Houmlgberg (1997)
suggests that the availability of different N sources in soils (ie nitrates versus
96
ammonia) with different residence times can also explain these δsup1⁵NFoliar values
Indeed Feigin et al (1974) described differences of up to 20permil between ammonia
and nitrates sources Denitrification and nitrification discriminate much more against
15N than N2 fixation does (Craine et al 2009) leading to soils (and plants after
uptake) enriched in 14N
In general multiple processes that affect the isotopic signal result in similar
δsup1⁵N values between the soil of the watershed and the sediments of the river
However POM isotope fluctuations allow to say that more negative δsup1⁵N values are
associated to lake productivity while more positive δsup1⁵N values are associated with
N input from the watershed
δ13C values in Lago Vichuqueacuten watershed reflect CO2 gas exchange between
C3 plants and algae with the atmosphere During photosynthesis plants
discriminate against the heavier stable isotope of carbon (13C) in favor of the lighter
isotope 12C which results in δ13CFoliar values between -220permil and -320permil (Browman
and Cook 2002 Liu et al 2007) Likewise the δ13CFoliar values from Lago Vichuqueacuten
oscillate around -27permil (Fig 6) Plants transform atmospheric CO2 into soil organic
carbon (C) which in turn reflects this initial discrimination against 13C during C
uptake and post photosynthetic fractionation (Farquhar et al 1982 1989 Badeck
et al 2005 Pausch and Kuzyakov 2017) Slightly more positive δ13CFoliar values
(about 15permil) were measured in comparison with their δ13CSoil values This may be
reflecting the C transference from plants to the soil but also a soil-atmosphere
interchange The preferential assimilation of the light isotopes (12C) during soil
respiration carried by the roots and the microbial biomass that is associated with the
decomposition of litter roots and soil organic matter explain this differential
(Berhardt et al 2006 Blagodatskaya et al 2010 Midwood and Millard 2011)
97
In general the δ13CSoil values from the Lago Vichuqueacuten watershed oscillated
around -290permil and did not vary with our plant classification types Here we use
these values as terrestrial-end members to track changes in source OM (Fig 6)
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from the terrestrial watershed By the other hand more positive δ13C
values most likely reflect an increased aquatic OM component as indicated by POM
isotope fluctuations (Fig 9)
53 Recently land use and cover change and its influences on N inputs to the lake
Tree plantations in the Lago Vichuqueacuten watershed have increased greatly in
the last few decades (from 1 in 1975 to 66 in 2016 Fig 7) replacing previous
native forests (from 26 to 7) meadows (17 to 5) and shrublands (53 to
17) In 1975 tree plantations were confined to the lake perimeter with discrete
patches distributed throughout the watershed The ldquoForest Lawrdquo (DFL 701- passed
in 1974) allocated state funding to afforestation efforts and management of tree
plantations which greatly favored the replacement native forests by introduced trees
This increase is marked by a sharp and steady decrease in lake sediment δ15N and
δ13C values because tree plantations function as a nutrient sink whereas other land
uses such as meadows favor nutrient loss from the watershed (Fig8) Bruland and
Mackenzie (2014) noted a decrease in wetland δ15N values when watershed
forested cover increased and concluded that N inputs to the wetlands are lower from
the forested areas as they generally do not export as much N as agricultural lands
A positive correlation between native vegetation and δ15Ncore values can be
explained by the relatively scarce arboreal cover in the watershed in 1975 when
native forest occupied just 26 of the watershed surface whereas shrublands and
98
meadows occupied more than the 70 of the surface of the watershed with the
concomitant elevated loss of N (Fig 7 and Fig 8)
54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years
Sedimentological compositional and geochemical indicators all show changes in
the N cycle associated to LUCC lake dynamics and climate changes (Fig 10) From
the pre-Columbian indigenous settlement including the Spanish colonial period up
to the start of the Republic (1300 - 1800 CE) the introduction of crops such as
quinoa and wheat but also the clearing of land for extensive agriculture would have
favored the entry of N into the lake Conversely major changes observed during the
last century were characterized by a sharp decrease of N input that were coeval
with the increase of tree plantations
From 1365 until 1550 CE (unit 2c 2b and half of 2a Fig 3 and Fig 10) pre-
Columbian agricultural societies planted mostly crops of corn and quinoa (Ramirez
and Vidal 1985) This period is characterized by large peaks seen in the δsup1⁵N record
(eg 1403 1452 1476 and 1505) with overall high δsup1⁵N values (up to 5permil) indicating
that N input from watershed was elevated and oscillating to the beat of the NT These
positive δsup1⁵N peaks could be due to several causes including a) the clearing of land
for farming b) N loss via denitrification which would be generally augmented in
anoxic environments (Diebel et al 2009) such as Lago Vichuqueacuten (low MnFe
values about 0001 Fig 4) or c) climate especially cold-wet winters or hot-dry
summers can also exert control on the δsup1⁵N record Indeed tree-ring records and
summer temperature reconstructions show overall wetcold conditions during this
period (Christie et al 2009 Von Gunten et al 2009 Fig 10) Increased
precipitation would bring more sediment (and nutrients) from the watershed into the
99
lake and increase lake productivity which is also detected by the geochemical
proxies for productivity (peaks of BrTi SrCa and TOC Fig 4 and Fig 10) (see also
Frugone-Alvarez et al 2017)
Figure 10 Changes in the N availability during the last six centuries in Lago
Vichuqueacuten a) lake sediment δsup1⁵N values displayed more positive values during the
prehistoric period Spanish Colony and the starting 19th century which is associated
with enhanced N input from the watershed by extensive clearing and crop
plantations The inset shows this relationship between sediment δsup1⁵N and
100
percentage of meadow cover over the last 30 years b) Summer temperature
reconstruction from central Chile (von Gunten et al 2009) showing a
correspondence with cold phases and high δsup1⁵N values at Vichuqueacuten except for the
last 100 years c) Palmer Drought Severity Index (PDSI) is an interannual moisture
variability reconstruction for late springndashearly summer during the last six centuries
(Christie et al 2009) Grey shadow indicating higher precipitation periods
From 1550 and 1810 CE (unit 2a) and during the Colonial period several peaks
of δ15N could be further related not only to high precipitation (Fig 10 grey shadow)
but also pulses of enhanced N input from the watershed linked to human land use
In 1550 CE Juan Cuevas was granted lands and indigenous workers under the
encomienda system for agricultural and mining development of the Vichuqueacuten
village (Ramiacuterez and Vidal 1985) Historical documents show that around 1579 CE
the Vichuqueacuten watershed was occupied by indigenous communities dedicated to
wheat plantations and vineyards wood extraction and gold mining (Odone 1998)
The introduction of the Spanish agricultural system implied not just a change in the
types of crops used (from quinoa to vineyards and wheat) but also a clearing of
native species for the continuous increase of agricultural surface and wood
extraction During the Colonial period wheat from Vichuqueacuten was exported to Peru
(Ramiacuterez and Vidal 1985) Armesto et al (2009) indicate that during the XVIII and
XIX centuries the extraction of wood for mining operations was important enough to
cause extensive loss of native forests The independence and instauration of the
Chilean Republic did not change this prevailing system Increases in the
contributions of N to the lake during the second half of the XIX century (peaks in
δsup1⁵N ca 1870 CE) could be due to expanding farm land dedicated for wheat
101
production and increased commercial trade with California and Canada (Ramiacuterez
and Vidal 1985)
In contrast LUCC in the last century are clearly related to the development of
large-scale tree plantations (Fig 7) The δsup1⁵N record reaches the lowest values of
the entire sequence in the last few decades (Fig 10) A marked increase in lake
productivity NT concentration and decreasing sediment input is synchronous (unit
1 Fig 4) with trees replacing meadows shrublands and areas with native forests
(Fig 7 and 8) The implementation of the lsquoForest Lawrsquo in 1974 had a stronger impact
on the landscape and lake ecosystem dynamics than the impacts of ongoing climate
change in the region which is much more recent (Garreaud et al 2018) although
the prevalence of hot dry summers seen over the last decade would also be
associated with low δsup1⁵N values and increase of NT (Fig 10) Heavy metals ratios
(FeTi CoTi and PbTi) show notable increases in lake sediments from 1992 to 2011
CE (Fig 4) Although this could be related to mining in the El Maule region the
closest mines are 60 Km away (Pencahue and Romeral) so local factors related to
shoreline urbanization for the summer homes and an increase in tourist activity
could also be a major factor
6 Conclusions
The N isotope signal in the watershed depends on the rates of exchange
between vegetation and soil cover (Fig 11) Plants preferentially uptake 14N and the
underlying soils become enriched in 15N especially when the terrestrial ecosystem
is N-limited andor significant N loss occurs (ie denitrification andor ammonia
volatilization) LUCC in the Lago Vichuqueacuten watershed have heavily modified the
links between terrestrial and aquatic ecosystems with agriculture practices
102
contributing more N to the lake than tree plantations or native forests In situ lake
processes can also fractionate N isotopes An increase of N-fixing species results
in OM depleted in 15N which results in POM with lower δsup1⁵N values during these
periods During winter phytoplankton is typically enriched in 15N due to the
decreased abundance of N-fixing species and increased N input from the
watershed This in turn generates the highest δsup1⁵NPOM values in Lago Vichuqueacuten
Periods of elevated productivity tend to deplete the dissolved inorganic N of 14N
resulting in even higher δ15N values
Our study illustrates the usefulness of δsup1⁵N as a tool for reconstructing the past
influence of LUCC on N availability in lake ecosystems To constrain the relative
roles of the diverse forcing mechanisms that can alter N cycling in mediterranean
ecosystems all main components of the N cycle should be monitored seasonally
(or monthly) including the measurements of δ15N values in land samples
(vegetation-soil) as well as POM
103
Figure 11 Summary of human and environmental factors controlling the δ15N
values of lake sediments Particulate organic matter(POM) δ15N values in
mediterranean lakes are driven by N input from the watershed that in turn depend
on land use and cover changes (ie forest plantation agriculture) andor seasonal
changes in N sources andor lake ecosystem processes (ie bioproductivity redox
condition denitrification) Arrows indicate the direction of N fluxes (inputoutput from
the N cycle) N cycle processes that deplete lake sediments of 15N are shown in
blue whereas those that enrich sediments in 15N are shown in red
104
Supplementary material
Figure S1 Pearson correlate coefficient between geochemical variables in core
VIC13-2B Positive and large correlations are in blue whereas negative and small
correlations are in red (p valuelt0001)
Figure S2 Principal Component Analysis of geochemical elements from core
VIC13-2B
105
Table S1 Lago Vichuqueacuten radiocarbon samples
RADIOCARBON
LAB CODE
SAMPLE
CODE
DEPTH
(m)
MATERIAL
DATED
14C AGE ERROR
D-AMS 029287
VIC13-2B-
1 043 Bulk 1520 24
D-AMS 029285
VIC13-2B-
2 085 Bulk 1700 22
D-AMS 029286
VIC13-2B-
2 124 Bulk 1100 29
Poz-63883 Chill-2D-1 191 Bulk 945 30
D-AMS 001133
VIC11-2A-
2 201 Bulk 1150 44
Poz-63884
Chill-2D-
1U 299 Bulk 1935 30
Poz-64089
VIC13-2D-
2U 463 Bulk 1845 30
Poz-64090
VIC13-2A-
3U 469 Bulk 1830 35
D-AMS 010068
VIC13-2D-
4U 667 Bulk 2831 25
Poz-63886
VIC13-2D-
4U 719 Bulk 3375 35
106
D-AMS 010069
VIC13-2D-
5U 775 Bulk 3143 27
Poz-64088
VIC13-2D-
5U 807 Bulk 3835 35
D-AMS-010066
VIC13-2D-
7U 1075 Bulk 6174 31
Poz-63885
VIC13-2D-
7U 1197 Bulk 6440 40
Poz-5782 VIC13-15 DIC 180 25
Table S2 Loadings of the trace chemical elements used in the PCA
Elementos PC1 PC2 PC3 PC4
Zr 0922 0025 -0108 -0007
Zn 0913 -0124 -0212 0001
Rb 0898 -0057 -0228 0016
K 0843 0459 0108 0113
Ti 0827 0497 0060 -0029
Al 0806 0467 0080 0107
Si 0803 0474 0133 0136
Y 0784 -0293 -0174 0262
V 0766 0455 0090 -0057
Br 0422 -0716 -0045 0226
Ca 0316 -0429 0577 0489
Sr 0164 -0420 0342 -0182
Cl 0151 -0781 -0397 0162
107
Mn -0121 -0091 0859 0095
S -0174 -0179 -0051 0714
Pb -0349 0414 -0282 0500
Fe -0700 0584 -0023 0280
Co -0704 0564 -0107 0250
Table S3 Stable Isotope values from vegetation of Lago Vichuqueacutenacutes watershed
Taxa Classification δsup1⁵N δsup1sup3C CN
molar
Poaceae Meadow 1216 -2589 3602
Juncacea Meadow 1404 -2450 3855
Cyperaceae Meadow 1031 -2596 1711
Taraxacum
officinale Meadow 836 -2400 2035
Poaceae Meadow 660 -2779 1583
Poaceae Meadow 453 -2813 1401
Poaceae Meadow 966 -2908 4010
Juncus Meadow 1247 -2418 3892
Poaceae Meadow 747 -3177 6992
Poaceae Meadow 942 -2764 3147
Poaceae Meadow 1479 -2634 2895
Poaceae Meadow 1113 -2776 1795
Poaceae Meadow 2215 -2737 7971
Poaceae Meadow 1121 -2944 2934
Poaceae Meadow 638 -3206 1529
108
Macrophytes Macrophytes 886 -3044 2286
Macrophytes Macrophytes 1056 -2720 2673
Macrophytes Macrophytes 769 -3297 1249
Macrophytes Macrophytes 967 -2763 1442
Macrophytes Macrophytes 959 -2670 2105
Macrophytes Macrophytes 334 -2728 1038
Acacia dealbata
Introduced
species 656 -2696 1296
Acacia dealbata
Introduced
species 487 -2941 1782
Acacia dealbata
Introduced
species 220 -2611 3888
Luma apiculata Native species 433 -2542 4135
Luma apiculata Native species 171 -2664 7634
Luma apiculata Native species -001 -2736 6283
Luma apiculata Native species 029 -2764 6425
Azara sp Native species 159 -2868 8408
Azara sp Native species 101 -2606 2885
Baccharis concava Native species 104 -2699 5779
Baccharis concava Native species 265 -2488 4325
Baccharis concava Native species 287 -2562 7802
Baccharis concava Native species 427 -2781 5204
Baccharis linearis Native species 190 -2610 4414
Baccharis linearis Native species 023 -2825 5647
109
Peumus boldus Native species 042 -2969 6327
Peumus boldus Native species 205 -2746 4110
Peumus boldus Native species 183 -2743 6293
Chusquea quila Native species 482 -2801 4275
Poaceae meadow 217 -2629 7214
Lobelia sp Native species 224 -2645 3963
Lobelia sp Native species -091 -2565 4538
Aristotelia chilensis Native species -035 -2785 5247
Aristotelia chilensis Native species -305 -2889 2305
Aristotelia chilensis Native species 093 -2836 5457
Chusquea quila Native species 173 -2754 3534
Chusquea quila Native species 045 -2950 6739
Quillaja saponaria Native species 223 -2838 9385
Scirpus meadow 018 -2820 7115
Sophora sp Native species -184 -2481 2094
Sophora sp Native species -181 -2717 1721
Pinus radiata
Introduced
trees 1581 -2602 3679
Pinus radiata
Introduced
trees 1431 -2784 4852
Pinus radiata
Introduced
trees -091 -2708 9760
Pinus radiata
Introduced
trees 153 -2568 3470
110
Salix sp
Introduced
trees 632 -2878 1921
LITERATURE CITED
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A historical
framework for land cover change in southwestern South America in the past 15000
years Land use policy 27 148ndash160
httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the next
carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014
httpsdoiorg101002eft2235
Blaauw M Christeny JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474
httpsdoiorg10121411-BA618
Bowman DMJS Cook GD 2002 Can stable carbon isotopes (δ13C) in soil
carbon be used to describe the dynamics of Eucalyptus savanna-rainforest
boundaries in the Australian monsoon tropics Austral Ecol
httpsdoiorg101046j1442-9993200201158x
Brahney J Ballantyne AP Turner BL Spaulding SA Otu M Neff JC 2014
Separating the influences of diagenesis productivity and anthropogenic nitrogen
deposition on sedimentary δ15N variations Org Geochem 75 140ndash150
httpsdoiorg101016jorggeochem201407003
111
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409
httpsdoiorg102134jeq20090005
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego R
Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and
environmental change from a high Andean lake Laguna del Maule central Chile
(36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the
Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from
tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Craine JM Elmore AJ Aidar MPM Dawson TE Hobbie EA Kahmen A
Mack MC Kendra K Michelsen A Nardoto GB Pardo LH Pentildeuelas J
Reich B Schuur EAG Stock WD Templer PH Virginia RA Welker JM
Wright IJ 2009 Global patterns of foliar nitrogen isotopes and their relationships
with cliamte mycorrhizal fungi foliar nutrient concentrations and nitrogen
availability New Phytol 183 980ndash992 httpsdoiorg101111j1469-
8137200902917x
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty
Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-
010-9453-1
Diebel M Vander Zanden MJ 2012 Nitrogen stable isotopes in streams stable
isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19
1127ndash1134 httpsdoiorg10189008-03271
112
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in freshwater
wetlands record long-term changes in watershed nitrogen source and land use SO
- Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash
2916
Fariacuteas L Castro-Gonzales M Cornejo M Charpentier J Faundez J
Boontanon N Yoshida N 2009 Denitrification and nitrous oxide cycling within the
upper oxycline of the oxygen minimum zone off the eastern tropical South Pacific
Limnol Oceanogr 54 132ndash144
Farquhar GD Ehleringer JR Hubick KT 1989 Carbon Isotope Discrimination
and Photosynthesis Annu Rev Plant Physiol Plant Mol Biol
httpsdoiorg101146annurevpp40060189002443
Farquhar GD OrsquoLeary MH Berry JA 1982 On the relationship between
carbon isotope discrimination and the intercellular carbon dioxide concentration in
leaves Aust J Plant Physiol httpsdoiorg101071PP9820121
Fogel ML Cifuentes LA 1993 Isotope fractionation during primary production
Org Geochem httpsdoiorg101007978-1-4615-2890-6_3
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A
Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-
resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS)
implications for past sea level and environmental variability J Quat Sci 32 830ndash
844 httpsdoiorg101002jqs2936
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924
httpsdoiorg104319lo20095430917
113
Gerhart LM McLauchlan KK 2014 Reconstructing terrestrial nutrient cycling
using stable nitrogen isotopes in wood Biogeochemistry 120 1ndash21
httpsdoiorg101007s10533-014-9988-8
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and oxygen
isotope fractionation during dissimilatory nitrate reduction by denitrifying bacteria 53
2533ndash2545 httpsdoiorg10230740058342
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater
eutrophic lake Limnol Oceanogr 51 2837ndash2848
httpsdoiorg104319lo20065162837
Harris D Horwaacuteth WR van Kessel C 2001 Acid fumigation of soils to remove
carbonates prior to total organic carbon or CARBON-13 isotopic analysis Soil Sci
Soc Am J 65 1853 httpsdoiorg102136sssaj20011853
Houmlgberg P 1997 Tansley review no 95 natural abundance in soil-plant systems
New Phytol httpsdoiorg101046j1469-8137199700808x
Kylander ME Ampel L Wohlfarth B Veres D 2011 High-resolution X-ray
fluorescence core scanning analysis of Les Echets (France) sedimentary sequence
New insights from chemical proxies J Quat Sci 26 109ndash117
httpsdoiorg101002jqs1438
Lara A Solari ME Prieto MDR Pentildea MP 2012 Reconstruccioacuten de la
cobertura de la vegetacioacuten y uso del suelo hacia 1550 y sus cambios a 2007 en la
ecorregioacuten de los bosques valdivianos lluviosos de Chile (35o - 43o 30acute S) Bosque
(Valdivia) 33 03ndash04 httpsdoiorg104067s0717-92002012000100002
Lehmann M Bernasconi S Barbieri A McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during
114
simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66
3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007
Liu G Li M An L 2007 The Environmental Significance Of Stable Carbon
Isotope Composition Of Modern Plant Leaves In The Northern Tibetan Plateau
China Arctic Antarct Alp Res httpsdoiorg1016571523-0430(07-505)[liu-
g]20co2
Marzecovaacute A Mikomaumlgi A Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54
httpsdoiorg103176eco2011105
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash
1643 httpsdoiorg1011770959683613496289
Meyers PA 2003 Application of organic geochemistry to paleolimnological
reconstruction a summary of examples from the Laurention Great Lakes Org
Geochem 34 261ndash289 httpsdoiorg101016S0146-6380(02)00168-7
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland
Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Odone C 1998 El Pueblo de Indios de Vichuqueacuten siglos XVI y XVII Rev Hist
Indiacutegena 3 19ndash67
Pausch J Kuzyakov Y 2018 Carbon input by roots into the soil Quantification of
rhizodeposition from root to ecosystem scale Glob Chang Biol
httpsdoiorg101111gcb13850
115
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98
httpsdoiorg1011772053019614564785
Sun W Zhang E Jones RT Liu E Shen J 2016 Biogeochemical processes
and response to climate change recorded in the isotopes of lacustrine organic
matter southeastern Qinghai-Tibetan Plateau China Palaeogeogr Palaeoclimatol
Palaeoecol httpsdoiorg101016jpalaeo201604013
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of
different trophic status J Paleolimnol 47 693ndash706
httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl
httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M 2009 High-resolution quantitative climate
reconstruction over the past 1000 years and pollution history derived from lake
sediments in Central Chile Philos Fak PhD 246
Woodward C Chang J Zawadzki A Shulmeister J Haworth R Collecutt S
Jacobsen G 2011 Evidence against early nineteenth century major European
induced environmental impacts by illegal settlers in the New England Tablelands
south eastern Australia Quat Sci Rev 30 3743ndash3747
httpsdoiorg101016jquascirev201110014
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K Yeager
KM 2016 Different responses of sedimentary δ15N to climatic changes and
116
anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau
J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
117
DISCUSION GENERAL
El Nitroacutegeno (N) es un elemento clave que controla la dinaacutemica y
funcionamiento de ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al
1997) Pese a que el N (N2) es muy abundante en la atmoacutesfera (78) es escaso
en los ecosistemas pues la mayoriacutea de la biota no puede acceder a su forma
molecular (Galloway et al 1995 Zaehle et al 2013) En este sentido la entrada
natural de N en los ecosistemas es viacutea fijacioacuten bioloacutegica del N2 por bacterias que lo
convierten en compuestos bioloacutegicamente disponibles (Battye et al 2017) Debido
a que la fijacioacuten es un proceso energeacuteticamente costoso los ecosistemas
comuacutenmente estaacuten limitados por N (Vitousek and Howarth 1991) Los LUCC
contribuyen al incremento del N disponible y son una de las principales causas de
eutroficacioacuten y contaminacioacuten de los cuerpos de agua (Woodward et al 2012)
En Chile central los LUCC principalmente relacionados con las actividades
agriacutecolas y forestales han causado grandes cambios en el ciclo del N debido al
118
reemplazo de la cobertura natural por un monocultivo y al uso de fertilizantes que
modifican los aportes de MO y N a los cuerpos de agua El programa de
estabilizacioacuten de laderas impulsada por el gobierno (Albert 1900) y la ley forestal
de 1974 han fomentado la forestacioacuten con plantaciones de Pinus radiata y
Eucalyptus globulus y en el caso de Lago Vichuqueacuten estas actualmente cubren maacutes
del 60 de su cuenca (Cap2 Fig 5) Ademaacutes en Laguna Matanzas la
sobreexplotacioacuten del agua en conjunto con el cambio climaacutetico actual el que ha
conducido a una de las megasequiacuteas maacutes severas en las uacuteltimas deacutecadas
(Garreaud et al 2017) generoacute una combinacioacuten letal que secoacute el lago en tan solo
10 antildeos (Cap1 Fig1) Las evidencias obtenidas a partir de estos dos lagos
permiten identificar las huellas del Antropoceno en Chile central basadas en el
registro sedimentario lacustre
La transferencia de N desde los ecosistemas terrestres a acuaacuteticos es un
proceso natural con controles bioacuteticos climaacuteticos y geoloacutegicos El enlace
hidroloacutegico entre los ecosistemas terrestres y acuaacuteticos hace que el estado troacutefico
de los lagos esteacute vinculado al de su cuenca (McLauchlan et al 2013) En Chile
central se sabe poco del impacto de los LUCC sobre el ciclo de nutrientes en los
ecosistemas lacustres aunque al menos desde la colonizacioacuten espantildeola se tienen
registros de influencia humana en las cuencas Durante la colonia espantildeola
Laguna Matanzas fue una hacienda ganadera que exportaba productos caacuternicos al
Peruacute (Contreras-Lopez et al 2014) A su vez en el Lago Vichuqueacuten se realizaban
extensas actividades agriacutecolas hasta comienzos del s XX como por ejemplo
cultivos de vid y trigo ademaacutes de extraccioacuten de madera y mineriacutea de oro (Odone
1997) En las uacuteltimas deacutecadas los LUCC han estado vinculados principalmente con
el desarrollo forestal al compaacutes de una estrategia gubernamental de fomentar con
119
incentivos financieros (Ley de Bosques DFL nordm265 de 1931 y DFL 701 de 1974)
esta actividad El incremento de la superficie forestal es especialmente fuerte en
ambos lagos a partir de la deacutecada de 1980 y hasta 2016 (Laguna de Matanzas 0-
17 Lago Vichuqueacuten 1-66) Este aumento habriacutea sido en detrimento del bosque
nativo y los pastizales (Cap 1 Fig 5 y Cap 2 Fig 8) El incremento de la superficie
forestal generariacutea una disminucioacuten de los aportes de sedimentos y nutrientes al lago
y en este sentido un cambio de estado en los flujos de N (e g tipping points) que
a su vez ha quedado registrado en la sentildeal isotoacutepica (δ15N) y la acumulacioacuten de
MO en los sedimentos lacustres
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central
Los sedimentos lacustres son verdaderos ldquosumiderosrdquo que pueden llegar a
registrar lo que ocurre en cuanto a la historia ambiental de su cuenca En esta tesis
se utilizan los isoacutetopos estables de N (15N y 14N) en sedimentos lacustres para
reconstruir los cambios en la disponibilidad de N en el tiempo y establecer la
magnitud de impacto generado por actividades humanas El fraccionamiento
cineacutetico en los lagos es mediado por procesos bioloacutegicos que mediante la
asimilacioacuten de N genera un producto (eg fitoplancton) maacutes ldquoligerordquo (valores maacutes
bajos de 15N) que su remanente (eg DIN) (Fogel y Cifuentes 1993) Sin embargo
en periacuteodos en que los lagos estaacuten limitados por N el fitoplancton discrimina menos
y la MO resultante queda enriquecida en 15N (Fig 1 a y b) Ademaacutes la
desnitrificacioacuten acentuada en lagos de fondo anoacutexicos (eg Laguna Matanzas
entre 1800 y 1940 Cap1 Fig 6) conduce a un enriquecimiento en 15N de los
sedimentos y de la columna de agua Esta informacioacuten queda reflejada en la MO
120
de los sedimentos lacustres lo que permite conocer la disponibilidad del N en el
tiempo a partir de las variaciones de 15N
En esta tesis analizamos la MO de los sedimentos lacustres para reconstruir
la influencia de los LUCC en los aportes de sedimentos y N al lago En teacuterminos de
asimilacioacuten de N se puede distinguir entre dos grupos principales de productores
primarios que componen el POM (Fig1)
1 Los que asimilan el DIN compuesto por amonio nitratos y nitritos Aquiacute el
δ15N resultante fluctuaraacute en torno a valores maacutes positivos en la medida que
la productividad se incrementa o no hay reposicioacuten del DIN (Fig 1)
2 Los fijadores de N Dado que es un proceso energeacuteticamente costoso en
ambientes que no estaacuten limitados por N muchas veces son excluiacutedas
competitivamente por el resto del fitoplancton Si el DIN queda agotado por
el incremento de la productividad (o bien si el ambiente estaacute limitado ya sea
por N o por otros nutrientes) los fijadores de N pueden florecer y la MO que
se genera a partir de ello tendraacute un δ15N con valores en torno a 0permil
De este modo la MO en los sedimentos lacustres dependeraacute de la
composicioacuten de productores primarios (ie fijadores vs asimiladores del DIN)
ademaacutes de los aportes desde la cuenca tanto de MO como del N de la cuenca que
pasa a formar parte del DIN (eg fertilizantes estieacutercol) (Fig 1)
La MO de los lagos estudiados en esta tesis ha sido analizada a partir de
variables geoquiacutemicas isotoacutepicas frotis y estaacute compuesta principalmente por
diatomeas y MO amorfa A partir de los datos obtenidos en esta tesis los valores
de δ15N aumentan cuando hay una mayor cobertura agriacutecola o pastizales A su vez
tiende a disminuir cuando aumenta la cobertura arboacuterea independiente si esta es
por plantaciones forestales o por bosque nativo
121
Por otra parte la dinaacutemica del lago tambieacuten imprime caracteriacutesticas
especiales en el POM observaacutendose variaciones estacionales en los valores
δ15NPOM Durante la estacioacuten caacutelida y seca se observan valores maacutes bajos que
durante la estacioacuten friacutea y huacutemeda posiblemente relacionado con el incremento de
la contribucioacuten de especies fijadoras de N (Lago Vichuqueacuten Fig 9 Cap 2) durante
el verano En la estacioacuten friacutea y huacutemeda el DIN seriacutea la principal fuente de N Las
mayores entradas de MO y N terrestre debidos a un incremento del lavado de la
cuenca como consecuencia de las precipitaciones y la mineralizacioacuten de la MO
podriacutean estimular la produccioacuten de MO lacustre por crecimiento del fitoplancton
Como consecuencia se observan tendencias decrecientes de los valores de
δ15N en la MO sedimentaria cuando la productividad en el lago estaacute relacionada
con especies fijadoras de N mientras que el δ15N muestra aumentos cuando la
productividad del lago estaacute asociada principalmente al consumo del DIN pero
tambieacuten cuando predominan procesos de perdida de N (eg desnitrificacioacuten Fig
1)
Hemos identificado tres fases en la dinaacutemica del δ15N para ambos lagos
Entre 1800 y 1940 CE la cuenca de laguna de Matanzas estaba dominada por
actividades agriacutecolas y ganaderas (δ15Nsuelo gt 4permil Cap 1 Fig 4 y 6) Las entradas
de sedimentos y MO desde la cuenca son altas (δ13C lt -209permil ZrTi ~029 AlTi
~013) y coincide con un clima maacutes huacutemedo y friacuteo (Von Gunten et al 2009
Christiansen et al 2011) El lago teniacutea un nivel de agua maacutes profundo dominado
por condiciones anoacutexicas (MnFe ~0013) que contribuiriacutean a mayores valores de
δ15N en la columna de agua y por ende de los sedimentos acumulados en el fondo
debido a la peacuterdida diferencial de 14N viacutea desnitrificacioacuten (Fig 7 Cap1) La baja
122
produccioacuten de MO lacustre (BrTi ~002 TOC~17) coincide con un bajo contenido
de NT (~478 μg) y valores maacutes positivos de δ15N (gt 4permil)
La actividad agriacutecola tambieacuten dominaba en la cuenca del Lago Vichuqueacuten
durante este periacuteodo (δ15Nsuelo gt4permil Cap2 Fig 6) El aporte sedimentario desde la
cuenca es alto (AlTi ~029 ZrTi ~032) y fue favorecido por mayores
precipitaciones a las actuales (Christiansen et al 2011) El Lago Vichuqueacuten es un
lago estratificado anoacutexico (MnFe ~002) y profundo (~30 m) por lo que la
desnitrificacioacuten juega un rol importante en el enriquecimiento de la MO
sedimentaria La baja produccioacuten de MO (BrTi ~007 TOC ~12) de origen
lacustre (δ13C ~ -268permil) coincide con un bajo contenido de NT (~309μg +6) y
valores maacutes positivos de δ15N (56permil +03)
Durante esta fase en ambos lagos los aportes de N de la cuenca parecen
ser los principales responsables en las fluctuaciones del δ15N Esta sentildeal podriacutea
estar amplificada por bajas temperaturas (que afectan la productividad lacustre) y
altas precipitaciones que favorecen el lavado de la cuenca y aumentan el aporte de
sedimentos y MO desde la cuenca predominantemente agriacutecola
Entre 1940 y 1980 CE en Laguna Matanzas se observa un incremento en
la acumulacioacuten de MO algal (TOC ~2 +1 δ13C ~-230+09permil) El ambiente
deposicional es ligeramente menos turbulento que el anterior (AlTi ~011+001
ZrTi ~03+001) y el lago estaacute en una fase de transicioacuten hacia condiciones maacutes
oacutexicas (MnFe ~002+0004) y maacutes productivas (BrTi ~004+0007) A su vez δ15N
tiende hacia valores maacutes negativos (δ15N ~30permil+03) mientras que el NT aumenta
oscilando en antifase con el δ15N
En Lago Vichuqueacuten en cambio se observa un ligero incremento en la
acumulacioacuten de MO (TOC ~13 +02) de origen terrestre (δ13C ~-27permil +04) La
123
productividad es similar al periacuteodo anterior (BrTi ~007+003) y el ambiente
deposicional es menos turbulento (AlTi ~02 +009 Zrti ~033 +007) El δ15N y el
NT oscilan en torno a valores ligeramente maacutes bajos (δ15N ~51permil +04 NT ~292μg
+6) Este periacuteodo se caracteriza por ser maacutes caacutelido que el anterior lo que
posibilitariacutea un incremento en la productividad lacustre en Laguna Matanzas pero
que no es observada en el Lago Vichuqueacuten
Entre 1980 y la actualidad en Laguna Matanzas habriacutea un incremento de la
acumulacioacuten de MO (TOC ~59 + 3) asociada a un incremento en la productividad
del lago (BrTi ~009+005) y a los aportes de MO terrestres (δ13C ~-24 + 2permil) El
lago se vuelve maacutes oacutexico (Mn ~0003 +0006) y las entradas de sedimento
disminuyen (AlTi~ 009 +002) El δ15N tiende a valores maacutes negativos (δ15N ~13permil
+1) y el NT aumenta de 20 a 55 μg (NT ~346 + 9 μg) tomando valores sin
precedentes en todo el registro En los sedimentos lacustres del Lago Vichuqueacuten
tambieacuten se observa un incremento de la acumulacioacuten de MO (TOC ~17 + 05)
asociada a un incremento en la productividad del lago (BrTi ~01 + 003) y las
entradas de sedimento desde la cuenca disminuyen (AlTi ~005 + 005) El δ15N
(~43 + 05permil) tiende a valores maacutes negativos y el NT incrementa de 20 a 55 μg (NT
~346 + 9 μg) oscilando en antifase
Durante esta fase en ambos lagos se observa un aumento en la
acumulacioacuten de MO sedimentaria un incremento en el NT y valores maacutes negativos
de δ15N que coincide con el incremento de la superficie forestal de las cuencas
(Laguna Matanzas gt17 Lago Vichuqueacuten gt60)
124
Figura 2 Comparacioacuten de los cambios del ciclo del N entre Laguna Matanzas y
Lago Vichuqueacuten con los procesos biogeoquiacutemicos contrastantes en rojo En L
Matanzas las condiciones oacutexicas podriacutean estar favoreciendo la nitrificacioacuten del
amonio En Vichuqueacuten las condiciones anoacutexicas favorecen un aumento en δ15N de
la MO en los sedimentos por desnitrificacioacuten y mineralizacioacuten
Los ambientes mediterraacuteneos en el que los lagos del presente estudio se
encuentran insertos estaacuten caracterizados por una estacioacuten seca prolongada y las
precipitaciones ocurren en eventos puntuales alcanzando altos montos
pluviomeacutetricos en poco tiempo Las lluvias favorecen un lavado de la cuenca la
perdida de N y otros nutrientes Por lo que es esperable que los sedimentos del
lago reflejen mejor los valores isotoacutepicos de los suelos de la cuenca durante los
periacuteodos con altos montos pluviomeacutetricos En el Lago Vichuqueacuten por ejemplo el
125
POM superficial (2 y 5 m de profundidad) despliega valores isotoacutepicos maacutes
positivos en invierno presumiblemente como resultado de mayores aportes de MO
y N de su cuenca (Fig 1b) En ambos lagos los valores maacutes positivos en los
sedimentos lacustres ocurren durante periacuteodos con mayores montos pluviomeacutetricos
(Cap1 Fig 6 y Cap 2 Fig 12)
Ademaacutes del efecto de la precipitacioacuten en el aporte de nutrientes al lago en
esta tesis hemos encontrado que los mayores aportes de N desde la cuenca se dan
cuando la cobertura vegetal del suelo es menor En Laguna Matanzas el desarrollo
de la actividad ganadera desde la llegada de los espantildeoles a la cuenca habriacutea
incrementado los aportes de N al lago Los valores de δ15N en los sedimentos
lacustres depositados durante este periacuteodo son los maacutes positivos de todo el registro
(~4permil) A su vez en Lago Vichuqueacuten los valores isotoacutepicos maacutes positivos se
registran entre 1300 y 1900 CE que coinciden con el mayor desarrollo de
actividades agriacutecola y de extraccioacuten de maderas de la cuenca (Fig 10 Cap 2)
Los recientes LUCC (a partir de 1975) en las cuencas de Laguna Matanzas
y Lago Vichuqueacuten estaacuten caracterizados por un incremento de la superficie forestal
y agriacutecola en reemplazo de la cobertura de pastizal (Laguna Matanzas) y bosque
nativo (Lago Vichuqueacuten) Los valores de δ15N en los testigos sedimentarios fueron
maacutes positivos cuanto mayor la superficie agriacutecola de la cuenca y maacutes negativos
cuanto mayor la superficie forestal Aunque por los alcances de esta tesis no
podemos ser concluyentes respecto del origen del NT (aloacutectonoautoacutectono) parece
ser que esta maacutes ligado a los aportes de la cuenca pese a la disminucioacuten del aporte
sedimentario observado en ambos lagos Las plantaciones forestales a diferencia
del bosque nativo son fertilizadas con una mezcla de N P y K (Toral et al 2005)
Por lo que pese a la disminucioacuten del aporte sedimentario la concentracioacuten de
126
nutrientes en el aporte al lago puede verse incrementada por el tipo de manejo
forestal con respecto al bosque nativo
Los resultados del primer capiacutetulo demuestran que 1) las plantaciones
forestales yo bosque nativo retiene maacutes sedimentos y MO mientras que el uso de
suelo agriacutecola y los pastizales destinados al desarrollo de la actividad ganadera lo
libera 2) Al parecer la desnitrificacioacuten es el proceso natural maacutes importante de
perdida de N en lagos costeros y favorece el enriquecimiento de 15N tanto en la
columna de agua (ie 15N-DIN) como en los sedimentos una vez enterrados y 3) la
desnitrificacioacuten depende de los cambios en las condiciones REDOX en el lago La
oxigenacioacuten en lagos poco profundos -como Laguna Matanzas- es altamente
fluctuante a escalas decadal-centenal depende de la profundidad de la laacutemina de
agua y de la variabilidad de la precipitacioacuten En este sentido en Laguna Matanzas
habriacutea condiciones anoacutexicas en ambientes cuando el lago alcanza niveles maacutes
altos Estas condiciones se observan entre 1820 y 1940 CE y son coetaacuteneas con
episodios maacutes huacutemedos y friacuteos (von Gunten et al 2009 Christie et al 2009) pero
tambieacuten con una fuerte actividad ganadera en la cuenca
Las fuentes variables de N y su fluctuacioacuten en el tiempo hacen necesario
contar con un anaacutelogo moderno que permite usar al δ15N de los sedimentos
lacustres como un indicador indirecto de los cambios en la disponibilidad de N en
el tiempo Por ello en el segundo capiacutetulo integramos muestras de suelo-
vegetacioacuten y un monitoreo de POM de la columna de agua en verano e invierno La
composicioacuten isotoacutepica de POM estariacutea reflejando cambios de la fuente de N (fijacioacuten
vs consumo del DIN) que variacutea durante el antildeo hidroloacutegico Durante el verano la
mayor temperatura y horas-luz favoreceriacutean las entradas de N al lago viacutea fijacioacuten
bioloacutegica de N2 atmosfeacuterico resultando en un incremento de la MO con un valor
127
isotoacutepico bajo (POM verano~5permil Fig 1b) El N2 (δ15N = 0permil) es fijado praacutecticamente
sin fraccionamiento isotoacutepico generando un δ15N de la OM cercano a 0permil (Leng et
al 2006) En invierno las precipitaciones pueden estar favoreciendo un incremento
en la entrada de nutrientes al lago El DIN de la columna de agua llega a contener
valores de δ15N maacutes altos reflejando estos mayores aportes de la cuenca y el POM
del Lago Vichuqueacuten queda enriquecido en 15N (δ15N ~11permil Cap 2 Fig 9) Estas
variaciones estacionales pueden estar evidenciando un cambio en la composicioacuten
de especies de POM desde especies fijadoras a especies que consumen el N de la
columna de agua Sin embargo para corroborar esta informacioacuten seriacutea deseable
contar con el anaacutelisis de la composicioacuten de especies de las muestras de agua
extraidas
Entre los resultados maacutes importantes estaacuten los valores isotoacutepicos del suelo
y la biomasa representativa de la cuenca que incluye un listado de las especies
nativas de la zona que hasta ahora no se contaba (Cap 2 Tabla en material
suplementario) Las especies colectadas despliegan valores de δ15Nfoliar maacutes
positivos (plantaciones forestales y pastizal ~89permil) que el suelo (35permil) salvo por
las especies nativas las que oscilan en torno a valores cercanos al atmosfeacuterico
(~14permil) La razoacuten isotoacutepica 15N14N refleja la dinaacutemica de intercambio de N entre la
vegetacioacuten y el suelo (Koba et al 2002) La discriminacioacuten es positiva en la mayoriacutea
de los sistemas bioloacutegicos por lo que las plantas tienen valores de δ15N maacutes bajos
que el suelo (Evans et al 2001) como sucede con las especies nativas en Lago
Vichuqueacuten En consecuencia las plantas quedan empobrecidas en 15N y conducen
a un enriquecimiento en 15N del suelo sobretodo si no hay reposicioacuten de N por otras
viacuteas (eg fijacioacuten de N) Por lo tanto los valores maacutes negativos de δsup1⁵Nfoliar de las
especies nativas pueden estar relacionados con el consumo preferencial de 14N del
128
suelo donde la discriminacioacuten isotoacutepica de la vegetacioacuten contra el 15N conduce a
valores del δsup1⁵Nsuelo maacutes altos (Houmlgberg 1997) Por el contrario los valores maacutes
positivos de δsup1⁵Nfoliar de las plantaciones forestales y pastos con respecto al suelo
puede deberse por una parte que el suelo no cuenta con mecanismos naturales de
reposicioacuten de N salvo por la descomposicioacuten de la hojarasca que puede ser maacutes
lenta en Pinus radiata que en bosque nativo (Lusk et al 2001) y por otra por el alto
impacto de los aportes de N (y otros nutrientes) derivado de las actividades
humanas (eg uso de fertilizantes) en el suelo
El alcance maacutes significativo de esta tesis se relaciona con un cambio en la
tendencia maacutes negativa de δsup1⁵N y el incremento del NT a partir de 1940 y a partir
de 1970 CE en ambos lagos La causa maacutes probable de esta tendencia es el
reemplazo de la cobertura natural o preexistente a 1940 CE por plantaciones
forestales
En la figura 2 se observa una siacutentesis de los principales procesos que
afectan la sentildeal isotoacutepica de la MO en los sedimentos lacustres de L Matanzas y
L Vichuqueacuten La entrada de N y MO desde la cuenca depende del tipo de cobertura
Las plantaciones forestales y el bosque nativo retienen maacutes nutrientes y sedimentos
en la cuenca mientras que la agricultura los favorece Sin embargo las praacutecticas
de manejo que incluyen la aplicacioacuten de N (y otros nutrientes) pueden aportar maacutes
nutrientes al lago que la cobertra de bosque nativo Cuando las actividades
forestales cubren una mayor superficie de la cuenca (1940 en adelante) δsup1⁵N oscila
en valores maacutes negativos y el NT tiende a incrementarse en los sedimentos
lacustres Cabe destacar que los valores de δsup1⁵N observados en los testigos
sedimentarios a fines del s XX no tienen precedente durante la conquista y colonia
espantildeola o durante el resto del periodo de la Repuacuteblica
129
Figura 2 Modelo de transferencia de N entre los ecosistemas terrestres y
acuaacuteticos El δ15N de los sedimentos lacustres depende de la interaccioacuten entre los
aportes de la cuenca y porcesos ldquoin-lakerdquo Flechas indican la direccioacuten del flujo de
N En azul las salidasentradas de 14N en rojo las salidasentradas de 15N δ15N de
la interaccioacuten plantas-suelo depende del tipo de cobertura de suelo
130
CONCLUSIONES GENERALES
La transferencia de N entre cuencas y lagos es un factor de control del ciclo
del N en los lagos y queda reflejado en la sentildeal isotoacutepica de los sedimentos
lacustres (Leng et al 2006 McLauchlan et al 2007) En Laguna Matanzas el
suelo de las especies nativas y las plantaciones forestales despliegan valores de
δ15N maacutes bajos (~1permil) que los de Lago Vichuqueacuten (~35permil) Coincidentemente los
sedimentos lacustres de Laguna Matanzas oscilan con valores de δ15N maacutes bajos
(-15 a +45permil Fig 1) que los de Lago Vichuqueacuten (+3 a +8permil)
Por otra parte el estudio de la MO de los sedimentos lacustres ha permitido
reconstruir los cambios en los aportes de MO y N desde la cuenca Aunque no es
posible afirmar queacute tipo de N estaacute entrando al lago (eg Nitroacutegeno orgaacutenico e
inorganico) si podemos decir que los cambios en las fluctuaciones de δ15N son
coincentes con el crecimiento de la actividad forestal (1980 - hasta 2016 L
Matanzas 0-17 y L Vichuqueacuten 1-66) El δ15N fluctuacutea hacia valores maacutes
negativos Ademaacutes se observa un incremento del NT en los sedimentos lacustres
cuanto mayor es la superficie forestal
Entre 1800-1940 las cuencas eran fundamentalmente agriacutecolas y
ganaderas (Cap1 Fig 7 y Cap 2 Fig 12) el δ15N de los sedimentos lacustres
oscila con valores maacutes positivos (L Matanzas +40 plusmn 05permil y L Vichuqueacuten 56 plusmn
033permil Fig 1) hay una baja bioproductividad en el lago (BrTi ~002 TOC~17)
lo que coincide con un bajo contenido de NT (~478 μg) Este es un periacuteodo de altas
precipitaciones (Christie et al 2009) que tienden a favorecer el lavado de la cuenca
131
y con ello el aporte de MO sedimentos y nutrientes desde la cuenca tiende a verse
favorecido Aunque las principales actividades humanas en estas cuencas son
diferentes (ganaderiacutea en Laguna Matanzas Contreras-Lopez et al 2014
agricultura en Lago Vichuqueacuten Odone 1997) en ambos casos hubo un reemplazo
de la cobertura natural de bosques nativo que favorecioacute mayores aportes de N y
sedimentos desde la cuenca en un efecto sumado con el aumento de las
precipitaciones
A partir de 1980 CE en ambos lagos se observa una disminucioacuten en los
valores de δ15N y un incremento del NT que no tiene precedentes en todo el registro
y pese a que ambos lagos son limnologicamente muy diferentes En Lago
Vichuqueacuten el δ15N tiende a oscilar en antifase con el NT (Fig 1) En el caso de
Laguna Matanzas se observa una tendencia hacia valores maacutes negativos y a partir
de 1990s a valores maacutes positivos mientras que el NT tiende a incrementarse de
manera sostenida Ello podriacutea estar relacionado con el crecimiento de la actividad
forestal habriacutea una mayor retencioacuten de N y sedimentos en la cuenca ligado al
incremento de la superficie forestal Ademaacutes en el caso de L Matanzas el
incremento de la actividad agriacutecola (sumado al de las plantaciones forestales)
podriacutea expicar el cambio en la tendencia en la deacutecada de los 1990s
En el contexto de Antropoceno esta tesis nos permite identificar un gran
impacto de las actividades humanas en Chile central a partir de la deacutecada de 1940
y una Gran Aceleracoacuten a partir de la deacutecada de 1970s En el registro sedimentario
de los lagos en estudio se observa una gran acumulacioacuten de MO el δ15N oscila
hacia valores maacutes negativos y un gran incremento del NT y el crecimiento de la
actividad forestal parece ser la principal responsable Ademaacutes la sobreexplotacioacuten
132
del agua junto al cambio climaacutetico global ha generado una combinacioacuten letal para
los lagos costeros de Chile central
Figura 1 Comparacioacuten de las fluctuaciones de δ15N durante los uacuteltimos 300
antildeos en Laguna Matanzas y Lago Vichuqueacuten
133
Referencias
Battye W Aneja VP Schlesinger WH 2017 Earthrsquos Future Is nitrogen the next carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia
UC de V 2014 Elementos de la historia natural del An Mus Hist Natulas Vaplaraiso 27 51ndash67
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Evans RD Evans RD 2001 Physiological mechanisms influencing
plant nitrogen isotope composition 6 121ndash126 Fogel ML Cifuentes LA 1993 Isotope fractionation during primary
production Org Geochem 73ndash98 httpsdoiorg101007978-1-4615-2890-6_3 Galloway JN Schlesinger WH Ii HL Schnoor L Tg N 1995
Nitrogen fixation Anthropogenic enhancement-environmental response reactive N Global Biogeochem Cycles 9 235ndash252
Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J
Christie D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Koba K Koyama LA Kohzu A Nadelhoffer KJ 2003 Natural 15 N
Abundance of Plants Coniferous Forest httpsdoiorg101007s10021-002-0132-6
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos
Transactions American Geophysical Union httpsdoiorg1010292007EO070007 McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE
2007 Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
134
Phytologist N Oct N 2006 Tansley Review No 95 15N Natural Abundance in Soil-Plant Systems Peter Hogberg 137 179ndash203
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Turner BL Kasperson RE Matson PA McCarthy JJ Corell RW
Christensen L Eckley N Kasperson JX Luers A Martello ML Polsky C Pulsipher A Schiller A 2003 A framework for vulnerability analysis in sustainability science Proc Natl Acad Sci U S A httpsdoiorg101073pnas1231335100
Vitousek PM Aber JD Howarth RW Likens GE Matson PA
Schindler DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
Vitousek PM Howarth RW 1991 Nitrogen limitation on land and in the
sea How can it occur Biogeochemistry httpsdoiorg101007BF00002772 von Gunten L Grosjean M Rein B Urrutia R Appleby P 2009 A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 Holocene 19 873ndash881 httpsdoiorg1011770959683609336573
Xu R Tian H Pan S Dangal SRS Chen J Chang J Lu Y Maria
Skiba U Tubiello FN Zhang B 2019 Increased nitrogen enrichment and shifted patterns in the worldrsquos grassland 1860-2016 Earth Syst Sci Data httpsdoiorg105194essd-11-175-2019
Zaehle S 2013 Terrestrial nitrogen-carbon cycle interactions at the global
scale Philos Trans R Soc B Biol Sci 368 httpsdoiorg101098rstb20130125
3
TABLA DE CONTENIDO
RESUMEN 9
ABSTRACT 11
INTRODUCCIOacuteN 13
Los ecosistemas mediterraacuteneos y el ciclo del N 15
Los lagos como sensores ambientales 16
El ciclo del N en lagos 18
Reconstruyendo el ciclo del N a partir de variaciones en δ15N 20
Referencias 25
CAPIacuteTULO 1 A COMBINED APPROACH TO ESTABLISHING THE TIMING AND MAGNITUDE OF ANTHROPOGENIC NUTRIENT ALTERATION IN A MEDITERRANEAN COASTAL LAKE- WATERSHED SYSTEM 29
Abstract 30 1 INTRODUCTION 31 2 STUDY SITE 35 3 RESULTS 38
31 Age Model 38 32 The sediment sequence 39 33 Sedimentary units 41 34 Isotopic signatures 42 35 Recent land use changes in the Laguna Matanzas watershed 44
4 DISCUSSION 45 41 N and C dynamics in Laguna Matanzas 45 42 Recent evolution of the Laguna Matanzas watershed 48
5 CONCLUSIONS 54 6 METHODS 55
References 58
CAPIacuteTULO 2 STABLE ISOTOPES TRACK LAND USE AND COVER CHANGES IN A MEDITERRANEAN LAKE IN CENTRAL CHILE OVER THE LAST 600 YEARS 71
Abstract 73 1 Introduction 73
4
2 Study Site 77 3 Methods 79 4 Results 84
41 Geochemistry and PCA analysis 84 42 Sedimentary units 86 43 Recent seasonal changes of particulate organic matter on water column 88 44 Stable isotope values across the Lake Vichuqueacuten watershed 89 45 Land use and cover change from 1975 to 2014 90
5 Discussion 93 51 Seasonal variability of POM in the water column 93 52 Stable isotope signatures in the Lake Vichuqueacuten watershed 95 53 Recently land use and cover change and its influences on N inputs to the lake 97 54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years 98
6 Conclusions 101
LITERATURE CITED 110
DISCUSION GENERAL 117
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central 119
CONCLUSIONES GENERALES 130
Referencias 133
5
A mis padres Arturo y Malena
A mis hijos Xavi y Panchito
6
AGRADECIMIENTOS
Quiero agradecer a mi tutor y mentor Dr Claudio Latorre por brindarme su
apoyo sin el cual no habriacutea logrado concluir esta tesis de doctorado Claudio tu
apoyo constante incentivo y el fijarme metas que a veces me pareciacutean imposibles
de alcanzar no solo han dado forma a esta tesis sino tambieacuten me ha hecho maacutes
exigente como cientiacutefica Claudio destacas no solo por ser un gran cientiacutefico si no
tambieacuten por tu gran calidad humana eres un gran ejemplo
Quiero agradecer Dr Blas Valero-Garceacutes por nuestras numerosas
conversaciones viacutea Skype que incluiacutean vacaciones y fines de semana para discutir
los resultados de la tesis y que han dado forma a esta investigacioacuten principalmente
al primer capiacutetulo Ademaacutes por haberme acogido como un miembro maacutes en el
laboratorio de Paleoambientes Cuaternarios durante las estancias que he realizado
en el transcurso de estos antildeos Blas eres un ejemplo para miacute conjugas ciencia de
calidad calidez y dedicacioacuten por tus estudiantes
A quienes han financiado mi doctorado la Comisioacuten Nacional de
Investigacioacuten Cientiacutefica y Tecnoloacutegica (CONICYT) con sus becas de manutencioacuten
doctoral (2013) gastos operacionales pasantiacutea (2016) postnatal (2017) y termino
de tesis doctoralrdquo (2013) A FONDECYT a traveacutes del proyecto 1160744 de C
Santoro Al Departamento de InvestigacioacutenAl Instituto de Ecologiacutea y Biodiversidad
(IEB) a traveacutes de del PIA financiamiento basal 170008 la Pontificia Universidad
Catoacutelica de Chile por la beca incentivo para tesis interdisciplinaria para doctorandos
(2015)
Agradezco a mis compantildeeros del laboratorio de Paleoecologiacutea y
Paleoclimatologiacutea Karla Matias Dani Carolina Mauricio y Pancho que han hecho
grato mi tiempo en el laboratorio Agradecimientos especiales a Carolina Matiacuteas y
Leo por acompantildearme a terreno
7
Agradezco a los miembros del laboratorio de Paleoambientes Cuaternarios
(IPE) en especial a Raquel Elena Fernando Ana Penelope Graciela Miguel
Sevilla Mariacutea y Miguel Bartolomeacute
Agradezco a los miembros de la comisioacuten Dr Joseacute Miguel Farintildea Dr Juan
Armesto y Dr Ricardo De Pol-Holz por la gran ayuda durante el desarrollo de mi
doctorado en especial por las correcciones finales de la tesis
Al departamento de Ecologiacutea en especial a Rosario Galaz Edita Luengo
Daniela Mora y Valeria Cavallero por su apoyo
A mis compantildeeros de doctorado Beleacuten Gallardo Pancho Diacuteaz Bryan Bularz
Bian Latorre Renato Borras Juan Carlos Hernaacutendez y Paulina Troncoso con
quienes estudieacute aprendiacute y compartimos muy buenos momentos durante los
primeros antildeos del doctorado
A Pablo Sarricolea mi compantildeero de vida Gracias por tu constante e
incondicional apoyo Sin ti durante este proceso todo habriacutea sido cuesta arriba
A mis padres Arturo y Malena gracias por su carintildeo apoyo y compromiso
Papaacute aunque no esteacutes a mi lado siempre te recuerdo con mucho carintildeo A mi
madre que ha sido un constante apoyo sobre todo en el cuidado de mis hijos xavi
y panchito
A mis hermanos Rodrigo y David por estar presentes durante toda esta
etapa Siempre con carintildeo y hermandad
A Rodrigo David Matiacuteas Jany Paola y Julio por cuidar a mis hijos con carintildeo
siempre que estuve ausente por el doctorado
8
ABREVIATURAS
N Nitroacutegeno (Nitrogen)
DIN Nitroacutegeno Inorgaacutenico Disuelto (Dissolved Inorganic Nitrogen)
C Carbono (Carbon)
TOC Carbono Inorgaacutenico Total (Total Organic Carbon)
TIC Carbono Inorgaacutenico Total (Total inorganic Carbon)
TC Carbono Total (Total Carbon)
TS Azufre Total (Total Sulfur)
LUCC Cambios de UsoCobertura del Suelo (Land Use and Cover Change)
OM Materia Orgaacutenica (Organic Matter)
POM Particulate Organic Matter (materia orgaacutenica particulada)
CE Common Era
BCE Before Common Era
Cal BP Calibrado en antildeos radiocarbono antes de 1950
ie id est (esto es)
e g Exempli gratia (por ejemplo)
9
RESUMEN
El ldquoAntropocenordquo se caracteriza por pertubaciones de origen humano que
conllevan impactos globales Por ejemplo los cambios de usocobertura del suelo
(LUCC) son conocidos por perturbar el ciclo del N a partir de la Revolucioacuten Industrial
pero especialmente desde la Gran Aceleracioacuten (1950 CE) en adelante Sin
embargo existen incertezas asociadas a la magnitud del impacto y su efecto
acoplado con los cambios climaacuteticos actuales (ie megasequiacutea disminucioacuten de las
precipitaciones) y acoplamientos con otros ciclos biogeoquiacutemicos (eg ciclo del
Carbono) Este impacto ha cambiado la disponibilidad de N en los ecosistemas
terrestres y acuaacuteticos y sus enlaces Los sedimentos lacustres contienen
informacioacuten de las condiciones paleoambientales del lago y su cuenca en el
momento en que se depositaron y el anaacutelisis de los isoacutetopos estables de N (δ15N)
en los sedimentos permiten reconstruir los cambios en la disponibilidad de N a
traveacutes del tiempo En este trabajo usamos una aproximacioacuten multiproxy que incluye
10
anaacutelisis sedimentoloacutegicos geoquiacutemicos e isotoacutepicos (δ15N δ13C) de los sedimentos
lacustres columna de agua y suelo-vegetacioacuten de la cuenca y la reconstruccioacuten de
los LUCC entre 1975 y 2016 a partir de imaacutegenes satelitales El objetivo de esta
tesis fue evaluar el rol de LUCC como principal control del ciclo del N en el sistema
cuenca-lago durante las uacuteltimas centurias en Chile central Nuestros principales
resultados muestran que valores maacutes positivos de δ15N en los sedimentos lacustres
estaacuten relacionados con grandes aportes de N de la cuenca que a su vez son
mayores cuanto mayor la superficie agriacutecola yo los pastizales mientras que tanto
las plantaciones forestales como los bosques favorecen la retencioacuten de nutrientes
en la cuenca (lo que se ve reflejado en un δ15N maacutes negativo) Esta tesis plantea
un cambio de estado en la dinaacutemica del N asociado a la introduccioacuten y expansioacuten
de las plantaciones forestales o bosques los que retienen el Nitroacutegeno en las
cuencas mientras que el clima juega un rol secundario
11
ABSTRACT
The Anthropocene is characterized by human disturbances at the global
scale For example changes in land use are known to disturb the N cycle since the
industrial revolution but especially since the Great Acceleration (1950 CE) onwards
This impact has changed N availability in both terrestrial and aquatic ecosystems
However there are some important uncertainties associated with the extent of this
impact and how it is coupled to ongoing climate change (ie megadroughts rainfall
variability and shifting stationarity) or to other nutrient cycles (e g carbon cycle)
Lake sediments contain paleoenvironmental information regarding the conditions of
the watershed and associated lakes and which the respective sediments are
deposited Nitrogen stable isotope analysis (δ15N) on lake sediments allows us to
reconstruct the changes in N availability through time Here we used a multiproxy
approach that uses sedimentological geochemical and isotopic analyses on
lacustrine sediments water column and soilvegetation from the watershed as well
12
as land use and cover change (LUCC) from 1975 to 2016 obtained from satellite
images The goal of this thesis is to evaluate the role of LUCC as the main driver for
N cycling in a coastal watershed system of central Chile over the last centuries Our
main results show that more positive δ15N values in lake sediments are related to
higher N contributions from the watershed which in turn increase with increased
agricultural andor pasture cover whereas either forest plantations or native forests
can favor nutrient retention in the watershed (δ15N more negative) This thesis
proposes that N dynamics are mainly driven by the introduction and expansion of
forest or tree plantations that retain nitrogen in the watershed whereas climate plays
a secondary role
13
INTRODUCCIOacuteN
El N es un elemento esencial para la vida y limita la productividad en
ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al 1997) Las actividades
humanas han tenido un profundo impacto sobre el ciclo del N global principalmente
a partir la revolucioacuten industrial (IPCC 2013 2007) Las concentraciones de N se
han incrementado por la fijacioacuten industrial de N atmosfeacuterico (viacutea el proceso Haber-
Bosch Erisman et al 2008) por el uso de fertilizantes sobre la base de N para
mejorar el rendimiento de los cultivos la produccioacuten ganadera y ademaacutes de los
cambios en el uso del suelo (abreviado en ingleacutes- LUCC) (Robertson y Vitousek
2009 Galloway et al 2008 Battye et al 2017 Xu et al 2019) Estas actividades
contribuyen significativamente al incremento de la disponibilidad bioloacutegica de N
cuyas consecuencias para los ecosistemas incluye la perdida de diversidad
modificacioacuten de la disponibilidad de nutrientes y acidificacioacuten de los suelos entre
otras Sin embargo se desconoce la cantidad destino y el alcance que ha tenido
14
el N movilizado entre los ecosistemas generado por la influencia de las actividades
humanas (Vitousek et al 1997)
La entrada natural de N en los ecosistemas terrestres y acuaacuteticos es viacutea
fijacioacuten atmosfeacuterica la que es llevada a cabo por microorganismos muchos de ellos
en relaciones simbioacuteticas (eg Rhizobium en leguminosas) y las algas (Vitousek et
al 1997 Battye et al 2017) Las principales salidas estaacuten relacionadas con la
desnitrificacioacuten volatilizacioacuten del amoniaco y por escurrimiento superficial y
subterraacuteneo (Galloway et al 2008 Diacuteaz et al 2016) En el caso de los sistemas
lacustres ademaacutes estaacuten la resuspencioacuten de N del fondo del lago los aportes desde
la cuenca (entradas de N) La depositacioacuten de MO en el fondo del lago que marca
la salida de N de la columna de agua Estas relaciones de intercambio de N tienen
un control geograacutefico (eg latitud exposicioacuten relieve etc) climaacutetico y biotico
(McLauchlan et al 2013) La alteracioacuten provocada por los LUCC no solo genera
las peacuterdidas de nutrientes del suelo por erosioacuten y escurrimiento superficial sino que
tambieacuten modifica la disponibilidad y transferencia de N entre los ecosistemas
terrestres y acuaacuteticos (Elbert et al 2012 McLauchlan et al 2013) Por ejemplo el
reemplazo de la vegetacioacuten nativa por la agricultura yo plantaciones forestales
altera el ciclo del N debido al despeje de los terrenos viacutea quema de la vegetacioacuten
pero tambieacuten debido al reemplazo de la vegetacioacuten preexistente por un
monocultivo (eg trigo pino) y el uso ineficiente de fertilizantes para mejorar el
rendimiento agriacutecola Estas actividades favorecen un incremento de los aportes de
N (y otros nutrientes) viacutea sistemas de drenajes a lagos los que actuacutean como
sumideros El incremento del N derivado de las actividades humanas tanto en los
ecosistemas terrestres como acuaacuteticos genera incertezas relacionados con la
trayectoria y la magnitud de la disponibilidad de N en ambos ecosistemas (Batye et
15
al 2017 Mclauchlan et al 2017) Por lo tanto se requiere de una liacutenea base de
largo plazo para saber coacutemo estas perturbaciones impactan la disponibilidad de N
en las uacuteltimas deacutecadassiglos en los ecosistemas lacustres y por lo tanto el alcance
real que los LUCC han tenido en el ciclo del N
Los ecosistemas mediterraacuteneos y el ciclo del N
Los ecosistemas mediterraacuteneos son especialmente sensibles a los LUCC
pues estos se encuentran sometidos a un estreacutes hiacutedrico durante largas temporadas
estivales y las precipitaciones se concentran en eventos puntuales y a veces con
altos montos pluviomeacutetricos Las precipitaciones tienen un alto potencial de arrastre
de sedimentos y nutrientes los que actuacutean como fertilizantes al llegar a los
ecosistemas lacustres A su vez un incremento en la disponibilidad de N puede
generar un desbalance con otros nutrientes (eg Fosforo) con efectos sobre la
productividad y diversidad de los lagos (Pentildeuelas et al 2012 Sardans et al 2012
McLauchlan 2013) En este sentido en Chile central se observa una disminucioacuten
de las precipitaciones especialmente fuerte en las uacuteltimas deacutecadas por lo que se ha
denoacuteminado como Megasequiacutea (Garreaud et al 2017) y el efecto de las
precipitaciones esporaacutedicas pueden generar un arrastre de nutrientes que no ha
sido evaluado
Los ecosistemas mediterraacuteneos concentran el 20 de la biodiversidad global
(Myers et al 2000) pero existe una escasez de conocimiento respecto a los
efectos del incremento de N en los cuerpos de agua como consecuencia de las
actividades humanas en el pasado histoacuterico (Ochoa-Hueso et al 2011) y coacutemo la
disponibilidad de N ha variado en el tiempo Los LUCC han aumentado el flujo de
N en las cuencas hidrograacuteficas viacutea escurrimiento superficial y lixiviacioacuten
16
favoreciendo la peacuterdida de sedimentos y nutrientes del suelo en el mundo entero
(Elser et al 2011 McLauchlan et al 2013 Xu et al 2017 y 2019) Ello ha
contribuido a la acidificacioacuten y eutrofizacioacuten generalizada de riacuteos y lagos
(McLauchlan et al 2013 Schindler et al 2008)
El ecosistema mediterraacuteneo de Chile central es una regioacuten ampliamente
intervenida desde al menos la colonizacioacuten espantildeola (1600-1800 CE) Los LUCC
han tenido efectos negativos en la disponibilidad de agua especialmente
observados con el desarrollo de las actividades minera (Prieto et al 2019) Aunque
se sabe de los impactos en los ecosistemas de bosque en teacuterminos de cobertura
debidos al desarrollo silvo-agropecuarias (Armesto et al 2010) se desconoce el
impacto de estas actividades sobre el ciclo del N en los cuerpos de agua o en queacute
momento cobran fuerza suficiente para alterar el flujo de N hacia los lagos de Chile
Central Por lo tanto en esta tesis nos centramos en entender coacutemo los LUCC han
afectado la disponibilidad de N en los lagos costeros de Laguna Matanzas y Lago
Vichuqueacuten desde la llegada de los espantildeoles a Chile hasta el presente
Los lagos como sensores ambientales
Los sedimentos lacustres son buenos sensores de cambios en los aportes
de nutrientes contaminacioacuten y eutroficacioacuten de los cuerpos de agua ya que son
capaces de preservar sentildeales quiacutemicas y fiacutesicas al momento de su formacioacuten y
ademaacutes con alta resolucioacuten y a largo plazo (Leng et al 2006) Por lo tanto
constituyen una herramienta uacutetil para reconstruir el flujo de N entre ecosistemas
terrestres y acuaacuteticos en el pasado sus consecuencias (eg incremento de la
productividad lacustre) y el rol del clima y los LUCC en el pasado (McLauchlan et
al 2013 von Gunten et al 2009) Por ejemplo el cultivo de maiacutez por parte de los
17
nativos iroqueses en Canadaacute provocoacute la eutrofizacioacuten del Lago Crawford (43degN)
durante los siglos XII y XV Entre las consecuencias registradas se documentoacute un
claro incremento de la productividad primaria y cambios en la estructura comunitaria
de diatomeas foacutesiles (incremento de Stephanodiscus complex y disminucioacuten de
Cyclotella bodanica en Ekdahl et al 2007) Otro ejemplo demuestra como las
actividades agriacutecolas en la cuenca del riacuteo Mississippi modificaron los patrones de
sedimentacioacuten y aporte de nutrientes al lago Pepin EE UU (44degN) a partir del
asentamiento europeo en 1840 CE y principalmente desde 1940 CE (Engstrom et
al 2009) Para Chile von Gunten et al (2009) a partir de indicadores
limnogeoloacutegicos evidencioacute la contaminacioacuten y eutroficacioacuten de cinco lagos cercanos
a la ciudad de Santiago (33ordmS) como consecuencia de la deposicioacuten atmosfeacuterica
de metales pesados (especialmente Cu) quema de combustible foacutesil y flujo de
nutrientes durante los uacuteltimos 200 antildeos
Caracteriacutesticas limnoloacutegicas de los lagos
Los procesos limnoloacutegicos afectan la distribucioacuten y abundancia de los
organismos en los lagos Estaacuten influenciados por forzamientos externos por
ejemplo la precipitacioacuten o la penetracioacuten de la luz y la morfometriacutea del lago En este
sentido el nivel del lago dependeraacute del balance entre las entradas y salidas de agua
(precipitaciones escurrimiento superficial y subterraacuteneo y la evaporacioacuten) y la forma
de la cuenca (profundidad pendiente aacuterea del espejo de agua)
En funcioacuten de la profundidad de penetracioacuten de la luz es posible diferenciar
dos aacutereas que determinan la distribucioacuten de los organismos la zona foacutetica (donde
penetra la luz y permite la fotosiacutentesis) y la zona afoacutetica (subyacente a la zona
foacutetica) Al depender de la radiacioacuten la zona foacutetica variacutea estacionalmente Ademaacutes
18
puede variar por el grado de turbidez la morfometriacutea del lago y la produccioacuten de
materia orgaacutenica en la columna de agua
Otro factor que influye en la productividad es el reacutegimen de mezcla de la
columna de agua pues favorece la oxigenacioacuten y el reciclado de nutrientes La
mezcla ocurre en condiciones de gran inestabilidad usualmente relacionado con el
reacutegimen de viento Por el contrario un lago estratificado resulta de grandes
diferencias en temperatura o salinidad entre la superficie (epilimnion) y el fondo del
lago (hipolimnion) que separa las masas de agua superficial y de fondo por una
termoclina (si la estratificacioacuten estaacute relacionada con diferencias de temperatura de
las masas de agua) o haloclina (diferencias de salinidad) En funcioacuten del reacutegimen
de mezcla los lagos se pueden clasificar en (Lewis 1983)
1 Amiacutecticos no hay mezcla vertical de la columna de agua
2 Monomiacutecticos friacuteos y caacutelidos se mezcla una vez al antildeo
3 Dimiacutecticos la columna de agua se mezcla dos veces al antildeo
4 Oligomiacutecticos Lagos estratificados la mayor parte del tiempo pero se mezclan a
intervalos irregulares mayores a 1 antildeo
5 Polimiacutecticos friacuteos y caacutelidos la columna de agua se mezcla varias veces al antildeo
El ciclo del N en lagos
Al estar presente en aacutecidos nucleicos aminoaacutecidos y proteiacutenas el N es un
nutriente esencial de los organismos Por lo tanto su disponibilidad en la columna
de agua es clave para la produccioacuten composicioacuten y acumulacioacuten de la MO a traveacutes
del tiempo (Olson 1998 Talbot 2002) Aunque el principal reservorio de N estaacute en
19
la atmosfera (N2) solo unos pocos organismos (eg cianobacterias) pueden fijarlo
directamente Asiacute el Nitroacutegeno Inorgaacutenico Disuelto (DIN) representa la principal
fuente de N disponible para la produccioacuten primaria en los ecosistemas acuaacuteticos
(Talbot et al 2002) El DIN estaacute compuesto por nitrato (NO3-) nitrito (N02
-) y amonio
(NH4+) y los cambios en su disponibilidad condicionan el tipo de produccioacuten primaria
(eg fijadores vs asimiladores de N ver Scott y Grant 2013 Scott 2008)
La Figura 1 resume los principales componentes en lagos del ciclo del N y
sus interacciones La principal entrada natural de N es viacutea fijacioacuten de N atmosfeacuterico
y es llevado a cabo principalmente por bacterias y algas verde-azules capaces de
romper el triple enlace entre los dos aacutetomos de N a un alto costo energeacutetico (Torres
et al 2012) El N fijado es reducido a NH3 Tras la muerte de los organismos el N
es liberado de sus tejidos por hongos y bacterias implicados en la descomposicioacuten
de la MO Una parte se deposita en el fondo del lago y la otra queda disponible para
ser asimilada por el fitoplancton como amonio mediante el proceso de
amonificacioacuten (Talbot 2002) La amonificacioacuten resulta de la reduccioacuten bacteriana
del amoniaco y se da preferentemente en condiciones anoacutexicas La volatizacioacuten del
amoniaco es un proceso por medio este se incorpora a la atmoacutesfera Este proceso
se da preferentemente en lagos aacutecidos (pH gt 85) El nitrato es otra forma de N
bioloacutegicamente disponible para el fitoplancton En los lagos el NO3 del DIN estaacute
compuesto por los aportes desde la cuenca hidrograacutefica en la que se circunscriben
por la resuspensioacuten desde el fondo del lago favorecido por procesos de mezcla
(descrito en seccioacuten anterior) y por los nitratos de la columna de agua Estos uacuteltimos
son el resultado de la nitrificacioacuten proceso realizado por bacterias aeroacutebicas
mediante el cual el amonio se oxida para formar nitrito y nitrato El ciclo se completa
20
con la desnitrificacioacuten que es la reduccioacuten bacteriana de nitrato al N atmosfeacuterico
Este proceso se da preferentemente en condiciones anoacutexicas
Figura 1 Materia Orgaacutenica sedimentaria y su relacioacuten con el ciclo del N y las
variaciones de δ15N (elaborado a partir de Talbot et al 2002) En la figura se
representan los factores clave en la acumulacioacuten de la MO sedimentaria y su
relacioacuten con el ciclo del N El δ15N de la MO es resultado de los aportes de MO
desde la cuenca MO de macroacutefitas y de la productividad bioloacutegica La productividad
en ecosistemas lacustres estaacute condicionada por DIN y la fijacioacuten de N atmosfeacuterico
El δ15N del pool del DIN disponible se va enriqueciendo por la asimilacioacuten
preferencial de 14N por parte del fitoplancton Ademaacutes el DIN remanente se va
enriqueciendo por accioacuten microbiana (amonificacioacuten nitrificacioacuten desnitrificacioacuten)
Reconstruyendo el ciclo del N a partir de variaciones en δ15N
La sentildeal isotoacutepica de N (δ15N) en los sedimentos lacustres puede ser usada
para reconstruir los cambios pasados del ciclo N la transferencia de N entre
ecosistemas terrestres y acuaacuteticos y el estado troacutefico del lago (Bruland y Mackenzie
2015 McLauchlan et al 2013 Woodward et al 2012 von Gunten et al 2009
Leng et al 2006 Meyers y Teranes 2001) La Figura 1 resume los principales
procesos y fuentes que afectan la sentildeal isotoacutepica δ15N (total o ldquobulkrdquo) en la MO de
21
los sedimentos lacustres Entre los que destacan las fuentes de MO (aloacutectona vs
autoacutectona) la bioproductividad lacustre y las entradas de N (ie fijacioacuten bioloacutegica
de N2) (Torres et al 2012) Estos procesos conducen a un fraccionamiento
isotoacutepico cineacutetico que favorece la disociacioacuten entre sus isotopos ldquopesadosrdquo (15N) y
ldquoligerosrdquo (14N) Asiacute por ejemplo los valores de δ15N de la MO se enriquecen en 15N
en la medida que los sedimentos lacustres estaacuten sometidos a perdidas de 14N viacutea
desnitrificacioacuten (Bruland y Mackenzie 2015 Diaz et al 2016 Talbot et al 2002)
Por otra parte el fitoplancton en la medida que aumenta su biomasa (eg
durante la estacioacuten caacutelida) puede llegar a asimilar el DIN hasta agotarse En este
caso el N remanente se va empobreciendo en 14N si no hay reposicioacuten de N (eg
aporte de NO3 de la cuenca) y la MO queda enriquecida en 15N Esta disminucioacuten
induce a una limitacioacuten de fitoplancton por N y los organismos diztroacuteficos pueden
verse estimulados por lo que se incrementa la entrada de N viacutea fijacioacuten de N2 (Scott
y Grantsz 2013) como resultado la MO oscila en valores maacutes bajos (en torno a 0permil)
La cantidad de MO que se deposita en el fondo del lago depende del
predominio de contribuciones provenientes desde la cuenca (aloacutectona) versus las
producidas dentro del mismo lago (autoacutectona) (Torres et al 2012) Aunque en
general los lagos reciben permanentemente aportes de sedimentos y MO desde
su cuenca los lagos pequentildeos tienden a reflejar maacutes los procesos que ocurren
solamente en su cuenca directa y reflejan los valores isotoacutepicos de la misma (Gu et
al 2006) Woodward et al (2012) realizaron un estudio en 50 lagos en Irlanda que
les permitioacute correlacionar diferentes patrones de LUCC (bosques de coniacuteferas
agricultura ganaderiacutea entre otros) con los valores de δ15N (y δ13C) en los
sedimentos superficiales de los lagos Encontraron que los valores de δ15N son maacutes
negativos en cuencas poco impactadas y maacutes positivos en aquellas con alto
22
impacto humano (eg usos de suelo agriacutecola y ganadero) McLauchlan et al (2007)
encontraron valores maacutes positivos de δ15N en los sedimentos del Mirror Lake New
Hampshire (29ordm N) a partir de la desforestacioacuten de su cuenca en 1790 CE (inicio
del asentamiento europeo en el lago) para el desarrollo de la actividad agriacutecola
Estos valores se volvieron maacutes negativos hacia valores similares al pre-
asentamiento europeo despueacutes del cese de la agricultura en la cuenca y la
recuperacioacuten del bosque a partir de 1929 CE
El anaacutelisis de δ15N en los sedimentos lacustres es una herramienta casi sin
explorar en Chile central Proponemos reconstruir la dinaacutemica de transferencia de
N cuenca-lago como consecuencia de los LUCC desde la colonizacioacuten espantildeola en
los lagos costeros de Laguna Matanzas (34ordmS) y Lago Vichuqueacuten (36ordmS) Como son
muacuteltiples los procesos que pueden afectar la sentildeal isotoacutepica de N (medido como
δ15N) en los sedimentos lacustres existen muchos problemas para su
interpretacioacuten paleoambiental (Leng et al 2006) Por ello en esta tesis utilizamos
un enfoque multiproxy que consiste en integrar el anaacutelisis limnoloacutegico y geoquiacutemico
de los sedimentos lacustres con los valores isotoacutepicos modernos de la columna de
agua (mediante monitoreo del POM) la relacioacuten vegetacioacuten-suelo de la cuenca y la
reconstruccioacuten de los principales usos de suelo de las cuencas desde 1975 CE
mediante el uso de imaacutegenes satelitales Considerando estas muacuteltiples liacuteneas de
evidencia nos planteamos utilizar las variaciones del δ15N para reconstruir los
cambios en la disponibilidad de N en los lagos costeros de Chile central y establecer
coacutemo y desde cuaacutendo las actividades humanas en la cuenca son lo suficientemente
importantes como para modificar el ciclo del N en los lagos desde la colonizacioacuten
espantildeola (siglo XVII)
23
Como hipoacutetesis planteamos que la dinaacutemica de transferencia de sedimentos
y nutrientes entre la cuenca y el lago tiene controles geoloacutegicos climaacuteticos y
bioloacutegicos que interactuacutean en el tiempo registraacutendose en la acumulacioacuten de MO de
los sedimentos lacustres (Fig1) Bajo este marco los LUCC modifican esta
dinaacutemica alterando la transferencia de N a los lagos ya que las actividades agriacutecolas
y ganaderas tienden a aportar maacutes N (aumentando δ15N) que la actividad forestal
(la que disminuye δ15N)
En el primer capiacutetulo titulado ldquoA combined approach to establishing the timing
and magnitude of anthropogenic nutrient alteration in a mediterranean coastal lake-
watershed systemrdquo se evaluoacute el rol de los LUCC en el control de la descarga de N
y sedimentos a Laguna Matanzas en un sistema cuenca-lago desde el siglo XVII
Para ello se realizoacute un anaacutelisis de la dinaacutemica sedimentaria lacustre mediante el
anaacutelisis de muacuteltiples proxys sedimentoloacutegicos (ie minerales y textura)
geoquiacutemicos (eg geoquiacutemica orgaacutenica e isotoacutepica) y bioloacutegicos (ie diatomeas) de
Laguna Matanzas que cubren los uacuteltimos 200 antildeos Ademaacutes se realizoacute una
reconstruccioacuten de los usos de suelo entre 1975 y 2016 a partir de imaacutegenes de
sateacutelites y se colectaron muestras de suelo de las principales coberturas de la
cuenca a los cuales se midioacute el δ15N
Entre los principales resultados obtenidos se destaca la influencia de la
ganaderiacutea y la denitrificacioacuten lo que explica los altos valores de δ15N evidenciados
por el registro lacustre durante la colonizacioacuten espantildeola e inicios de la repuacuteblica A
partir de la Gran Aceleracioacuten (1950 CE) la desforestacioacuten y el reemplazo de la
ganaderiacutea por plantaciones forestales tienen un correlato en el registro
sedimentario con valores de δ15N maacutes bajos Estos resultados demuestran que los
LUCC son el factor de primer orden para explicar los cambios observados en
24
nuestro registro de δ15N y a su vez implica un rol secundario del clima como posible
control de la disponibilidad de N en Laguna Matanzas Este capiacutetulo fue sometido
a la revista Scientific Reports y estaacute en revisioacuten desde el 26 de marzo de 2019 En
la elaboracioacuten del mauscrito han colaborado Claudio Latorre Blas Valero-Garceacutes
Matiacuteas Frugone Pablo Sarricolea Santiago Giralt Manuel Contreras-Lopez
Ricardo Prego y Patricia Bernardez
El segundo capiacutetulo se titula ldquoStable isotopes track land use and cover
changes in a mediterranean lake in central Chile over the last 600 yearsrdquo Aquiacute
evaluamos la dinaacutemica del N actual en el sistema cuenca-lago considerando los
valores isotoacutepicos modernos tanto de la vegetacioacuten como del suelo ademaacutes de los
cambios estacionales del POM de la columna de agua Esta informacioacuten se utiliza
como un anaacutelogo moderno para conocer el rol de los LUCC en la disponibilidad de
N en los uacuteltimos 600 antildeos Para ello se colectaron muestras filtradas de la columna
de agua a 2 5 y 20 m dos veces por antildeo entre noviembre de 2015 y julio de 2018
y se obtuvo el δsup1⁵N del POM Ademaacutes se colectoacute muestras de vegetacioacuten y suelo
de la cuenca diferenciando entre especies nativas plantaciones forestales y
vegetacioacuten herbaacutecea Esta informacioacuten se analizoacute en conjunto con la reconstruccioacuten
de los LUCC entre 1975 y 2014 y el anaacutelisis limnoloacutegico del lago Ello nos permitioacute
evaluar coacutemo el δ15N de los sedimentos lacustres refleja los valores δ15N de la
cuenca en el Lago Vichuqueacuten y el control que ejerce el clima en la sentildeal isotoacutepica
de POM Este capiacutetulo seraacute sometido a la revista Science of total environmet
proximamente En la elaboracioacuten del mauscrito han colaborado Claudio Latorre
Blas Valero-Garceacutes Matiacuteas Frugone Pablo Sarricolea Leonardo Villacis y M Laura
Carrevedo
25
Entre los principales resultados encontramos que el δ15N en los sedimentos
lacustres es maacutes positivo con presencia de agricultura en la cuenca y maacutes negativo
cuando esta es reemplazada por cobertura arboacuterea bien sea plantaciones
forestales bien sea bosque nativo A su vez durante la estacioacuten seca y caacutelida la
mayor entrada al lago es viacutea fijacioacuten de N atmosfeacuterico (bajos valores de δ15NPOM)
Durante la estacioacuten huacutemeda en cambio prevalecen los aportes de la cuenca con
altos valores de δ15NPOM Ello estariacutea evidenciando un cambio estacional en la
composicioacuten de especies en verano ligado a especies fijadoras de N y en invierno
las algas y microorganismos que consumen el DIN de la columna de agua
Referencias
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea AM Marquet PA 2010 From the Holocene to the Anthropocene A historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the
next carbon Earth rsquo s Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409 httpsdoiorg102134jeq20090005
Diacuteaz FP Frugone M Gutieacuterrez RA Latorre C 2016 Nitrogen cycling in an
extreme hyperarid environment inferred from δ15 N analyses of plants soils and herbivore diet Sci Rep 6 1ndash11 httpsdoiorg101038srep22226
Ekdahl EJ Teranes JL Wittkop CA Stoermer EF Reavie ED Smol JP
2007 Diatom assemblage response to Iroquoian and Euro-Canadian eutrophication of Crawford Lake Ontario Canada J Paleolimnol 37 233ndash246 httpsdoiorg101007s10933-006-9016-7
Elbert W Weber B Burrows S Steinkamp J Buumldel B Andreae MO
Poumlschl U 2012 Contribution of cryptogamic covers to the global cycles of carbon and nitrogen Nat Geosci 5 459ndash462
26
httpsdoiorg101038ngeo1486 Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506
httpsdoiorg101126science1215567 ARTICLE Engstrom DR Almendinger JE Wolin JA 2009 Historical changes in
sediment and phosphorus loading to the upper Mississippi River Mass-balance reconstructions from the sediments of Lake Pepin J Paleolimnol 41 563ndash588 httpsdoiorg101007s10933-008-9292-5
Erisman J W Sutton M A Galloway J Klimont Z amp Winiwarter W (2008)
How a century of ammonia synthesis changed the world Nature Geoscience 1(10) 636
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892
httpsdoiorg101126science1136674 Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J Christie
D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592 Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
IPCC 2013 Climate Change 2013 The Physical Science BasisContribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change Cambridge University Press Cambridge United Kingdom and New York NY USA
httpsdoiorg101017CBO9781107415324 Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007 Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32iumliquestfrac12S 3050iumliquestfrac12miumliquestfrac12asl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470
27
httpsdoiorg101073pnas0701779104 McLauchlan KK Gerhart LM Battles JJ Craine JM Elmore AJ Higuera
PE Mack MC McNeil BE Nelson DM Pederson N Perakis SS 2017 Centennial-scale reductions in nitrogen availability in temperate forests of the United States Sci Rep 7 1ndash7 httpsdoiorg101038s41598-017-08170-z
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Myers N Mittermeler RA Mittermeler CG Da Fonseca GAB Kent J
2000 Biodiversity hotspots for conservation priorities Nature httpsdoiorg10103835002501
Pentildeuelas J Poulter B Sardans J Ciais P Van Der Velde M Bopp L
Boucher O Godderis Y Hinsinger P Llusia J Nardin E Vicca S Obersteiner M Janssens IA 2013 Human-induced nitrogen-phosphorus imbalances alter natural and managed ecosystems across the globe Nat Commun 4 httpsdoiorg101038ncomms3934
Prieto M Salazar D Valenzuela MJ 2019 The dispossession of the San
Pedro de Inacaliri river Political Ecology extractivism and archaeology Extr Ind Soc httpsdoiorg101016jexis201902004
Robertson GP Vitousek P 2010 Nitrogen in Agriculture Balancing the Cost of
an Essential Resource Ssrn 34 97ndash125 httpsdoiorg101146annurevenviron032108105046
Sardans J Rivas-Ubach A Pentildeuelas J 2012 The CNP stoichiometry of
organisms and ecosystems in a changing world A review and perspectives Perspect Plant Ecol Evol Syst 14 33ndash47 httpsdoiorg101016jppees201108002
Schindler DW Howarth RW Vitousek PM Tilman DG Schlesinger WH
Matson PA Likens GE Aber JD 2007 Human Alteration of the Global Nitrogen Cycle Sources and Consequences Ecol Appl 7 737ndash750 httpsdoiorg1018901051-0761(1997)007[0737haotgn]20co2
Scott E and EM Grantzs 2013 N2 fixation exceeds internal nitrogen loading as
a phytoplankton nutrient source in perpetually nitrogen-limited reservoirs Freshwater Science 2013 32(3)849ndash861 10189912-1901
28
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Talbot M R (2002) Nitrogen isotopes in palaeolimnology In Tracking
environmental change using lake sediments (pp 401-439) Springer Dordrecht
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable
isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K
Yeager KM 2016 Different responses of sedimentary δ15N to climatic changes and anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
29
CAPIacuteTULO 1 A COMBINED APPROACH TO ESTABLISHING THE TIMING
AND MAGNITUDE OF ANTHROPOGENIC NUTRIENT ALTERATION IN A
MEDITERRANEAN COASTAL LAKE- WATERSHED SYSTEM
30
A combined approach to establishing the timing and magnitude of anthropogenic
nutrient alteration in a mediterranean coastal lake- watershed system
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugoneabc Pablo
Sarricolead Santiago Giralte Manuel Contreras-Lopezf Ricardo Prego g Patricia
Bernaacuterdez g Blas Valero-Garceacutesch
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Institute of Earth Sciences Jaume Almera (ICTJA-CSIC) CLluis Solegrave Sabaris sn Barcelona E-
08028 Spain
f Facultad de Ingenieriacutea y Centro de Estudios Avanzados Universidad de Playa Ancha Traslavintildea
450 Vintildea del Mar Chile
g Instituto de Investigaciones Marinas (CSIC) CEduardo Cabello 6 36208 Vigo Spain
h Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding author
E-mail address
clatorrebiopuccl magdalenafuentealbagmailcom
Abstract
Since the industrial revolution and especially during the Great Acceleration (1950
CE) human activities have profoundly altered the global nutrient cycle through land
use and cover changes (LUCC) However the timing and intensity of recent N
variability together with the extent of its impact in terrestrial and aquatic ecosystems
and coupled effects of regional LUCC and climate are not well understood Here
we used a multiproxy approach (sedimentological geochemical and isotopic
31
analyses historical records climate data and satellite images) to evaluate the role
of LUCC as the main control for N cycling in a coastal watershed system of central
Chile during the last few centuries The largest changes in N dynamics occurred in
the mid-1970s associated with the replacement of native forests and grasslands for
livestock grazing by plantations of introduced Monterey pine (Pinus radiata) and
eucalyptus (Eucalyptus globulus) Lake productivity increased and is evidenced by
an increase trend in δ15N values Our study shows that anthropogenic land
usecover changes are key in controlling nutrient supply and N availability in
Mediterranean watershed ndash lake systems and that large-scale forestry
developments during the mid-1970s likely caused the largest changes in central
Chile
Keywords Anthropocene Organic geochemistry watershedndashlake system Stable
Isotope Analyses Land usecover change Nitrogen cycle Mediterranean
ecosystems central Chile
1 INTRODUCTION
Human activities have become the most important driver of the nutrient cycles in
terrestrial and aquatic ecosystems since the industrial revolution (Gruber and
Galloway 2008 Galloway et al 2008 Holtgrieve et al 2011 Fowler et al 2013
Goyette et al 2016) Among these N is a common nutrient that limits productivity
in terrestrial and aquatic ecosystems (Vitousek and Howarth 1991 McLauchlan et
al 2013) With the advent of the Haber-Bosch industrial N fixation process in the
early 20th century total N fluxes have surpassed previous planetary boundaries
32
(Galloway et al 2008 Elser 2011) reaching unprecedented values (ie tipping
points) in the Earth system especially during what is now termed the Great
Acceleration (which began in the 1950s) which intensified in the 1970s (Howarth
2004 Steffen et al 2015) Yet dramatic changes in LUCC have occurred in the last
few centuries mainly linked to forestry and agro-pastoral activities (Poraj-Goacuterska et
al 2017 Ge et al 2019 Cheng et al 2019) These have had large impacts on N
(and Carbon -C-) availability in terrestrial and aquatic ecosystems but the synergic
effect with climate change and global N dynamics has not been established
(Vitousek et al 1997 Mclauchlan et al 2007 2013 Pongratz et al 2010
Woodward et al 2012 Mclauchlan et al 2017)
The onset of the Anthropocene poses significant challenges in mediterranean
regions that have a strong seasonality of hydrological regimes and an annual water
deficit (Stocker et al 2013) Mediterranean climates occur in all continents
(California central Chile Australia South Africa circum-Mediterranean regions)
providing a unique opportunity to investigate global change processes during the
Anthropocene in similar climate settings but with variable geographic and cultural
contexts The effects of global change in mediterranean watersheds have been
analyzed from different perspectives hydrology (Giorgi 2006 Arnell and Gosling
2013 Prudhomme et al 2014) vegetation dynamics (Lenihan et al 2003 Vicente-
Serrano et al 2013 Matesanz and Valladares 2014) sediment dynamics (Garciacutea-
Ruiz et al 2013 Garciacutea-Ruiz 2010 Syvitski et al 2005) changes in
biogeochemical cycles (Thomas et al 2013 Fowler et al 2013 Frank et al 2015)
carbon storage (Muntildeoz-Rojas et al 2015) and biodiversity (Hooper et al 2012) A
recent review showed an extraordinarily high variability of erosion rates in
mediterranean watersheds positive relationships with slope and annual
33
precipitation and the paramount effect of land use (Garciacutea-Ruiz et al 2015)
However the temporal context and effect of LUCC on nutrient supply to
mediterranean lakes has not been analyzed in much detail
Major LUCC in central Chile occurred during the Spanish Colonial period
(17th-18th centuries) (Armesto 1994 Gurnell et al 2001 Figueroa et al 2004
Martiacuten-Foreacutes et al 2012 Contreras-Loacutepez et al 2014) with the onset of
industrialization and mostly during the mid to late 20th century (von Gunten et al
2009 Gayo et al 2012) Recent copper pollution caused by 20th century mining
and industrial smelters has been documented in cores throughout the Andes
(Laguna el Ocho and Laguna Ensuentildeo von Gunten et al 2009) and also from our
surveys in the central valley (Batuco wetland) coastal range (Cordillera de Name)
and along the coast itself (Bucalemito and Colejuda) (Valero-Garceacutes et al 2010
unpublished data)
Paleolimnological studies have shown how these systems respond to
climate LUCC and anthropogenic impacts during the last millennia (Jenny et al
2002 2003 Villa Martiacutenez et al 2002 Frugone-Aacutelvarez et al 2017 Contreras et
al 2018) Furthermore changes in sediment and nutrient cycles have also been
identified in associated terrestrial ecosystems dating as far back as the Spanish
Conquest and related to fire clearance and wood extraction practices of the native
forests (Armesto 1994 Contreras-Loacutepez et al 2014) Nevertheless pollen and
limnological evidence argue for a more recent timing of the largest anthropogenic
impacts in central Chile For example paleo records show that during the mid-20th
century increased soil erosion followed replacement of native forest by Pinus
radiata and Eucalyptus globulus plantations at Laguna Matanzas Aculeo and
34
Vichuqueacuten lakes (Jenny et al 2002 Villa-Martiacutenez et al 2002 2003 Frugone-
Aacutelvarez et al 2017)
Lakes are a central component of the global carbon cycle Lakes act as a
sink of the carbon cycle both by mineralizing terrestrially derived organic matter and
by storing substantial amounts of organic carbon (OC) in their sediments (Anderson
et al 2009) Paleolimnological studies have shown a large increase in OC burial
rates during the last century (Heathcote et al 2015) however the rates and
controls on OC burial by lakes remain uncertain as do the possible effects of future
global change and the coupled effect with the N cycle LUCC intensification of
agriculture and associated nutrient loading together with atmospheric N-deposition
are expected to enhance OC sequestration by lakes Climate change has been
mainly responsible for the increased algal productivity since the end of the 19th
century and during the late 20th century in lakes from both the northern (Ruumlhland et
al2015) and southern hemispheres (Michelutti et al 2015 Carrevedo et al 2015)
but many studies suggest a complex interaction of global warming and
anthropogenic influences and it remains to be proven if climate is indeed the only
factor controlling these transitions (Catalaacuten et al 2013) Alternative causes for
recent N (Galloway et al 2008) increases in high altitude lakes such as catchment
mediated processes cannot be ruled out (Camarero and Catalan 2012 Fritz and
Anderson 2013) Few lake-watershed systems have robust enough chronologies of
recent changes to compare variations in C and N with regional and local processes
and even fewer of these are from the southern hemisphere (McLauchlan et al
2007 Holtgrieve et al 2011)
In this paper we present a multiproxy lake-watershed study including N and
C stable isotope analyses on a series of short cores from Laguna Matanzas in
35
central Chile focused in the last 200 years We complemented our record with land
use surveys satellite and aerial photograph studies Our major objectives are 1) to
reconstruct the dynamics among climate human activities and changes in the N
cycle over the last two centuries 2) to evaluate how human activities have altered
the N cycle during the Great Acceleration (since the mid-20th century)
2 STUDY SITE
Laguna Matanzas (33ordm45rsquoS 71ordm 40acuteW 7 m asl Fig 1) is a coastal lake located
in central Chile near to a large populated area (Santiago gt6106 inhabitants) The
lake has a surface area of 15 km2 with a max depth of 3 m and a watershed of 30
km2 The lake basin is emplaced over Pleistocene-Holocene aeolian and alluvial fan
deposits (SERNAGEOMIN 2003) A recent phase of dune activity occurred from the
mid to late Holocene which mostly sealed off the basin from the ocean (Villa-
Martinez 2002) Climate in Laguna Matanzas is characterized by cool-wet winters
and hot-dry summers with annual precipitation of ~510 mm and a mean annual
temperature of 12ordmC Central Chile is in the transition zone between the southern
hemisphere mid-latitude westerlies belt and the South Pacific Anticyclone (or SPA)
(Garreaud et al 2009) In winter precipitation is modulated by the north-west
displacement of the SPA the northward shift of the westerlies wind belt and an
increased frequency of storm fronts stemming off the Southern Hemisphere
Westerly Winds (SWW) (Frugone-Aacutelvarez et al 2017) Austral summers are
typically dry and warm as a strong SPA blocks the northward migration of storm
tracks stemming off the SWW
36
Historic land cover changes started after the Spanish conquest with a Jesuit
settlement in 1627 CE near El Convento village and the development of a livestock
ranch that included the Matanzas watershed After the Jesuits were expelled from
South America in 1778 CE the farm was bought by Pedro Balmaceda and had more
than 40000 head of cattle around 1800 CE (Contreras-Lopez et al 2014) The first
Pinus radiata and Eucalyptus globulus trees were planted during the second half of
the 19th century and were mostly used for dune stabilization (Albert 1900 Gibson
1972) However the main plantation phase occurred 60 years ago (Villa-Martinez
2002) as a response to the application of Chilean Forestry Laws promulgated in
1931 and 1974 and associated state subsidies
Major land cover changes occurred recently from 1975 to 2008 as shrublands
were replaced by more intensive land uses practices such as farmland and tree
plantations (Schulz et al 2010) Laguna Matanzas is part of the Reserva Nacional
Humedal El Yali protected area (Ramsar Ndeg 878) Despite this protected status the
lake and its watershed have been heavily affected by intense agricultural and
farming activities during the last decades The main inlet ldquoLas Rosasrdquo has been
diverted for crop irrigation causing a significant loss of water input to the lake
Consequently the flooded area of the lake has greatly decreased in the last couple
of decades (Fig 1b) Exotic tree species cover a large surface area of the
watershed Recently other activities such as farms for intensive chicken production
have been emplaced in the watershed
37
Figure 1 Laguna Matanzas a) Digital Elevation Model showing the watershed and
the surface hydrologic connection with Colejuda and Cabildo lakes b) Climograph
depicting the warm dry season in austral summer c) Annual precipitation from
1965-2015 (arrow shows start of the most recent ldquomega-droughtrdquo- see Garreaud et
al 2017) d) A decline of lake area occurs between 2007 and 2018 Lake surface
area decreased first along the western sector (in 2007) followed by more inland
areas (in 2018)
38
3 RESULTS
31 Age Model
The age model for the Matanzas sequence was developed using Bacon software
to establish the deposition rates and age uncertainty (Blaauw and Christen 2011)
It is based on 210Pb and 137Cs dates and two 14C dates (Table 2) According to this
age model the lake sequence spans the last 1000 years (Fig 2) A major breccia
layer (unit 3b) was deposited during the early 18th century which agrees with
historic documents indicating that a tsunami impacted Laguna Matanzas and its
watershed in 1730 CE (Contreras-Loacutepez et al 2014) Here we focus in the last 200
years were the most important changes occurred in terms of LUCC (after the
sedimentary hiatus caused by the tsunami) The Spanish colonial period (17th -18th
century) brought new forms of territorial management along with an intensification
of watershed use which remained relatively unchanged until the 1900s
39
Figure 2 a) Age-depth model obtained for the Laguna de Matanzas sedimentary
sequence A hiatus occurs at 80 cm depth (unit 3b) The section of core used for our
analysis is highlighted in a red rectangle b) Close up of the age model used for
analysis of recent anthropogenic influences on the N cycle c) Information regarding
the 14C dates used to construct age model
Lab code Sample ID
Depth (cm) Material Fraction of modern C
Radiocarbon age
Pmc Error BP Error
D-AMS 021579
MAT11-6A 104-105 Bulk Sediment
8843 041 988 37
D-AMS 001132
MAT11-6A 1345-1355
Bulk Sediment
8482 024 1268 21
POZ-57285
MAT13-12 DIC Water column 10454 035 Modern
Table 2 Laguna Matanzas radiocarbon dates
32 The sediment sequence
Laguna Matanzas sediments consist of massive to banded mud with some silt
intercalations They are composed of silicate minerals (plagioclase quartz and clay
minerals) with relatively high TOC content (Fig 3) Pyrite is a common mineral
indicating dominant anoxic conditions in the lake sediments whereas aragonite
occurs only in the uppermost section Mineralogical analyses visual descriptions
texture and geochemical composition were used to characterize five main facies
(Fig 3) F1 (organic-rich mud) represents baseline sedimentation in a shallow well-
mixed brackish highly productive lake and F1acute (dark orange) is a less organic facies
than F1 (more details see table in the supplementary material) F2 (massive to
banded silty mud) indicates periods of higher clastic input into the lake but finer
(mostly clay minerals) likely from suspension deposition associated with flooding
40
events Aragonite (up to 15 ) occurs in both facies but only in samples from the
uppermost 30 cm and it is interpreted as endogenic linked to higher MgFe waters
and elevated biologic productivity
Figure 3 Sedimentary facies and units mineralogy grain size elemental geochemical
and isotopic composition of Laguna Matanzas core Values highlighted in gray indicate
that these are above average
The banded to laminated fining upward silty clay layers (F3) reflect
deposition by high energy turbidity currents The presence of aragonite suggests
that littoral sediments were incorporated by these currents Non-graded laminated
coarse silt layers (F4) do not have aragonite indicating a dominant watershed
41
sediment source Both facies are interpreted as more energetic flood deposits but
with different sediment sources A unique breccia layer with coarse silt matrix and
cm-long soft clasts (F5) is interpreted as a high-energy event (ie a tsunami)
capable of eroding the littoral zone and depositing coarse clastic material in the
distal zone of the lake Similar coarse breccia layers have been found at several
coastal sites in Chile and interpreted as tsunami-related deposits (Vargas et al
2005 Le Roux et al 2008)
33 Sedimentary units
Three main units and six subunits have been defined (Fig 3) based on
sedimentary facies and sediment composition We use ZrTi as an indicator of the
mineral fraction transported from the watershed (Marzecoacuteva et al 2011) as higher
ZrTi (F3 and F4) is commonly associated with coarser sediments (Cuven et al
2010) A high correlation among Br BrTi and TOC (r = 046-087 p value=0)
supports the use of BrTi as an indicator of lake productivity (Marzecoacuteva et al 2011
Frugone-Aacutelvarez et al 2017) The MnFe ratio is indicative of lake bottom
oxygenation (Naecher et al 2013) as under reducing conditions Mn mobilizes more
than Fe leading to a decreased MnFe ratio (Marzecoacuteva et al 2011) SrTi indicates
periods of increased aragonite formation as Sr is preferentially included in the
aragonite mineral structure (Veizer et al 1971) (See supplementary material)
The basal unit (3c 129 to 99 cm) is relatively organic-rich (TOC mean=26
BioSi mean=5) and composed by F1 (without aragonite) with some coarser F4
flood layers (ZrTi mean=025) Unit 3b (98 to 80 cm) is interpreted as a tsunami or
storm surge deposit (breccia F5 grading into massive to banded silt F2) Unit 3a
(79 to 73 cm) is characterized by relatively low productivity (TOC=2 BrTi=002
42
BioSi=4) under anoxic conditions (MnFe=001) Unit 2a (72 to 31cm) has
relatively less organic content and more intercalated clastic facies F3 and F4 The
top of this unit (43-30 cm) has elevated TS values The Subunit 1b (30-20 cm)
shows increasing TOC BioSi and BrTi values (TOC mean = 29 BioSi mean =
54 BrTi mean=004) and the upper subunit 1a (19-0 cm) has the highest TOC
(mean = 64) and BioSi (mean = 56) high BrTi mean = 010 and the presence
of aragonite More frequent anoxic conditions (MnFe lower than 001) during units
3 and 2 shifted towards more oxic episodes (MnFe = 003) in unit 1 (Fig 3)
34 Isotopic signatures
Figure 4 shows the isotopic signature from soil samples of the major land
usescover present in the Laguna Matanzas used as an end member in comparison
with the lacustrine sedimentary units δ15N from cropland samples exhibit the
highest values whereas grassland and soil samples from lake shore areas have
intermediate values (Fig 4) Tree plantations and native forests have similarly low
δ15N values (+11 permil SD=24) All samples (except those from the lake shore)
exhibit low δ13C values (from ndash285permil to ndash298permil) CNmolar from agriculture land
lakeshore area and non-vegetation areas samples display the lowest values (about
18) CNmolar from tree plantations and native forest have the highest values (383
and 267 respectively)
43
Figure 4 C-N stable isotope plot showing a comparison of lake sediments grouped
by sedimentary units (MAT11-6A) with the soil end members of present-day (lake
shore and land usecover) from Laguna Matanzas
The δ15N values from sediment samples (MAT11-6A) range from ndash15 and
+53permil (mean= +35 SD=05) δ13C values range from ndash266permil to ndash202permil (mean=
ndash240 SD=14) In unit 3c δ15N and δ13C show relatively high values (mean=
+41permil and ndash233permil respectively) δ15N and δ13C from unit 3b and 3a fluctuate at
slightly lower values than in 3c (mean δ15N from 3a= +38permil and 3b mean= +39permil
mean δ13C from 3a= ndash242permil and 3b mean= ndash244permil) In unit 2a δ15N values are
relatively high (mean= +38permil) but show a slightly decreasing trend (from +52 to
+24permil at 66 cm and 36 cm respectively) δ13C also decreases (mean= ndash242permil)
reaching minimum values at 45 cm (ndash266permil) with a sharp increase towards the top
of this unit with maximum values ca ndash210permil Unit 1 exhibits the lowest δ15N values
(1a mean= +11permil 1b mean= +28permil) with negative values in the uppermost
44
sediments (ndash04permil at 14 cm) δ13C values show a decreasing trend over most of
subunit 1b and increase only near the very top of this unit
35 Recent land use changes in the Laguna Matanzas watershed
Major LUCC from 1975 (unit 1b) to 2016 CE in the Laguna Matanzasacutes
watershed is summarized in Figure 5 The watershed has a surface area of 30 km2
of which native forest (36) and grassland areas (44) represented 80 of the
total surface in 1975 The area occupied by agriculture was only 02 and tree
plantations were absent Isolated burned areas (33) were located mostly in the
northern part of the watershed By 1989 tree plantations surface area had increased
to 5 burned areas to 17 and agricultural fields to 9 (a 45-fold increase) and
native forest and grassland sectors decreased to 23 and 27 respectively By
2016 agricultural land and tree plantations have increased to 17 of the total area
whereas native forests decreased to 21
45
Figure 5 Land uses and cover changes from 1975 to 2016 in Laguna Matanzas
watershed from natural cover and areas for livestock grazing (grassland) to the
expansion of agriculture and forest plantation
4 DISCUSSION
41 N and C dynamics in Laguna Matanzas
Small lakes with relatively large watersheds such as Laguna Matanzas would
be expected to have relatively high contributions of allochthonous C to the sediment
OM (Gu et al 2006) Terrestrial C3 plants (δ13C =ndash26 to ndash28permil Meyers and Terranes
2001 Ku et al 2007) are dominant in the Laguna Matanzas watershed Likewise
our soil samples ranged across similar although slightly more negative values
46
(δ13C = ndash30 to ndash28permil Fig 4) to those proposed by Meyers and Terranes (2001)
and are used here as terrestrial end members oil samples were taken from the lake
shore (δ13C = ndash22permil) and POM from surface water (δ13C = ndash24permil) were more
positive than the terrestrial end member and are used as lacustrine end members
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from terrestrial vegetation and more positive δ13C values have increased
aquatic OM (Gu et al 2006 Torres et al 2012) Phytoplankton preferentially uptake
12C leaving the DIC pool enriched in 13C (Gu et al 2006) especially when there are
no important external sources of C (eg decreased C input from the watershed)
Therefore during events of elevated primary productivity the phytoplankton uptakes
12C until its depletion and are then obligated to use the heavier isotope resulting in
an increase in δ13C Changes in lake productivity thus greatly affect the C isotope
signal (Torres et al 2012) with high productivity leading to elevated δ13C values
(Torres et al 2012 Gu et al 2006)
In a similar fashion the N isotope signatures in Laguna Matanzas reflect a
combination of factors including different N sources (autochthonousallochthonous)
and lake processes such as productivity isotope fractionation in the water column
and sediment denitrification Elevated δ15N values from a POM sample (+22permil) and
average values from the lake shore (mean=+34permil SD=028) are used as aquatic
end members whereas terrestrial samples have values from +10 +24 (tree species)
to +77permil +35 (agriculture) which we use as terrestrial end members (Fig 4)
Autochthonous OM in aquatic ecosystems typically displays low δ15N values
when the OM comes from N-fixing species Atmospheric fixation of N2 by
cyanobacteria ends in OM with δ15N values close to 0permil (Leng et al 2006)
Phytoplankton preferentially uptake 14N from Dissolved Inorganic Nitrogen (DIN) in
47
the water column and derived OM typically have δ15N values lower than DIN values
When productivity increases the remaining DIN becomes depleted in 14N which in
turn increases the δ15N values of phytoplankton over time especially if the N not
replenished (Torres et al 2012) Thus high POM δ15N values from Laguna
Matanzas reflect elevated phytoplankton productivity with a 14N- depleted DIN In
addition N-watershed inputs also contribute to high δ15N values Heavily impacted
watersheds by human activities are often reflected in isotope values due to land use
changes and associated modified N fluxes For example the input of N runoff
derived from the use of inorganic fertilizers leads to the presence of elevated δ15N
(between ndash4 to +4permil) values in the water bodies (Elliott and Brush 2006 Diebel and
Vander Zanden 2009) Widory et al (2004) reported a direct relationship between
elevated δ15N values and increased nitrate concentration from manure in the
groundwater of Brittany (France) Terranes and Bernasconi (2000) found a good
correlation between augmented nutrient loading and a progressive increase in δ15N
values of sedimentary OM related to agricultural land use
Post-depositional diagenetic processes can further affect C and N isotope
signatures Several studies have shown a decrease in δ13C values of OM in anoxic
environments particularly during the first years of burial related to the selective
preservation of OM depleted in 13C (Hollander and Smith 2001 Lehmann et al
2002 Gӓlman et al 2009 Torres et al 2012) Diagenetic processes can also lead
to post-burial N isotope enrichment of the sediments Indeed 14N is consumed more
rapidly than 15N by denitrifying bacteria which intensifies under anoxic conditions
(Diebel and Vander Zanden 2009) Thus the remaining OM pool will be enriched
in 15N leaving to elevated δ15N values (Granger et al 2008 Menzel et al 2013)
48
In summary the relatively high δ15N values in sediments of Laguna Matanzas
reflect N input from an agriculturegrassland watershed with positive synergetic
effects from increased lake productivity enrichment of DIN in the water column and
most likely denitrification The increase of algal productivity associated with
increased N terrestrial input andor recycling of lake nutrients (and lesser extent
fixing atmospheric N) and denitrification under anoxic conditions can all increase
δ15N values (Fig 3) In addition elevated lake productivity without C replenishing
(eg by terrestrial C input) produces shifts towards positive δ13C values whereas C
input from the watershed generates more negative δ13C values
42 Recent evolution of the Laguna Matanzas watershed
Sedimentological compositional and geochemical indicators show three
depositional phases in the lake evolution under the human influence in the Laguna
Matanzas over the last two hundred years Although the record is longer (around
1000 years) we analyzed the last two centuries (unit 2 and 1) to provide a recent
historical context for the large changes detected during the 20th century
The first phase lasted from the beginning of the 19th century until ca 1940
(unit 2a Fig 3 4) and was characterized by moderate productivity with elevated
sediment input from the watershed as indicated by our geochemical proxies (BrTi
= 004 AlTi gt 01 Fig 6) Higher lake levels and dominant anoxic bottom conditions
(MnFe = 001 Fig3) can be related to increased precipitation (mean= 557 mmyr)
and lower temperatures (summer annual temperature lt19ordmC) During the Spanish
colonial period the Laguna Matanzas watershed was used as a livestock farm
(Jesuits 1627-1780 unit 3 followed by the Pedro Balmaceda era 1780-1810- unit
2a) (Contreras-Loacutepez et al 2014) The main buildings and farm facilities at the El
49
Convento village During this period livestock grazing and lumber extraction for
mining would have involved extensive deforestation and loss of native vegetation
(eg Armesto et al 1994 2010) However the Matanzas pollen record does not
show any significant regional deforestation during this period (Villa Martiacutenez 2002)
suggesting that the impact may have been highly localized
Lake productivity sediment input and elevated precipitation (Fig 6) all
suggest that N availability was related to this increased input from the watershed
The N from cow manure and soil particles would have led to higher δ15N values
(Elliot and Brush 2016) In addition anoxic lake bottom conditions would lead to
even further enrichment of buried sediment N The δ13C values lend further support
to our interpretation of increased sediment input -and N- from the watershed
Decreased δ13C values reach their lowest values in the entire record (ndash270permil) at
ca 1910 CE (Fig 4 6)
During most of the 19th century human activities in Laguna Matanzas were
similar to those during the Spanish Colonial period However the appearance of
Pinus radiata and Eucalyptus globulus pollen ca 1890 CE (Gibson 1972) the dune
stabilization-afforestation program which began ca 1900 CE (Albert 1900) and the
application of the Chilean forestry law of 1931 (DFL nordm265) contributed to an
increased capacity of the surrounding vegetation to retain nutrients and sediments
The law subsidized forest plantations in areas devoid of vegetation and prohibited
the cutting of forest on slopes greater than 45ordm These land use changes were coeval
with decreased sediment inputs (AlTi trend) from the watershed slightly increased
lake productivity (BrTi from 001 to 003) and a decrease in annual precipitation
(Fig 6) N isotope values become more negative during this period although they
remained high (from +49permil to +37permil) whereas the δ13C trend towards more
50
positive values reflects changes in the N source from watershed to in-lake dynamics
(e g increased endogenic productivity)
The second phase started after 1940 and is clearly marked by an abrupt
change in the general trend of δ15N that oscillates around +41permil (SD = 04permil) during
the first phase to +24permil (SD = 14permil) during the second These δ15N trends reflect
the lowest watershed nutrient and sediment inputs (based on the AlTi record)
decreased precipitation (mean = 318 mm year) and a slight increase in lake
productivity (increased BrTI) Depositional dynamics in the lake likely crossed a
threshold as human activity intensified throughout the watershed and lake levels
decreased
During the Great Acceleration δ15N values shifted towards higher values to
ca 3permil with an increase in δ13C values that are not reflected either in lake
productivity or lake level As the sediment input from the watershed increased and
precipitation remained as low as the previous decade δ15N values during this period
are likely related to watershed clearance which would have increased both nutrient
and sediment input into the lake
The δ13C trend to more positive values reaching the peaks in the 1960s (ndash
212permil) at the same time as the peaks in temperature (196 ordmC mean annual) with a
downward trend in precipitation A shift in OM origin from macrophytes and
watershed input influences to increased lake productivity could explain this trend
(Fig 4 1b)
In the 1970s the Laguna Matanzasacute watershed was mostly covered by native
forest (36) and grassland areas were intended for livestock grazing (44 Fig 5)
Both soils have δ15NSoil lt 5permil (Fig 4) Farming fields occupied a very small area and
tree plantations were almost nonexistent The decreasing trend in δ15N values seen
51
in our record is interrupted by several large peaks that occurred between ca 1975
and ca 1989 when the native forest and grassland areas fell by 23 and 27
respectively largely due to fires affecting 17 of the forests Agriculture fields
increased by 4 and forest plantations increased by 9 (unit 1a) Concomitantly
sediment ndash and likely N - inputs from the watershed decreased (as indicated by the
trend in AlTi) the precipitation was still relatively low (Fig 6) These changes are
likely related to the increase of vegetation cover especially of tree plantations (which
have more negative δ15N values) The small increase in productivity in the lake could
have been favored by increased temperature (von Gunten et al 2009) After 1989
the increase in agriculture land (17 in 2016) correlates with increasing δ15N δ13C
and TOC trends in spite of declining rainfall The increase of forest plantations was
mostly in response to the implementation of the Law Decree of Forestry
Development (DL 701 of 1974) that subsidized forest plantation After 1989 the
increase in agricultural land (17 in 2016) is synchronous with increasing δ15N
δ13C TOC and MnFe trends despite the decline in rainfall and overall lower lake
levels as more water is used for irrigation
The third phase started c 1990 CE (unit 1a) when OM accumulation rates
increase and δ13C δ15N decreased reaching their lowest values in the sequence
around 2000 CE Afterward during the 21st century δ13C and δ15N values again
began to increase The onset of unit 1 is marked by increased lake productivity and
decreased sediment input (AlTi lt 02) synchronous with intensive farming replacing
forestry and extensive agriculture (Fig 5 6)
A change in the general trend of δ15N values which decreased until 1990
(unit 1a) as the native forest and grassland area fell to 23 and 27 respectively
is most likely due to deforestation and fires Agriculture surface increased to 4 and
52
forest plantations increased by 9 (Fig 5) Concomitantly sediment ndash and likely N
ndash inputs from the watershed decreased probably related to the low precipitation (Fig
1b) and the increase of vegetation cover in the watershed in particularly by tree
plantations (with more negative δ15N Fig 4)
At present agriculture and tree plantations occupy around 34 of the
watershed surface whereas native forests and grassland cover 21 and 25
respectively Lake productivity as indicated by BrTi (Fig6) is higher and generates
OM with lower δ15N and higher δ13C values (about +2permil and ndash20permil ca 2000 CE
respectively) due to in-lake processes (ie biological N fixation and nutrient
recycling) and driven by changes in the arboreal cover which diminishes nutrient
flux into the lake (Fig6)
53
Figure 6 Anthropogenic and climatic forcing and lake dynamics response
(productivity sediment input N and C cycles) at Matanzas Lake over the last two
54
centuries Mean annual precipitation reconstructed and temperatures (von Gunten
et al 2009) Vertical gray bars indicate mega-droughts
5 CONCLUSIONS
Human activities have been the main factor controlling the N and C cycle in
the Laguna Matanzas during the last two centuries The N isotope signature in the
lake sediments reflects changes in the watershed fluxes to the lake but also in-lake
processes such as productivity and post-depositional changes Denitrification could
have been a dominant process during periods of increased anoxic conditions which
were more frequent prior to Great Acceleration (1950 CE) Higher δ15N and lower
δ13C values are associated with increased nutrient input from the watershed due to
increased livestock grazing and agriculture pressure (phase 1 Fig 7a) whereas
lower isotope values occurred during periods of increased forest plantations (phase
3 Fig 7c) During periods of increased lake productivity - such as in the last few
decades - δ15N values increased significantly
The most important change in C and N dynamics in the lake occurred after the
1950s (Phase 2 and 3) as tree plantations dominated the watershed Recent
changes in N dynamics can be explained by the higher nutrient contribution
associated with intensive agriculture (i e fertilizers) since the 1990s Although the
replacement of livestock activities with forestry and farming seems to have reduced
nutrient and soil export from the watershed to the lake the inefficient use of fertilizer
(by agriculture) can be the ultimate responsible for lake productivity increase during
the last decades
55
Figure 7 Schematic diagrams illustrating the main factors controlling the
isotope N signal in sediment OM of Laguna Matanzas N input from watershed
depends on human activities and land cover type Agriculture practices and cattle
(grassland development) contribute more N to the lake than native forest and
plantations Periods of higher productivity tend to deplete the dissolved inorganic N
in 14N resulting in higher δ15N (OM) The denitrification processes are more effective
in anoxic conditions associated with higher lake levels
6 METHODS
Short sediment cores were recovered from Laguna Matanzas using an Uwitec
gravity piston corer in 2011 and 2013 (Fig 1a) Sediment cores (MAT11-6A 129 cm
MAT13-2A 149 cm MAT13-3A 33 cm MAT13-4A 99 cm) were split
photographed sub-sampled and stored at the Pyrenean Institute of Ecology (IPE-
CSIC Spain) Core MAT11-6A was obtained from the central sector of the lake and
56
was selected for detailed multiproxy analyses (including elemental geochemistry C
and N isotope analyses XRF and 14C dating)
The isotope analyses (δ13C and δ15N) were performed at the Laboratory of
Biogeochemistry and Applied Stable Isotopes (LABASI-PUC Chile) using a Delta
V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via a
Conflo IV interface Isotope results are expressed in standard delta notation (δ) in
per mil (permil) relative to the standards Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Sediment samples
for δ13Corg were pre-treated with 50 ml of HCl and 50 ml of deionized water and
dried at 60ordmC for 4 hr to remove carbonates (Harris et al 2011)
Total Carbon (TC) Total Organic Carbon (TOC) Total Inorganic Carbon (TIC)
and Total Sulphur (TS) were measured every cm with a LECO SC 144 DR at IPE-
CSIC XRF measurements were carried out every 4 mm in MAT11-1A-1G core using
an AVAATECH X-ray Fluorescence II core scanner at the University of Barcelona
(Spain) Results are expressed as element intensities in counts per second (cps)
Tube voltage was operated at 30 kV and 10 kV to obtain the abundances of 15
elements (Al Si S Cl K Ca Ti V Mn Fe Br Rb Sr Y Zr) with an average at
least of 1600 cps (less for Br=1000)
Biogenic silica content mineralogy and grain size were measured every 4
cm Biogenic silica was measured following Mortlock and Froelich (1989) and
Bernaacuterdez et al (2005) using an Auto Analyzer Technicon AAII for dissolved silicate
analysis Mineralogy was analyzed with a Siemens D-500 X-ray diffractometer (Cu
kα 40 kV 30 mA graphite monochromator) at the ICTJA-CSIC (Spain) Grain size
analyses were performed in a Beckmann Coulter LS 13 320 Particle Size Analyzer
57
at the IPE-CSIC The samples were classified according to textural classes as
follows clay (lt2 μm) silt (20-2 microm) and sand (gt2 microm) fractions
The age-depth model for the Laguna Matanzas sedimentary sequence was
constructed using 210Pb and 137Cs dating techniques (MAT13-4A) as well as two 14C
AMS dates on bulk sediment samples (MAT11-6A fig 2) We dated the dissolved
inorganic carbon (DIC) in the water column and no significant reservoir effect is
present in the modern-day water column (10454 + 035 pcmc Table 2) An age-
depth model was obtained with the Bacon R package to estimate the deposition
rates and associated age uncertainties along the core (Blaauw and Christen 2011)
To estimate LUCC of Laguna Matanzasacutes watershed we use satellite images
Landsat MSS for 1975 Landsat TM for 1989 and Landsat OLI for 2016 all taken in
summer or autumn (Table 1) We performed supervised classification of land uses
(maximum likelihood algorithm) for each year (1975 1989 and 2016) and the results
were mapped using software ArcGIS 102 in 2017
Satellite Images Acquisition Date
Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat OLI 20160404 30 m
Table 1 Landsat imagery
Surface water samples were filtered for obtained particulate organic matter In
addition soil samples from the main land usecover present in the Laguna Matanzas
watershed were collected Elemental C N and their corresponding isotopes from
POM and soil were obtained at the LABASI and used here as end members
Daily precipitation at Santo Domingo (33ordm39`S 71ordm3 6`W)- the nearest weather
station to Laguna Matanzasndash was compiled using the redPrec R package (Fig1 d
Serrano-Notivoli et al 2017 Sarricolea et al 2019) To extend our precipitation
58
reconstruction back to 1824 we correlated this dataset with that available for
Santiago The Santiago data was compiled from data published in the Anales of
Universidad of Chile (1850) for the years 1824 to 1849 Almeyda (1940) for the years
1849 to 1864 and the Quinta Normal series from 1866 to the present (Direccioacuten
Meteoroloacutegica de Chile) We generated a linear regression model between the
presentday Santo Domingo station and the compiled Santiago data with a Pearson
coefficient of 087 and p-valuelt 001
Acknowledgments This research was funded by grants CONICYT AFB170008
to the Institute of Ecology and Biodiversity (IEB) and FONDECYT 1150763 (to CL)
Doctoral grant Becas Chile 21150224 MEDLANT (Spanish Ministry of Economy
and Competitiveness grant CGL2016-76215-R) Additional funding was provided
by the Laboratorio Internacional de Cambio Global (LINCGlobal PUC-CSIC) We
thank R Lopez E Royo and M Gallegos for help with sample analyses We thank
the Laboratory of Biogeochemistry and Applied Stable Isotopes (LABASI) of the
Department of Ecology (PUC) for sample analyses
References
Albert F 1900 Las dunas o sean arenas volantes voladeros arenas muertas invasion de las arenas playas I meacutedanos del centro de Chile comprendiendo el litoral desde el liacutemite norte de la provincia de Aconcagua hasta el liacutemite sur de la de Arauco Memorias cientiacuteficas I Lit
Anderson NJ W D Fritz SC 2009 Holocene carbon burial by lakes in SW
Greenland Glob Chang Biol httpsdoiorg101111j1365-2486200901942x
Armesto J Villagraacuten C Donoso C 1994 La historia del bosque templado
chileno Ambient y Desarro 66 66ndash72 httpsdoiorg101007BF00385244
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A
59
historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Arnell NW Gosling SN 2013 The impacts of climate change on river flow
regimes at the global scale J Hydrol 486 351ndash364 httpsdoiorg101016JJHYDROL201302010
Bernaacuterdez P Prego R Franceacutes G Gonzaacutelez-Aacutelvarez R 2005 Opal content in
the Riacutea de Vigo and Galician continental shelf Biogenic silica in the muddy fraction as an accurate paleoproductivity proxy Cont Shelf Res httpsdoiorg101016jcsr200412009
Blaauw M Christen JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474 httpsdoiorg10121411-BA618
Brush GS 2009 Historical land use nitrogen and coastal eutrophication A
paleoecological perspective Estuaries and Coasts 32 18ndash28 httpsdoiorg101007s12237-008-9106-z
Camarero L Catalan J 2012 Atmospheric phosphorus deposition may cause
lakes to revert from phosphorus limitation back to nitrogen limitation Nat Commun httpsdoiorg101038ncomms2125
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego
R Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and environmental change from a high Andean lake Laguna del Maule central Chile (36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Catalan J Pla-Rabeacutes S Wolfe AP Smol JP Ruumlhland KM Anderson NJ
Kopaacuteček J Stuchliacutek E Schmidt R Koinig KA Camarero L Flower RJ Heiri O Kamenik C Korhola A Leavitt PR Psenner R Renberg I 2013 Global change revealed by palaeolimnological records from remote lakes A review J Paleolimnol httpsdoiorg101007s10933-013-9681-2
Chen Y Y Huang W Wang W H Juang J Y Hong J S Kato T amp
Luyssaert S (2019) Reconstructing Taiwanrsquos land cover changes between 1904 and 2015 from historical maps and satellite images Scientific Reports 9(1) 3643 httpsdoiorg101038s41598-019-40063-1
Contreras S Werne J P Araneda A Urrutia R amp Conejero C A (2018)
Organic matter geochemical signatures (TOC TN CN ratio δ 13 C and δ 15 N) of surface sediment from lakes distributed along a climatological gradient on the western side of the southern Andes Science of The Total Environment 630 878-888 httpsdoiorg101016jscitotenv201802225
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia UC
60
de V 2014 ELEMENTOS DE LA HISTORIA NATURAL DEL An Mus Hist Natulas Vaplaraiso 27 51ndash67
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-010-9453-1
Diebel M Vander Zanden MJ 201209 Nitrogen stable isotopes in streams
stable isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19 1127ndash1134 httpsdoiorg10189008-03271
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506 Espizua LE Pitte P 2009 The Little Ice Age glacier advance in the Central
Andes (35degS) Argentina Palaeogeogr Palaeoclimatol Palaeoecol 281 345ndash350 httpsdoiorg101016JPALAEO200810032
Figueroa JA Castro S A Marquet P A Jaksic FM 2004 Exotic plant
invasions to the mediterranean region of Chile causes history and impacts Invasioacuten de plantas exoacuteticas en la regioacuten mediterraacutenea de Chile causas historia e impactos Rev Chil Hist Nat 465ndash483 httpsdoiorg104067S0716-078X2004000300006
Fowler D Coyle M Skiba U Sutton MA Cape JN Reis S Sheppard
LJ Jenkins A Grizzetti B Galloway N Vitousek P Leach A Bouwman AF Butterbach-bahl K Dentener F Stevenson D Amann M Voss M 2013 The global nitrogen cycle in the twenty- first century Philisophical Trans R Soc B httpsdoiorghttpdxdoiorg101098rstb20130164
Frank D Reichstein M Bahn M Thonicke K Frank D Mahecha MD
Smith P van der Velde M Vicca S Babst F Beer C Buchmann N Canadell JG Ciais P Cramer W Ibrom A Miglietta F Poulter B Rammig A Seneviratne SI Walz A Wattenbach M Zavala MA Zscheischler J 2015 Effects of climate extremes on the terrestrial carbon cycle Concepts processes and potential future impacts Glob Chang Biol httpsdoiorg101111gcb12916
Fritz SC Anderson NJ 2013 The relative influences of climate and catchment
processes on Holocene lake development in glaciated regions J Paleolimnol httpsdoiorg101007s10933-013-9684-z
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
61
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS) implications for past sea level and environmental variability J Quat Sci 32 830ndash844 httpsdoiorg101002jqs2936
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892 httpsdoiorg101126science1136674
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924 httpsdoiorg104319lo20095430917
Garciacutea-Ruiz JM 2010 The effects of land uses on soil erosion in Spain A
review Catena httpsdoiorg101016jcatena201001001
Garciacutea-Ruiz JM Loacutepez-Moreno JI Lasanta T Vicente-Serrano SM
Gonzaacutelez-Sampeacuteriz P Valero-Garceacutes BL Sanjuaacuten Y Begueriacutea S Nadal-Romero E Lana-Renault N Goacutemez-Villar A 2015 Los efectos geoecoloacutegicos del cambio global en el Pirineo Central espantildeol una revisioacuten a distintas escalas espaciales y temporales Pirineos httpsdoiorg103989Pirineos2015170005
Garciacutea-Ruiz JM Nadal-Romero E Lana-Renault N Begueriacutea S 2013
Erosion in Mediterranean landscapes Changes and future challenges Geomorphology 198 20ndash36 httpsdoiorg101016jgeomorph201305023
Garreaud RD Vuille M Compagnucci R Marengo J 2009 Present-day
South American climate Palaeogeogr Palaeoclimatol Palaeoecol httpsdoiorg101016jpalaeo200710032
Gayo EM Latorre C Jordan TE Nester PL Estay SA Ojeda KF
Santoro CM 2012 Late Quaternary hydrological and ecological changes in the hyperarid core of the northern Atacama Desert (~21degS) Earth-Science Rev httpsdoiorg101016jearscirev201204003
Ge Y Zhang K amp Yang X (2019) A 110-year pollen record of land use and land
cover changes in an anthropogenic watershed landscape eastern China Understanding past human-environment interactions Science of The Total Environment 650 2906-2918 httpsdoiorg101016jscitotenv201810058
Goyette J Bennett EM Howarth RW Maranger R 2016 Global
Biogeochemical Cycles 1000ndash1014 httpsdoiorg1010022016GB005384Received
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and
oxygen isotope fractionation during dissimilatory nitrate reduction by
62
denitrifying bacteria 53 2533ndash2545 httpsdoiorg10230740058342
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
Gurnell AM Petts GE Hannah DM Smith BPG Edwards PJ Kollmann
J Ward J V Tockner K 2001 Effects of historical land use on sediment yield from a lacustrine watershed in central Chile Earth Surf Process Landforms 26 63ndash76 httpsdoiorg1010021096-9837(200101)261lt63AID-ESP157gt30CO2-J
Heathcote A J et al Large increases in carbon burial in northern lakes during the
Anthropocene Nat Commun 610016 doi 101038ncomms10016 (2015) Hollander DJ Smith MA 2001 Microbially mediated carbon cycling as a
control on the δ13C of sedimentary carbon in eutrophic Lake Mendota (USA) New models for interpreting isotopic excursions in the sedimentary record Geochim Cosmochim Acta httpsdoiorg101016S0016-7037(00)00506-8
Holtgrieve GW Schindler DE Hobbs WO Leavitt PR Ward EJ Bunting
L Chen G Finney BP Gregory-Eaves I Holmgren S Lisac MJ Lisi PJ Nydick K Rogers LA Saros JE Selbie DT Shapley MD Walsh PB Wolfe AP 2011 A coherent signature of anthropogenic nitrogen deposition to remote watersheds of the Northern Hemisphere Science (80) 334 1545ndash1548 httpsdoiorg101126science1212267
Hooper DU Adair EC Cardinale BJ Byrnes JEK Hungate BA Matulich
KL Gonzalez A Duffy JE Gamfeldt L Connor MI 2012 A global synthesis reveals biodiversity loss as a major driver of ecosystem change Nature httpsdoiorg101038nature11118
Horvatinčić N Sironić A Barešić J Sondi I Krajcar Bronić I Borković D
2016 Mineralogical organic and isotopic composition as palaeoenvironmental records in the lake sediments of two lakes the Plitvice Lakes Croatia Quat Int httpsdoiorg101016jquaint201701022
Howarth RW 2004 Human acceleration of the nitrogen cycle Drivers
consequences and steps toward solutions Water Sci Technol httpsdoiorg1010382Fscientificamerican0490-56
Jenny B Valero-Garceacutes BL Urrutia R Kelts K Veit H Appleby PG Geyh
M 2002 Moisture changes and fluctuations of the Westerlies in Mediterranean Central Chile during the last 2000 years The Laguna Aculeo record (33deg50primeS) Quat Int 87 3ndash18 httpsdoiorg101016S1040-
63
6182(01)00058-1
Jenny B Wilhelm D Valero-Garceacutes BL 2003 The Southern Westerlies in
Central Chile Holocene precipitation estimates based on a water balance model for Laguna Aculeo (33deg50primeS) Clim Dyn 20 269ndash280 httpsdoiorg101007s00382-002-0267-3
Leng M J Lamb A L Heaton T H Marshall J D Wolfe B B Jones M D
amp Arrowsmith C 2006 Isotopes in lake sediments In Isotopes in palaeoenvironmental research (pp 147-184) Springer Dordrecht
Le Roux JP Nielsen SN Kemnitz H Henriquez Aacute 2008 A Pliocene mega-
tsunami deposit and associated features in the Ranquil Formation southern Chile Sediment Geol 203 164ndash180 httpsdoiorg101016jsedgeo200712002
Lehmann M Bernasconi S Barbieri a McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66 3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Lenihan JM Drapek R Bachelet D Neilson RP 2003 Climate change
effects on vegetation distribution carbon and fire in California Ecol Appl httpsdoiorg101890025295
McLauchlan K K Gerhart L M Battles J J Craine J M Elmore A J
Higuera P E amp Perakis S S (2017) Centennial-scale reductions in nitrogen availability in temperate forests of the United States Scientific reports 7(1) 7856 httpsdoi101038s41598-017-08170-z
Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32ordmS 3050 masl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
Martiacuten-Foreacutes I Casado MA Castro I Ovalle C Pozo A Del Acosta-Gallo
B Saacutenchez-Jardoacuten L Miguel JM De 2012 Flora of the mediterranean basin in the chilean ESPINALES Evidence of colonisation Pastos 42 137ndash160
Marzecovaacute A Mikomaumlgi a Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54 httpsdoiorg103176eco2011105
Matesanz S Valladares F 2014 Ecological and evolutionary responses of
Mediterranean plants to global change Environ Exp Bot httpsdoiorg101016jenvexpbot201309004
64
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Menzel P Gaye B Wiesner MG Prasad S Stebich M Das BK Anoop A
Riedel N Basavaiah N 2013 Influence of bottom water anoxia on nitrogen isotopic ratios and amino acid contributions of recent sediments from small eutrophic Lonar Lake central India Limnol Oceanogr 58 1061ndash1074 httpsdoiorg104319lo20135831061
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Michelutti N Wolfe AP Cooke CA Hobbs WO Vuille M Smol JP 2015
Climate change forces new ecological states in tropical Andean lakes PLoS One httpsdoiorg101371journalpone0115338
Muntildeoz-Rojas M De la Rosa D Zavala LM Jordaacuten A Anaya-Romero M
2011 Changes in land cover and vegetation carbon stocks in Andalusia Southern Spain (1956-2007) Sci Total Environ httpsdoiorg101016jscitotenv201104009
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Poraj-Goacuterska A I Żarczyński M J Ahrens A Enters D Weisbrodt D amp
Tylmann W (2017) Impact of historical land use changes on lacustrine sedimentation recorded in varved sediments of Lake Jaczno northeastern Poland Catena 153 182-193 httpdxdoiorg101016jcatena201702007
Pongratz J Reick CH Raddatz T Claussen M 2010 Biogeophysical versus
biogeochemical climate response to historical anthropogenic land cover change Geophys Res Lett 37 1ndash5 httpsdoiorg1010292010GL043010
Prudhomme C Giuntoli I Robinson EL Clark DB Arnell NW Dankers R
Fekete BM Franssen W Gerten D Gosling SN Hagemann S Hannah DM Kim H Masaki Y Satoh Y Stacke T Wada Y Wisser D 2014 Hydrological droughts in the 21st century hotspots and uncertainties from a global multimodel ensemble experiment Proc Natl Acad Sci httpsdoiorg101073pnas1222473110
65
Ruumlhland KM Paterson AM Smol JP 2015 Lake diatom responses to
warming reviewing the evidence J Paleolimnol httpsdoiorg101007s10933-
015-9837-3
Sarricolea P Meseguer-Ruiz Oacute Serrano-Notivoli R Soto M V amp Martin-Vide
J (2019) Trends of daily precipitation concentration in Central-Southern Chile Atmospheric research 215 85-98 httpsdoiorg101016jatmosres201809005
Sernageomin 2003 Mapa geologico de chile version digital Publ Geol Digit Serrano-Notivoli R de Luis M Begueriumlia S 2017 An R package for daily
precipitation climate series reconstruction Environ Model Softw httpsdoiorg101016jenvsoft201611005
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Stine S 1994 Extreme and persistent drought in California and Patagonia during
mediaeval time Nature 369 546ndash549 httpsdoiorg101038369546a0
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL
Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Syvitski JPM Voumlroumlsmarty CJ Kettner AJ Green P 2005 Impact of humans
on the flux of terrestrial sediment to the global coastal ocean Science 308(5720) 376-380 httpsdoiorg101126science1109454
Thomas RQ Zaehle S Templer PH Goodale CL 2013 Global patterns of
nitrogen limitation Confronting two global biogeochemical models with observations Glob Chang Biol httpsdoiorg101111gcb12281
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Urbina Carrasco MX Gorigoitiacutea Abbott N Cisternas Vega M 2016 Aportes a
la historia siacutesmica de Chile el caso del gran terremoto de 1730 Anuario de Estudios Americanos httpsdoiorg103989aeamer2016211
Vargas G Ortlieb L Chapron E Valdes J Marquardt C 2005 Paleoseismic
inferences from a high-resolution marine sedimentary record in northern Chile
66
(23degS) Tectonophysics httpsdoiorg101016jtecto200412031 399(1-4) 381-398httpsdoiorg101016jtecto200412031
Valero-Garceacutes BL Latorre C Morelliacuten M Corella P Maldonado A Frugone M Moreno A 2010 Recent Climate variability and human impact from lacustrine cores in central Chile 2nd International LOTRED-South America Symposium Reconstructing Climate Variations in South America and the Antarctic Peninsula over the last 2000 years Valdivia p 76 (httpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-yearshttpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-years
Vicente-Serrano SM Gouveia C Camarero JJ Begueria S Trigo R
Lopez-Moreno JI Azorin-Molina C Pasho E Lorenzo-Lacruz J Revuelto J Moran-Tejeda E Sanchez-Lorenzo A 2013 Response of vegetation to drought time-scales across global land biomes Proc Natl Acad Sci httpsdoiorg101073pnas1207068110
Villa Martiacutenez R P 2002 Historia del clima y la vegetacioacuten de Chile Central
durante el Holoceno Una reconstruccioacuten basada en el anaacutelisis de polen sedimentos microalgas y carboacuten PhD
Villa-Martiacutenez R Villagraacuten C Jenny B 2003 Pollen evidence for late -
Holocene climatic variability at laguna de Aculeo Central Chile (lat 34o S) The Holocene 3 361ndash367
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Human alteration of the global nitrogen cycle sources and consequences Ecological applications 7(3) 737-750 httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Rein B Urrutia R amp Appleby P (2009) A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 The Holocene 19(6) 873-881 httpsdoiorg1011770959683609336573
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Widory D Kloppmann W Chery L Bonnin J Rochdi H Guinamant JL
2004 Nitrate in groundwater An isotopic multi-tracer approach J Contam Hydrol 72 165ndash188 httpsdoiorg101016jjconhyd200310010
67
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
68
Supplementary material
Facie Name Description Depositional Environment
F1 Organic-rich
mud
Massive to banded black
organic - rich (TOC up to 14 )
mud with aragonite in dm - thick
layers Slightly banded intervals
contain less OM (TOClt4) and
aragonite than massive
intervals High MnFe (oxic
bottom conditions) High CaTi
BrTi and BioSi (up to 5)
Distal low energy environment
high productivity well oxygenated
and brackish waters and relative
low lake level
F2 Massive to
banded silty clay
to fine silt
cm-thick layers mostly
composed by silicates
(plagioclase quartz cristobalite
up to 65 TOC mean=23)
Some layers have relatively high
pyrite content (up to 25) No
carbonates CaTi BrTi and
BioSi (mean=48) are lower
than F1 higher ZrTi (coarser
grain size)
Deposition during periods of
higher sediment input from the
watershed
69
F3 Banded to
laminated light
brown silty clay
cm-thick layers mostly
composed of clay minerals
quartz and plagioclase (up to
42) low organic matter
(TOC mean=13) low pyrite
and BioSi content
(mean=46) and some
aragonite
Flooding events reworking
coastal deposits
F4 Laminated
coarse silts
Thin massive layers (lt2mm)
dominated by silicates Low
TOC (mean=214 ) BrTi
(mean=002) MnFe (lt02)
TIC (lt034) BioSi
(mean=46) and TS values
(lt064) and high ZrTi
Rapid flooding events
transporting material mostly
from within the watershed
F5 Breccia with
coarse silt
matrix
A 17 cm thick (80-97 cm
depth) layer composed by
irregular mm to cm-long ldquosoft-
clastsrdquo of silty sediment
fragments in a coarse silt
matrix Low CaTi BrTi and
MnFe ratios and BioSi
Rapid high energy flood
events
70
(mean=43) and high ZrTi
(gt018)
Table Sedimentological and compositional characteristics of Laguna Matanzas
facies
71
CAPIacuteTULO 2 STABLE ISOTOPES TRACK LAND USE AND COVER
CHANGES IN A MEDITERRANEAN LAKE IN CENTRAL CHILE OVER THE
LAST 600 YEARS
72
Stable isotopes track land use and cover changes in a mediterranean lake in
central Chile over the last 600 years
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugone-Aacutelvarezabc Pablo
Sarricolead Leonardo Villaciacutese M Laura Carrevedob Blas Valero-Garceacutescg
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Departamento de Ecologiacutea Universidad de ChileLas Palmeras 3425 Nuntildeoa Santiago Chile
f Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding authors E-mail addresses clatorrebiopuccl magdalenafuentealbagmailcom
Key words Anthropocene lake sediment δsup1⁵N δsup1sup3C Great Acceleration organic
geochemistry watershedndashlake system Stable Isotope Analyses land usecover
change Nitrogen cycle mediterranean ecosystems central Chile
73
Abstract
Nutrient fluxes in many aquatic ecosystems are currently being overridden by
anthropic controls especially since the industrial revolution (mid-1800s) and the
Great Acceleration (mid-1900s) Land use and cover changes (LUCC) alter the
availability and fluxes of nutrients such as nitrogen that are transferred via runoff
and groundwater into lakes By altering lake productivity and trophic status these
changes are often preserved in the sedimentary record Here we use
geolimnological proxies and stable isotope analyses (δsup1⁵N δsup13C) on lake sediments
to reconstruct the lake-watershed nutrient transfer in a central Chilean lake (Lago
Vichuqueacuten) over the past 600 years We compare our multiproxy record from recent
lake sediments to the soilvegetation relationship across the watershed as well as
land usecover changes from 1975 to 2014 derived from satellite images Our results
show that lake sediment δsup1⁵N values increased with meadow cover but decreased
with tree plantations suggesting increased nitrogen retention when trees dominate
the watershed Our δsup1⁵N record from central Chile can thus be interpreted as a proxy
for nutrient availability over the last 600 years mainly derived from land use changes
coupled with climate drivers Although variable sources of organic matter and in situ
fractionation often hinder straightforward environmental interpretations of stable N
isotope signatures our study shows that lake sediment δsup1⁵N can be a useful tool for
assessing the contribution of past human activities in nutrient and nitrogen cycling
1 Introduction
Nitrogen (N) is a limiting nutrient in terrestrial and aquatic ecosystems (Vitousek
et al 1997) Changes in its availability can drive eutrophication and increase
pollution in these ecosystems (McLauchlan et al 2013) Although recent human
74
impacts on the global N cycle have been significant the consequences of increased
anthropogenic N on these ecosystems are unclear (Gerhart and Mclauchlan 2014
Battye et al 2017) Isotopic signatures are key to understand N dynamics in lakes
nevertheless in situ andor diagenetic fractionation along with multiple sources of
organic matter (OM) often hinder straightforward environmental interpretations from
isotopes Monitoring δ15N and δ13C values as components of the N cycle
specifically those related to the link between terrestrial and aquatic ecosystems can
help differentiate between effects from processes versus sources in stable isotope
values (eg from Particulate Organic Matter -POM- soil and vegetation) and
improve how we interpret variations in δ15N (and δ13C) values at longer temporal
scales
The main processes controlling stable N isotopes in bulk lake OM are source
lake productivity diagenesis and denitrification (Talbot 2001 Leng et al 2006
Fariacuteas et al 2009 Torres et al 2012 Brahney et al 2014) N sources depend on
contributions from the watershed (ie soil and biomass) the transfer of atmospheric
N to the lake (ie atmospheric N2 fixation) and nutrient recycling (Leng et al 2006)
Atmospheric N2 (isotopically defined as δ15N =0) is fixed by cyanobacteria with
minimal fractionation resulting in δ15NOM values that fluctuate from -2 to +2permil (Fogel
and Cifuentes 1993 Talbot 2001 Elliot and Brush 2006) Besides N2 fixation by
cyanobacteria phytoplankton species assimilate dissolved inorganic nitrogen (DIN)
and values typically range from +7 to +10permil (Xu et al 2016 Meyers et al 2003) In
addition seasonal changes in POM occur in the lake water column Gu et al (2006)
sampled the water column of Lake Wauberg (Florida USA) during a hydrologic year
and found a higher development of N fixing species during the summer A major
factor behind this increase are human activities in the watershed which control the
75
inputs of the sediment and nutrients to the lake (Woodward et al 2011) Some
studies have shown higher δ15N values in lake sediments from watersheds that are
highly affected by human activities (Bruland and Mackenzie 2010 Woodward et al
2011) Syntenic fertilizers displayed δ15N values between -4 to +4permil cattle manure
around +8 to +22permil and surface and groundwater OM between 0 to +10permil (Elliott
and Brush 2006 Leng et al 2006) Although relatively low δ15N values from
fertilizers constitute major N input to human-altered watersheds the elevated loss
of 14N via volatilization of ammonia and denitrification leaves the remaining total N
input enriched in 15N (Bruland and Mackenzie 2010)
In addition to the different sources and variations in lake productivity early
diagenesis at the sedimentndashwater interface in the sediment can further alter
sediment OM isotope values (Brahney et al 2014 Galman et al 2009) During
diagenetic loss of N in lacustrine systems kinetic fractionation occurs and the
remaining N is enriched in 15N leading to higher δ15N values (Galman et al 2006
Talbot 2001) Denitrification tends to enrich lake sediments in 15N due to the
assimilation of nitrates by denitrifying bacteria (Granger et al 2008) and is more
prevalent in anoxic environments (Brahney et al 2016 Lehman et al 2002)
Carbon isotopes in lake sediments can also provide useful information about
paleoenvironmental changes OM origin and depositional processes (Meyers et al
2003) Allochthonous organic sources (high CN ratios) produce isotope values
similar to values from catchment vegetation Autochthonous organic matter (low CN
ratio) is influenced by fractionation both in the lake and the watershed leading up to
carbon assimilation by lacustrine biota (Woodward et al 2011) Changes in
productivity greatly affect δ13COM values (Torres et al 2012) as during C uptake
plankton preferentially assimilate 12C leaving the dissolved inorganic carbon (DIC)
76
pool enriched in 13C (Gu et al 2006) with phytoplankton averaging ca 20 permil lower
than the respective δ13CDIC values (Leng et al 2006) Under conditions of low to
moderate primary productivity plankton preferentially uptake the lighter 12C
resulting in low δ13C values displayed by the OM (Torres et al 2012) Conversely
during high primary productivity phytoplankton will uptake 12C until its depletion and
is then forced to assimilate the heavier isotope resulting in an increase in δ13C
values Higher productivity in C-limited lakes due to slow water-atmosphere
exchange of CO2 also results in high δ13C values (Galman et al 2009) In these
cases algae are forced to uptake dissolved bicarbonate with δ13C values between
7 to 10permil higher than those of the atmospheric CO2 dissolved in water (Sun et al
2016 Torres et al 2012 Galman et al 2009)
Stable isotope analyses from lake sediments are thus useful tools to
reconstruct shifts in lake-watershed dynamics caused by changes in limnological
parameters and LUCC Our knowledge of the current processes that can affect
stable isotope signals in a watershed-lake system is limited however as monitoring
studies are scarce Besides in order to use stable isotope signatures to reconstruct
past environmental changes we require a multiproxy approach to understand the
role of the different variables in controlling these values Hence in this study we
carried out a detailed survey of current N dynamics in a coastal central Chilean lake
(Lago Vichuqueacuten) and a multiproxy study of a sediment record spanning the last
600 years The characterization of the recent changes in the watershed since 1970s
is based on satellite images to compare recent changes in the lake and assess how
these are related with climate variability and an ever increasing human footprint
(Lara et al 2012 Gallardo et al 2015 Steffen et al 2015) Our main goal is to
investigate how stable isotope values from lake sediment reflect changes in the lake
77
ndash watershed system during periods of high watershed disruption (eg Spanish
Conquest late XIX century Great Acceleration) and recent climate change (eg
Little Ice Age and current global warming)
2 Study Site
Lago Vichuqueacuten (34ordm49`S 72ordm04W 4 m asl 30 m of depth Fig 1) is a
mesotrophic to eutrophic lake with dominant anoxic bottom conditions that is
stratified year around (Frugone-Aacutelvarez et al 2017) The main inlets are the
Vichuqueacuten and Huintildee rivers and the only outlet is the Llico estuary that drains into
the Pacific Ocean High tides can sporadically shift the flow direction of the Llico
estuary which increases the marine influence in the lake Dune accretion gradually
limited ocean-lake connectivity until the estuary was almost completely closed off
by ca 10 ky BP and the current lake regime formed (Frugone-Aacutelvarez et al 2017)
The area is characterized by a mediterranean climate with cold-wet winters and
hot-dry summers and an annual precipitation of ~650 mm and a mean annual
temperature of 15ordmC During the austral winter months (June - August) precipitation
is modulated by north-west shifts of the South Pacific Anticyclone (SPA) followed by
an increased frequency of storm fronts stemming off the South Westerly Winds
(SWW) A strengthened SPA during austral summers (December - March) which
are typically dry and warm blocks the northward migration of storm tracks stemming
off the SWW (Fig1 Frugone-Aacutelvarez et al 2017)
78
Figure 1 a) The location of the Lago Vichuqueacuten in central Chile b) Map of land
uses and cover in 2016 c) Ombrothermic diagram mediterranean climates are
characterized by cold-wet winters with surplus moisture from June to August and
hot-dry summers d) Lake bathymetry showing location of cores and water sampling
sites used in this study
Although major land cover changes in the area have occurred since 1975 to the
present as the native forests were replaced by tree (Monterey pine and eucalyptus)
plantations the region was settled before the Spanish conquest (Frugone-Alvarez
et al 2018) Historical documents indicate that Vichuqueacuten was occupied by a
Mitimaes colony as part of the Inka empire (Odone 1998) Unlike other Andean
areas during the Inka period the introduction of corn and quinoa in the Vichuqueacuten
watershed do not seem to have intensified land use The Spanish colonial period in
Chile lasted from 1542 CE to the independence in 1810 CE The first historical
document (1550 CE) shows that the areas around Vichuqueacuten were settled by the
Spanish early on as the Vichuqueacuten watershed was given as an ldquoencomiendardquo
system to Juan Cuevas The ldquoencomiendardquo system provided the owners with land
79
and indigenous people to work but also the introduction of wheat wine cattle
grazing and logging of native forests for lumber extraction and increasing land for
agricultural activities (Odone et al 1998 Armesto et al 2010) During the 19th
century (the Republic) the export of wheat to Australia and Canada generated
intensive changes in land cover use The town of Vichuqueacuten became the regional
capital and plans to build a maritime port in the bay of Vichuqueacuten were drawn
However the fall of international markets in 1880 paralyzed these plans During the
20th century the plantation of trees (Pinus radiata and Eucalyptus globulus) in areas
cleared of native forests was subsidized by two national laws DFL nordm265 (1931) and
DFL nordm 701 (1974) both of which provided funds for such plantations During the last
decades the urbanization with summer vacation homes along the shorelines of
Lago Vichuqueacuten has greatly increased Extensive lake eutrophication is currently a
large environmental problem (EULA 2008)
3 Methods
Coring campaigns were organized from 2011 to 2015 (Fig 1d) We recovered
12 short cores and 4 long cores (up to 14 m long) at several sites using a hammer-
modified UWITEC gravity corer Sediment cores (VIC11-2A 170 cm VIC13-2B 170
cm VIC13-2D 1200 cm and VIC15-1A 119 cm) were split photographed sub-
sampled and stored in the Pyrenean Institute of Ecology (IPE-CSIC Spain) Core
VIC13-2B was selected for detailed multiproxy analyses (including elemental
geochemistry C and N isotopes XRF and 14C dating) Stable isotope analyses
(δ13C δ15N) were performed at the Laboratory of Biogeochemistry and Applied
Stable Isotopes (LABASI-PUC Chile) Sediment samples for δ13Corg were pre-
treated with 50 ml of HCl and 50 ml of deionized water and dried at 60ordmC for 4 hr to
remove carbonates (Harris et al 2011) Isotope analyses were conducted using a
80
Delta V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via
a Conflo IV interface Isotope results are expressed in standard delta notation (δ)
and expressed in per mil (permil) relative to the Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Total carbon (TC)
Total Organic Carbon (TOC) Total Inorganic Carbon (TIC) and Total Sulfur (TS)
were measured every 2 cm with a LECO SC 144 DR at the IPE-CSIC
An AVAATECH X-Ray Fluorescence II core scanner with a Rh X-ray tube from
the University of Barcelona was used to obtain XRF logs every 4 mm of resolution
Results are expressed as element intensities in counts per second (cps) Tube
voltage was operated at 30 kV and 10 kV to obtain the abundances of 18 elements
(Pb Al Si S Cl K Ca Ti V Mn Fe Co Zn Br Rb Sr Y Zr) with an average of
at least of 1800 cps (less for Pb=437 and Zn=934 cps) Pb was included due to
similar behavior with Co and Fe Element ratios were calculated to describe changes
in redox conditions (MnFe) endogenic productivity (BrTi) carbonate formation
(SrCa and CaTi) and sediment input (ZrTi) (Barreiro-Lostres et al 2014
Carrevedo et al 2015 Frugone-Aacutelvarez et al 2017 Kylander et al 2011 Moreno
et al 2007a)
Several campaigns were carried out to sample the POM from the water column
two per hydrologic year from November 2015 to August 2018 A liter of water was
recovered in three sites through to the lake two are from the shallower areas (with
samples taken at 2 and 5 m depth at each site) and one in the deeper central portion
(at 2 5 and 20 m depth) of the lake (Fig 1d) The samples were filtered using glass
fiber filters (25 m) and then frozen to obtain the stable carbon and nitrogen isotope
signal of lacustrine POM Additionally soil and vegetation samples from the
following communities native species meadow hydrophytic vegetation and
81
Monterrey pine (Pinus radiata) were taken for stable isotope analyses (see in
supplementary material)
The age model for the complete Lago Vichuqueacuten sedimentary sequence is
based on Frugone-Aacutelvarez et al (2017) with the last 1000 years based on
210Pb137Cs dating techniques in VIC15-1A and 14C AMS dates on bulk sediment
samples (Supplementary Table S1) The 14C measurements of lake water DIC show
a moderate reservoir effect of 180 plusmn 25 14C yr) The revised age-depth model used
here includes three more 14C AMS dates performed with the program Bacon to
establish the deposition rates along the core (Blaauw and Christen 2011 Fig 2)
The age-depth model indicates that average resolution between 0 to 87 cm is lt2
cm per year and from 88 to 170 cm it is lt47 cm per year
82
Figure 2 Chronological age-depth model for the Lago Vichuqueacuten sedimentary
sequence spanning the last 1000 yr (see also Frugone-Alvarez et al 2017)
To estimate land use changes in the watershed we use Landsat MSS images
for 1975 and Landsat TM for 1989 and 2014 all taken in either summer or autumn
(Table 1) We performed supervised classification of land uses (maximum likelihood
83
algorithm) for each year (1975 1989 and 2014) and results were mapped using
ArcGIS 102
Table 1 Images using for LUCC reconstruction
Source of LUCC
Acquisition
Date Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat TM 19991226 30 m
CONAF 2009 30 m
Land cover Chile 2014 30 m
CONAF 2016 30 m
Previous Work on Lago Vichuqueacuten sedimentary sequence
The sediments are organic-poor dark brown to brown laminated silt with some
intercalated thin coarser clastic layers Lacustrine facies have been classified
according to elemental composition (TOC TS TIC and TN) grain size and
sedimentary textures (Frugone-Aacutelvarez et al 2017) Four main types of lacustrine
facies were identified in this short core Facies L1 is a laminated (1cm) black to dark
brown diatom-poor silt with relatively higher TOC percentages (TOC about 185)
TS (about of 18) and TOCTN (about 92) Facies L2 is characterized by a
homogeneous black organic-poor silty clay with low TOC (mean= 13) TS (mean=
13) TOCTN (mean= 88) values Facies L3 is a laminated (1cm) black organic-
poor (TOC mean 14) silts with higher TS (mean= 21) and TOCTN ratios
(mean= 88) Facies L3 is interpreted as ldquobaselinerdquo deposition on the distal areas
84
of Vichuqueacuten lake Mineralogy is dominated by mica (muscovite) clay minerals
(chlorite) and plagioclase (albite) Quartz and calcite occur at the top and pyrite
occurs in the lower part of the sequence Facies T is composed by massive banded
sediments 4 cm thick and coarse grain size It is interpreted as an instantaneous
depositional event (for more detail see Frugone-Aacutelvarez et al 2017) In this work
we identified four subunits based on geochemical and stable isotope signals
4 Results
41 Geochemistry and PCA analysis
High positive correlations exist between Al Si K and Ti (r = 078 ndash 096
supplementary Fig S1) and between Zn Zr Rb and Y (r = 072 - 096) which reflect
the silicate content of the watershed-derived sediments (Marzecoacuteva et al 2011) Zr
is commonly associated with minerals more abundant in coarser deposits Thus the
ZrTi profile could reflect grain size (Cuven et al 2010) showing higher variability
in the upper part of the Lake Vichuqueacuten sequence and in the alternation between
facies L1 and L2 Another group of elements made up of Co Fe and Pb displayed
positive correlations (r = 067ndash 097) and represents the input of heavy metals Br
Cl and Ca show a positive correlation (r = 049 ndash 06 p-value = 00001) BrTi ratio
is interpreted as a productivity indicator due to Br having a strong affinity with humic
and organic compounds (Marzecoacuteva et al 2011 Frugone-Aacutelvarez et al 2017) In
our Lago Vichuqueacuten sequence BrTi values are high from 170 to 132 cm and from
36 cm to the top of core where it shows several sharps peaks (Fig 3) The MnFe
ratio is indicative of lake bottom oxygenation (Naecher et al 2013) as under
reducing conditions Mn tends to become more mobile than Fe leading to a
decreased MnFe (Marzecoacuteva et al 2011) Several sharp peaks in MnFe occurred
85
from 127 to 120 cm and from 57 to 30 cm (MgFegt03) The silicate group and the
Br Cl Ca Mn group are negatively correlated (r= -012 and -066)
Principal Component Analysis (PCA) was undertaken on the XRF
geochemical data to investigate the main factors controlling sediment deposition in
Lago Vichuqueacuten (Fig 3) The four main eigenvectors explain 801 of the variance
(supplementary material Table S2) The principal component (PC1) explain 437
of the total of variance and grouped elements are associated with terrigenous input
to the lake Positive values of the biplot have been attributed to higher heavy metals
deposition (Pb Co and Fe) and negative values to higher silicate input (Al K Ti and
Si) in a high energy catchment environment (Zr) The PC2 explains 198 of the
total of variance and highlights the endogenic productivity in the lake The positive
loading is compound by Br Cl Ca and Sr and to a lesser extent by Y Zn Rb and
Mn The PC2 may explain both the precipitation of carbonates (Ca Sr) and biological
production (Br)
86
Figure 3 Principal Component Analysis of XRF geochemical measurements in
VIC13-2B Lago Vichuqueacuten lake sediments
42 Sedimentary units
Based on geochemical and stable isotope analysis we identified four
lithological subunits in the short core sedimentary sequence Our PCA analyses and
Pearson correlations pointed out which variables were better for characterizing the
subunits (Fig 4) in terms of endogenic productivity heavy metals and terrestrial
input The unit 2c (170-131 cm) is characterized by the occurrence of some clastic
layers (L2 from 160 and 150 cm) high δsup1⁵N values (mean= +59 plusmn 04permil) with
Magnetic Susceptibility (MS) BrTi ZrTi and SrCa values increase towards the top
Also it has relatively low δsup1sup3C values (mean= -265 plusmn 11permil) and CNmolar ratios
(mean=78 plusmn 04) The ZrTi show upwards increasing values reaching the highest
values of the sequence at the top of this unit suggesting a coarsening upward trend
and relatively higher depositional energy The MS trend also indicates higher
erosion in the watershed and enhanced delivery of ferromagnetic minerals likely
from soil particles particularly from 150 to 160 cm (Frugone-Aacutelvarez at al 2017)
The subunit 2b (130-118 cm) is also composed of black silts but it has the
lowest MS values of the whole sequence and its onset is marked by a sharp
decrease in ZrTi This subunit is marked by sharp peaks of MnFe (at 126 and 120
cmgt027) SrCa (at 125 and 120 cmgt033) CaTi (at 124 and 120 cmgt199) TOC
(about 26 at 130 124 122 and 118 cm) and δsup1⁵N (gt+67permil at 126 and 120 cm)
BrTi oscillates around low values (about 01) whereas the δsup1sup3C values range
between -262 and -282permil
87
The unit 2a (58-117 cm) shows increasing and then decreasing MS values
and displayed sharps peaks of δsup1⁵N (gt64 at 102 to 99 87 and 75 cm) and CN
(about 10 at 86 74 66 and 60 cm) SrCa (mean = 04 plusmn 013) BrTi (mean = 008
plusmn 002) MnFe (mean = 002 plusmn 001) and δsup1sup3C (mean = -260 plusmn 07permil) oscillate in
low values The upper part of this unit (72 ndash 57 cm) shows an increase of SrCa
(from 03 to 05)
The upper unit 1a (0 - 57 cm) starts with a fine clastic layer (T at 57 to 54
cm) interpreted as deposition during a high-energy event It is characterized by
lower BrTi (mean = 003 plusmn 002) TOC (mean = 12 plusmn 03) and δsup1sup3C (mean= -
266 plusmn 06permil) higher δsup1⁵N (mean = +55 plusmn 08 permil) This unit is made of alternating
fine (lt 1cm) black and dark brown laminae with carbonate (L1 facies) Recently
deposited sediments show decreasing MS values increasing TOC (mean= 15 plusmn
04 ) sharp peaks of BrTi (gt015 at 33 21 5 and 1 cm) and the lowest δsup1⁵N values
of the entire record (mean= 45 plusmn 06 permil) and δsup1sup3C values (mean= -277 plusmn 09 permil)
Heavy metal content increases abruptly in the upper section of the core (2 - 20 cm)
(peaks of FeTi CoTi and PbTi)
88
Figure 4 Sedimentary units of Lago Vichuqueacuten sedimentary sequence Selected
variables illustrate watershed input (Magnetic Susceptibility ZrTi CoTi and MnFe)
endogenic productivity (SrCa CaTi TS) organic matter source (BrTi TOC
CNmolar and stable isotope records (δ13Corg and δ15Nbulk)
43 Recent seasonal changes of particulate organic matter on water column
The average of δ15NPOM from the three sites across to the lake was +86 plusmn 58
permil oscillating widely from -14 to +193permil (Fig 5) Important seasonal differences
occur at 2 and 5 m depth with more negative values during summer (~53 plusmn 52 permil)
than winter (~122 plusmn 40permil) At 20 m depth δ15NPOM values exhibit small seasonal
ranges (mean = +10permil for winter and +87permil for summer) The δ13CPOM average was
-296 plusmn 33permil with slightly seasonal and water column depth differences However
more positive δ13CPOM values occurred during winter (mean=-290 plusmn 41permil) than in
summer (mean= -301 plusmn 19 permil) Like the δ15NPOM values CNPOM (mean= 83 plusmn 17)
displayed important seasonal and water depth differences Lower CNPOM ratios
89
occurred during summer at 2 and 5 m depth (mean= 95 plusmn 17) and higher and more
constant values during winter (mean = 73 plusmn 09) at the same depth At 20 m CNPOM
shows similar values in both winter (70) and summer (74)
Figure 5 Seasonal POM averages from samples taken of the Lago Vichuqueacuten
water column Samples were taken from 2015 to 2018 at 2 (n=16) 5 (n=14) and 20
(n=8) meters depth
44 Stable isotope values across the Lake Vichuqueacuten watershed
Figure 6 shows modern vegetation soil and sediment isotope values found for
the watershed Huintildee tributary and Vichuqueacuten lake The δsup1⁵NFoliar values from
meadow plantations and macrophytes have similar range values with a mean of
+89 plusmn50 permil (n=18) +74 plusmn 75 permil (n=5) +82 plusmn 36 permil (n=6) respectively Native
vegetation has overall lower δsup1⁵NFoliar values with a mean of 14 plusmn 21 permil (n=28 see
Table 2 in supplementary material) The δsup1sup3CFoliar from terrestrial samples exhibit
similar values across the different plant communities (tree plantation mean=-274 plusmn
13permil meadow mean = -275 plusmn 23 permil native species mean=-272 plusmn 14permil) whereas
macrophytes display slightly more negative values with a mean of -287 plusmn 23permil
Macrophytes substrate displayed the highest δsup1⁵N values with a mean of +60 plusmn
14permil (n=3) Similar and relatively high δsup1⁵N values for soil of plantation (mean= +54
plusmn 13permil n=4) meadow (mean= +49 plusmn 04permil n=8) and surface river sediment
90
(mean= +52 plusmn 17permil n=10) Native forest soils oscillate around slightly more
negative values with a mean of +45 plusmn 12permil (n=4) Slightly more positive soil δsup1sup3C
values occur both underneath native forests and in tree plantations with means of -
284 plusmn 23permil and -298 plusmn 13permil respectively compared to either meadow soils
(mean= -302 plusmn 11permil) aquatic sediments underneath macrophytes (-311 plusmn 10permil)
or from surface river sediments (mean= -312 plusmn 10permil)
Figure 6 Stable isotope content of modern soil river sediment and foliar vegetation
used as end members in the sedimentary sequence of Lago Vichuqueacuten a)
Scatterplot of foliar δsup1⁵N and δsup13C values for vegetation from Lago Vichuqueacuten
watershed (plantation meadow and native species) and macrophytes on Lake
Vichuqueacuten See supplementary material for more detail of vegetation types b)
Scatterplot of soil δsup1⁵N and δsup13C values across the watershed the sediments of the
Huintildee and Vichuqueacuten rivers and the fluvial sediment corresponding to the
macrophyte vegetation
45 Land use and cover change from 1975 to 2014
Major land use changes between 1975 CE and 2016 CE in the Lago
Vichuqueacuten watershed are summarized in Figure 7 The watershed has a surface
area of 535329 km2 of which native vegetation (26) and shrublands (53)
represent 79 of the total surface in 1975 Meadows are confined to the valley and
91
represent 17 of watershed surface Tree plantations initially occupied 1 of the
watershed and were first located along the lake periphery By 1989 the areas of
native forests shrublands and meadows had decreased to 22 31 and 14
respectively whereas tree plantations had expanded to 30 These trends
continued almost invariably until 2016 when shrublands and meadows reached 17
and 5 of the total areas while tree plantations increased to 66 Native forests
had practically disappeared by 1989 and then increased up to 7 of the total area
in 2016 (Fig 6) preferentially occupying the southeastern sector of the watershed
Figure 7 Changes in land use and cover between 1975 and 2016 in the Lago
Vichuqueacuten watershed as measured from satellite images The major change is
represented by the replacement of native forest shrubland and meadows by
plantations of Monterrey pine (Pinus radiata)
Figure 8 shows correlations between lake sediment stable isotope values and
changes in the soil cover from 1975 to 2013 Positive relationships occurred
between sediment δsup1⁵N and δsup13C values from core VIC13-2B (unit 1) and the
92
percentage of native forest (r=089 for δsup1⁵N 082 for δsup13C) shrublands (r = 071 for
δsup1⁵N 057 for δsup13C) and meadow cover (r = 088 for δsup1⁵N 081 for δsup13C) All of these
correlations are significant (p value lt 0001) In contrast significant negative
correlations (p lt0001) occurred between tree plantation cover and lake sediment
stable isotope values (δsup1⁵N r = -082 and δsup13C r = -067) native forest (r = -097)
meadows (r = -086) and shrubland (r =-093)
Figure 8 Correlation plots of land use and cover change versus lake sediment
stable isotope values The δsup1⁵N values are positively correlated with native forests
agricultural fields and meadow cover across the watershed Total Plantation area
increases are negatively correlated with native forest meadow and shrubland total
area Significance levels are indicated by the symbols p-values (0 0001 001
005 01 1) lt=gt symbols ( )
93
5 Discussion
51 Seasonal variability of POM in the water column
The stable isotope values of POM can vary during the annual cycle due to
climate and biologic controls namely temperature and length of the photoperiod
which affect phytoplankton growth rates and isotope fractionation in the water
column (Gu et al 2006 Leng et al 2006) POM from Lago Vichuqueacuten surface
samples (2 and 5 m depth) displayed lower δ13CPOM values during the summer than
in winter During C uptake phytoplankton preferentially utilize 12C leaving the
DICpool enriched in 13C Therefore as temperature increases during the summer
phytoplankton growth generates OM enriched in 12C until this becomes depleted
and then the biomas come to enriched u At the onset of winter the DICpool is now
enriched in 13C and despite an overall decrease in phytoplankton production the
OM produced will also be enriched in 13C In contrast δ13CPOM values at 20 m depth
did not reflect these seasonal differences probably due to water-column
stratification that maintains similar temperatures and biological activity throughout
the year
Lake N availability depends on N sources including inputs from the
watershed and the atmosphere (ie deposition of N compounds and fixation of
atmospheric N2) which varies during the hydrologic year The fixation of atmospheric
N2 is an important natural source of N to the lake occurring mainly during the
summer season associated with higher temperature and light (Gu et al 2006)
Because the δ15NPOM of N2 fixers is close to 0permil with almost no overall isotope
fractionation (Leng et al 2006) when N2 is the major N source δ15NPOM values are
typically low However when DIN concentrations are high or alternatively when little
94
N2 fixation occurs δ15NPOM values will be higher (Gu et al 2006) δ15NPOM values
from summer Lago Vichuqueacuten samples were lower than those from winter with large
differences between the seasons (~5permil Fig 5 and 9) Furthermore δ15NPOM values
were high when monthly average temperature was low and monthly precipitation
was high (Fig 9) This pattern is consistent with the dominance of high N2 fixation
by cyanobacteria associated with increased summer temperatures This correlation
of δ15NPOM values with temperature further suggests a functional group shift i e
from N fixers to phytoplankton that uptake DIN The correlation between wetter
months and higher δ15NPOM values could be caused by increased N input from the
watershed due to increased runoff during the winter season The lack of data of the
δ15NPOM between the 30 mm and the 160 mm of precipitation is due to the
mediterranean-type climate that concentrates precipitations in the winter months
Finally Lago Vichuqueacuten is characterized by pronounced seasonality leading to
higher phytoplankton biomass in summer characterized by low δ15NPOM In winter
low biomass production and increased input from watershed is associated to high
δ15NPOM
95
Figure 9 Seasonal changes can drive δ15NPOM values in Lago Vichuqueacuten Data
correspond to average monthly temperature and total monthly precipitation for the
months when the water samples were taken (years 2015 - 2018) P-valuelt005
52 Stable isotope signatures in the Lake Vichuqueacuten watershed
The natural abundance of 15N14N isotopes of soil and vegetation samples
from the Lago Vichuqueacuten watershed appear to result from a combination of factors
isotope fractionation different N sources for plants and soil microorganisms (eg N2
fixation Nitrates) depth in the soil where N uptake by vegetation occurs and N loss
mechanisms (ie denitrification leaching and ammonia volatilization Hogberg
1997) The lowest δsup1⁵Nfoliar values are associated with native species and are
probably due to the contribution of N-fixing species (eg Sophora) (Fig6 and for
more detail see Table S3 in supplementary material) The number native N-fixers
species present in the Chilean mediterranean vegetation are not well known
however so there could be other N-fixing species in this group Alternatively δsup1⁵Nfoliar
values reflect soil N uptake (Kahmen et al 2008) In environments limited by N
plants have δsup1⁵N values similar to the soil δsup1⁵N values however the denitrification
and volatilization of ammonia can lead to the remain N of soil to come enriched in
15N (Cifuentes et al 1989 Craine et al 2009 Bruland and Mckenzie 2010) Soil N
isotope samples from native species communities tends to display relatively high
δsup1⁵N values respect to foliar samples due to loss of N-soil
The higher foliar and soil δsup1⁵N values obtained from samples of meadows
aquatic macrophytes and tree plantations can be attributed to the presence of
greater amounts of nitrate available for plant uptake (Fig 6) Houmlgberg (1997)
suggests that the availability of different N sources in soils (ie nitrates versus
96
ammonia) with different residence times can also explain these δsup1⁵NFoliar values
Indeed Feigin et al (1974) described differences of up to 20permil between ammonia
and nitrates sources Denitrification and nitrification discriminate much more against
15N than N2 fixation does (Craine et al 2009) leading to soils (and plants after
uptake) enriched in 14N
In general multiple processes that affect the isotopic signal result in similar
δsup1⁵N values between the soil of the watershed and the sediments of the river
However POM isotope fluctuations allow to say that more negative δsup1⁵N values are
associated to lake productivity while more positive δsup1⁵N values are associated with
N input from the watershed
δ13C values in Lago Vichuqueacuten watershed reflect CO2 gas exchange between
C3 plants and algae with the atmosphere During photosynthesis plants
discriminate against the heavier stable isotope of carbon (13C) in favor of the lighter
isotope 12C which results in δ13CFoliar values between -220permil and -320permil (Browman
and Cook 2002 Liu et al 2007) Likewise the δ13CFoliar values from Lago Vichuqueacuten
oscillate around -27permil (Fig 6) Plants transform atmospheric CO2 into soil organic
carbon (C) which in turn reflects this initial discrimination against 13C during C
uptake and post photosynthetic fractionation (Farquhar et al 1982 1989 Badeck
et al 2005 Pausch and Kuzyakov 2017) Slightly more positive δ13CFoliar values
(about 15permil) were measured in comparison with their δ13CSoil values This may be
reflecting the C transference from plants to the soil but also a soil-atmosphere
interchange The preferential assimilation of the light isotopes (12C) during soil
respiration carried by the roots and the microbial biomass that is associated with the
decomposition of litter roots and soil organic matter explain this differential
(Berhardt et al 2006 Blagodatskaya et al 2010 Midwood and Millard 2011)
97
In general the δ13CSoil values from the Lago Vichuqueacuten watershed oscillated
around -290permil and did not vary with our plant classification types Here we use
these values as terrestrial-end members to track changes in source OM (Fig 6)
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from the terrestrial watershed By the other hand more positive δ13C
values most likely reflect an increased aquatic OM component as indicated by POM
isotope fluctuations (Fig 9)
53 Recently land use and cover change and its influences on N inputs to the lake
Tree plantations in the Lago Vichuqueacuten watershed have increased greatly in
the last few decades (from 1 in 1975 to 66 in 2016 Fig 7) replacing previous
native forests (from 26 to 7) meadows (17 to 5) and shrublands (53 to
17) In 1975 tree plantations were confined to the lake perimeter with discrete
patches distributed throughout the watershed The ldquoForest Lawrdquo (DFL 701- passed
in 1974) allocated state funding to afforestation efforts and management of tree
plantations which greatly favored the replacement native forests by introduced trees
This increase is marked by a sharp and steady decrease in lake sediment δ15N and
δ13C values because tree plantations function as a nutrient sink whereas other land
uses such as meadows favor nutrient loss from the watershed (Fig8) Bruland and
Mackenzie (2014) noted a decrease in wetland δ15N values when watershed
forested cover increased and concluded that N inputs to the wetlands are lower from
the forested areas as they generally do not export as much N as agricultural lands
A positive correlation between native vegetation and δ15Ncore values can be
explained by the relatively scarce arboreal cover in the watershed in 1975 when
native forest occupied just 26 of the watershed surface whereas shrublands and
98
meadows occupied more than the 70 of the surface of the watershed with the
concomitant elevated loss of N (Fig 7 and Fig 8)
54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years
Sedimentological compositional and geochemical indicators all show changes in
the N cycle associated to LUCC lake dynamics and climate changes (Fig 10) From
the pre-Columbian indigenous settlement including the Spanish colonial period up
to the start of the Republic (1300 - 1800 CE) the introduction of crops such as
quinoa and wheat but also the clearing of land for extensive agriculture would have
favored the entry of N into the lake Conversely major changes observed during the
last century were characterized by a sharp decrease of N input that were coeval
with the increase of tree plantations
From 1365 until 1550 CE (unit 2c 2b and half of 2a Fig 3 and Fig 10) pre-
Columbian agricultural societies planted mostly crops of corn and quinoa (Ramirez
and Vidal 1985) This period is characterized by large peaks seen in the δsup1⁵N record
(eg 1403 1452 1476 and 1505) with overall high δsup1⁵N values (up to 5permil) indicating
that N input from watershed was elevated and oscillating to the beat of the NT These
positive δsup1⁵N peaks could be due to several causes including a) the clearing of land
for farming b) N loss via denitrification which would be generally augmented in
anoxic environments (Diebel et al 2009) such as Lago Vichuqueacuten (low MnFe
values about 0001 Fig 4) or c) climate especially cold-wet winters or hot-dry
summers can also exert control on the δsup1⁵N record Indeed tree-ring records and
summer temperature reconstructions show overall wetcold conditions during this
period (Christie et al 2009 Von Gunten et al 2009 Fig 10) Increased
precipitation would bring more sediment (and nutrients) from the watershed into the
99
lake and increase lake productivity which is also detected by the geochemical
proxies for productivity (peaks of BrTi SrCa and TOC Fig 4 and Fig 10) (see also
Frugone-Alvarez et al 2017)
Figure 10 Changes in the N availability during the last six centuries in Lago
Vichuqueacuten a) lake sediment δsup1⁵N values displayed more positive values during the
prehistoric period Spanish Colony and the starting 19th century which is associated
with enhanced N input from the watershed by extensive clearing and crop
plantations The inset shows this relationship between sediment δsup1⁵N and
100
percentage of meadow cover over the last 30 years b) Summer temperature
reconstruction from central Chile (von Gunten et al 2009) showing a
correspondence with cold phases and high δsup1⁵N values at Vichuqueacuten except for the
last 100 years c) Palmer Drought Severity Index (PDSI) is an interannual moisture
variability reconstruction for late springndashearly summer during the last six centuries
(Christie et al 2009) Grey shadow indicating higher precipitation periods
From 1550 and 1810 CE (unit 2a) and during the Colonial period several peaks
of δ15N could be further related not only to high precipitation (Fig 10 grey shadow)
but also pulses of enhanced N input from the watershed linked to human land use
In 1550 CE Juan Cuevas was granted lands and indigenous workers under the
encomienda system for agricultural and mining development of the Vichuqueacuten
village (Ramiacuterez and Vidal 1985) Historical documents show that around 1579 CE
the Vichuqueacuten watershed was occupied by indigenous communities dedicated to
wheat plantations and vineyards wood extraction and gold mining (Odone 1998)
The introduction of the Spanish agricultural system implied not just a change in the
types of crops used (from quinoa to vineyards and wheat) but also a clearing of
native species for the continuous increase of agricultural surface and wood
extraction During the Colonial period wheat from Vichuqueacuten was exported to Peru
(Ramiacuterez and Vidal 1985) Armesto et al (2009) indicate that during the XVIII and
XIX centuries the extraction of wood for mining operations was important enough to
cause extensive loss of native forests The independence and instauration of the
Chilean Republic did not change this prevailing system Increases in the
contributions of N to the lake during the second half of the XIX century (peaks in
δsup1⁵N ca 1870 CE) could be due to expanding farm land dedicated for wheat
101
production and increased commercial trade with California and Canada (Ramiacuterez
and Vidal 1985)
In contrast LUCC in the last century are clearly related to the development of
large-scale tree plantations (Fig 7) The δsup1⁵N record reaches the lowest values of
the entire sequence in the last few decades (Fig 10) A marked increase in lake
productivity NT concentration and decreasing sediment input is synchronous (unit
1 Fig 4) with trees replacing meadows shrublands and areas with native forests
(Fig 7 and 8) The implementation of the lsquoForest Lawrsquo in 1974 had a stronger impact
on the landscape and lake ecosystem dynamics than the impacts of ongoing climate
change in the region which is much more recent (Garreaud et al 2018) although
the prevalence of hot dry summers seen over the last decade would also be
associated with low δsup1⁵N values and increase of NT (Fig 10) Heavy metals ratios
(FeTi CoTi and PbTi) show notable increases in lake sediments from 1992 to 2011
CE (Fig 4) Although this could be related to mining in the El Maule region the
closest mines are 60 Km away (Pencahue and Romeral) so local factors related to
shoreline urbanization for the summer homes and an increase in tourist activity
could also be a major factor
6 Conclusions
The N isotope signal in the watershed depends on the rates of exchange
between vegetation and soil cover (Fig 11) Plants preferentially uptake 14N and the
underlying soils become enriched in 15N especially when the terrestrial ecosystem
is N-limited andor significant N loss occurs (ie denitrification andor ammonia
volatilization) LUCC in the Lago Vichuqueacuten watershed have heavily modified the
links between terrestrial and aquatic ecosystems with agriculture practices
102
contributing more N to the lake than tree plantations or native forests In situ lake
processes can also fractionate N isotopes An increase of N-fixing species results
in OM depleted in 15N which results in POM with lower δsup1⁵N values during these
periods During winter phytoplankton is typically enriched in 15N due to the
decreased abundance of N-fixing species and increased N input from the
watershed This in turn generates the highest δsup1⁵NPOM values in Lago Vichuqueacuten
Periods of elevated productivity tend to deplete the dissolved inorganic N of 14N
resulting in even higher δ15N values
Our study illustrates the usefulness of δsup1⁵N as a tool for reconstructing the past
influence of LUCC on N availability in lake ecosystems To constrain the relative
roles of the diverse forcing mechanisms that can alter N cycling in mediterranean
ecosystems all main components of the N cycle should be monitored seasonally
(or monthly) including the measurements of δ15N values in land samples
(vegetation-soil) as well as POM
103
Figure 11 Summary of human and environmental factors controlling the δ15N
values of lake sediments Particulate organic matter(POM) δ15N values in
mediterranean lakes are driven by N input from the watershed that in turn depend
on land use and cover changes (ie forest plantation agriculture) andor seasonal
changes in N sources andor lake ecosystem processes (ie bioproductivity redox
condition denitrification) Arrows indicate the direction of N fluxes (inputoutput from
the N cycle) N cycle processes that deplete lake sediments of 15N are shown in
blue whereas those that enrich sediments in 15N are shown in red
104
Supplementary material
Figure S1 Pearson correlate coefficient between geochemical variables in core
VIC13-2B Positive and large correlations are in blue whereas negative and small
correlations are in red (p valuelt0001)
Figure S2 Principal Component Analysis of geochemical elements from core
VIC13-2B
105
Table S1 Lago Vichuqueacuten radiocarbon samples
RADIOCARBON
LAB CODE
SAMPLE
CODE
DEPTH
(m)
MATERIAL
DATED
14C AGE ERROR
D-AMS 029287
VIC13-2B-
1 043 Bulk 1520 24
D-AMS 029285
VIC13-2B-
2 085 Bulk 1700 22
D-AMS 029286
VIC13-2B-
2 124 Bulk 1100 29
Poz-63883 Chill-2D-1 191 Bulk 945 30
D-AMS 001133
VIC11-2A-
2 201 Bulk 1150 44
Poz-63884
Chill-2D-
1U 299 Bulk 1935 30
Poz-64089
VIC13-2D-
2U 463 Bulk 1845 30
Poz-64090
VIC13-2A-
3U 469 Bulk 1830 35
D-AMS 010068
VIC13-2D-
4U 667 Bulk 2831 25
Poz-63886
VIC13-2D-
4U 719 Bulk 3375 35
106
D-AMS 010069
VIC13-2D-
5U 775 Bulk 3143 27
Poz-64088
VIC13-2D-
5U 807 Bulk 3835 35
D-AMS-010066
VIC13-2D-
7U 1075 Bulk 6174 31
Poz-63885
VIC13-2D-
7U 1197 Bulk 6440 40
Poz-5782 VIC13-15 DIC 180 25
Table S2 Loadings of the trace chemical elements used in the PCA
Elementos PC1 PC2 PC3 PC4
Zr 0922 0025 -0108 -0007
Zn 0913 -0124 -0212 0001
Rb 0898 -0057 -0228 0016
K 0843 0459 0108 0113
Ti 0827 0497 0060 -0029
Al 0806 0467 0080 0107
Si 0803 0474 0133 0136
Y 0784 -0293 -0174 0262
V 0766 0455 0090 -0057
Br 0422 -0716 -0045 0226
Ca 0316 -0429 0577 0489
Sr 0164 -0420 0342 -0182
Cl 0151 -0781 -0397 0162
107
Mn -0121 -0091 0859 0095
S -0174 -0179 -0051 0714
Pb -0349 0414 -0282 0500
Fe -0700 0584 -0023 0280
Co -0704 0564 -0107 0250
Table S3 Stable Isotope values from vegetation of Lago Vichuqueacutenacutes watershed
Taxa Classification δsup1⁵N δsup1sup3C CN
molar
Poaceae Meadow 1216 -2589 3602
Juncacea Meadow 1404 -2450 3855
Cyperaceae Meadow 1031 -2596 1711
Taraxacum
officinale Meadow 836 -2400 2035
Poaceae Meadow 660 -2779 1583
Poaceae Meadow 453 -2813 1401
Poaceae Meadow 966 -2908 4010
Juncus Meadow 1247 -2418 3892
Poaceae Meadow 747 -3177 6992
Poaceae Meadow 942 -2764 3147
Poaceae Meadow 1479 -2634 2895
Poaceae Meadow 1113 -2776 1795
Poaceae Meadow 2215 -2737 7971
Poaceae Meadow 1121 -2944 2934
Poaceae Meadow 638 -3206 1529
108
Macrophytes Macrophytes 886 -3044 2286
Macrophytes Macrophytes 1056 -2720 2673
Macrophytes Macrophytes 769 -3297 1249
Macrophytes Macrophytes 967 -2763 1442
Macrophytes Macrophytes 959 -2670 2105
Macrophytes Macrophytes 334 -2728 1038
Acacia dealbata
Introduced
species 656 -2696 1296
Acacia dealbata
Introduced
species 487 -2941 1782
Acacia dealbata
Introduced
species 220 -2611 3888
Luma apiculata Native species 433 -2542 4135
Luma apiculata Native species 171 -2664 7634
Luma apiculata Native species -001 -2736 6283
Luma apiculata Native species 029 -2764 6425
Azara sp Native species 159 -2868 8408
Azara sp Native species 101 -2606 2885
Baccharis concava Native species 104 -2699 5779
Baccharis concava Native species 265 -2488 4325
Baccharis concava Native species 287 -2562 7802
Baccharis concava Native species 427 -2781 5204
Baccharis linearis Native species 190 -2610 4414
Baccharis linearis Native species 023 -2825 5647
109
Peumus boldus Native species 042 -2969 6327
Peumus boldus Native species 205 -2746 4110
Peumus boldus Native species 183 -2743 6293
Chusquea quila Native species 482 -2801 4275
Poaceae meadow 217 -2629 7214
Lobelia sp Native species 224 -2645 3963
Lobelia sp Native species -091 -2565 4538
Aristotelia chilensis Native species -035 -2785 5247
Aristotelia chilensis Native species -305 -2889 2305
Aristotelia chilensis Native species 093 -2836 5457
Chusquea quila Native species 173 -2754 3534
Chusquea quila Native species 045 -2950 6739
Quillaja saponaria Native species 223 -2838 9385
Scirpus meadow 018 -2820 7115
Sophora sp Native species -184 -2481 2094
Sophora sp Native species -181 -2717 1721
Pinus radiata
Introduced
trees 1581 -2602 3679
Pinus radiata
Introduced
trees 1431 -2784 4852
Pinus radiata
Introduced
trees -091 -2708 9760
Pinus radiata
Introduced
trees 153 -2568 3470
110
Salix sp
Introduced
trees 632 -2878 1921
LITERATURE CITED
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A historical
framework for land cover change in southwestern South America in the past 15000
years Land use policy 27 148ndash160
httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the next
carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014
httpsdoiorg101002eft2235
Blaauw M Christeny JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474
httpsdoiorg10121411-BA618
Bowman DMJS Cook GD 2002 Can stable carbon isotopes (δ13C) in soil
carbon be used to describe the dynamics of Eucalyptus savanna-rainforest
boundaries in the Australian monsoon tropics Austral Ecol
httpsdoiorg101046j1442-9993200201158x
Brahney J Ballantyne AP Turner BL Spaulding SA Otu M Neff JC 2014
Separating the influences of diagenesis productivity and anthropogenic nitrogen
deposition on sedimentary δ15N variations Org Geochem 75 140ndash150
httpsdoiorg101016jorggeochem201407003
111
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409
httpsdoiorg102134jeq20090005
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego R
Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and
environmental change from a high Andean lake Laguna del Maule central Chile
(36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the
Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from
tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Craine JM Elmore AJ Aidar MPM Dawson TE Hobbie EA Kahmen A
Mack MC Kendra K Michelsen A Nardoto GB Pardo LH Pentildeuelas J
Reich B Schuur EAG Stock WD Templer PH Virginia RA Welker JM
Wright IJ 2009 Global patterns of foliar nitrogen isotopes and their relationships
with cliamte mycorrhizal fungi foliar nutrient concentrations and nitrogen
availability New Phytol 183 980ndash992 httpsdoiorg101111j1469-
8137200902917x
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty
Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-
010-9453-1
Diebel M Vander Zanden MJ 2012 Nitrogen stable isotopes in streams stable
isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19
1127ndash1134 httpsdoiorg10189008-03271
112
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in freshwater
wetlands record long-term changes in watershed nitrogen source and land use SO
- Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash
2916
Fariacuteas L Castro-Gonzales M Cornejo M Charpentier J Faundez J
Boontanon N Yoshida N 2009 Denitrification and nitrous oxide cycling within the
upper oxycline of the oxygen minimum zone off the eastern tropical South Pacific
Limnol Oceanogr 54 132ndash144
Farquhar GD Ehleringer JR Hubick KT 1989 Carbon Isotope Discrimination
and Photosynthesis Annu Rev Plant Physiol Plant Mol Biol
httpsdoiorg101146annurevpp40060189002443
Farquhar GD OrsquoLeary MH Berry JA 1982 On the relationship between
carbon isotope discrimination and the intercellular carbon dioxide concentration in
leaves Aust J Plant Physiol httpsdoiorg101071PP9820121
Fogel ML Cifuentes LA 1993 Isotope fractionation during primary production
Org Geochem httpsdoiorg101007978-1-4615-2890-6_3
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A
Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-
resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS)
implications for past sea level and environmental variability J Quat Sci 32 830ndash
844 httpsdoiorg101002jqs2936
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924
httpsdoiorg104319lo20095430917
113
Gerhart LM McLauchlan KK 2014 Reconstructing terrestrial nutrient cycling
using stable nitrogen isotopes in wood Biogeochemistry 120 1ndash21
httpsdoiorg101007s10533-014-9988-8
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and oxygen
isotope fractionation during dissimilatory nitrate reduction by denitrifying bacteria 53
2533ndash2545 httpsdoiorg10230740058342
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater
eutrophic lake Limnol Oceanogr 51 2837ndash2848
httpsdoiorg104319lo20065162837
Harris D Horwaacuteth WR van Kessel C 2001 Acid fumigation of soils to remove
carbonates prior to total organic carbon or CARBON-13 isotopic analysis Soil Sci
Soc Am J 65 1853 httpsdoiorg102136sssaj20011853
Houmlgberg P 1997 Tansley review no 95 natural abundance in soil-plant systems
New Phytol httpsdoiorg101046j1469-8137199700808x
Kylander ME Ampel L Wohlfarth B Veres D 2011 High-resolution X-ray
fluorescence core scanning analysis of Les Echets (France) sedimentary sequence
New insights from chemical proxies J Quat Sci 26 109ndash117
httpsdoiorg101002jqs1438
Lara A Solari ME Prieto MDR Pentildea MP 2012 Reconstruccioacuten de la
cobertura de la vegetacioacuten y uso del suelo hacia 1550 y sus cambios a 2007 en la
ecorregioacuten de los bosques valdivianos lluviosos de Chile (35o - 43o 30acute S) Bosque
(Valdivia) 33 03ndash04 httpsdoiorg104067s0717-92002012000100002
Lehmann M Bernasconi S Barbieri A McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during
114
simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66
3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007
Liu G Li M An L 2007 The Environmental Significance Of Stable Carbon
Isotope Composition Of Modern Plant Leaves In The Northern Tibetan Plateau
China Arctic Antarct Alp Res httpsdoiorg1016571523-0430(07-505)[liu-
g]20co2
Marzecovaacute A Mikomaumlgi A Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54
httpsdoiorg103176eco2011105
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash
1643 httpsdoiorg1011770959683613496289
Meyers PA 2003 Application of organic geochemistry to paleolimnological
reconstruction a summary of examples from the Laurention Great Lakes Org
Geochem 34 261ndash289 httpsdoiorg101016S0146-6380(02)00168-7
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland
Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Odone C 1998 El Pueblo de Indios de Vichuqueacuten siglos XVI y XVII Rev Hist
Indiacutegena 3 19ndash67
Pausch J Kuzyakov Y 2018 Carbon input by roots into the soil Quantification of
rhizodeposition from root to ecosystem scale Glob Chang Biol
httpsdoiorg101111gcb13850
115
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98
httpsdoiorg1011772053019614564785
Sun W Zhang E Jones RT Liu E Shen J 2016 Biogeochemical processes
and response to climate change recorded in the isotopes of lacustrine organic
matter southeastern Qinghai-Tibetan Plateau China Palaeogeogr Palaeoclimatol
Palaeoecol httpsdoiorg101016jpalaeo201604013
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of
different trophic status J Paleolimnol 47 693ndash706
httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl
httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M 2009 High-resolution quantitative climate
reconstruction over the past 1000 years and pollution history derived from lake
sediments in Central Chile Philos Fak PhD 246
Woodward C Chang J Zawadzki A Shulmeister J Haworth R Collecutt S
Jacobsen G 2011 Evidence against early nineteenth century major European
induced environmental impacts by illegal settlers in the New England Tablelands
south eastern Australia Quat Sci Rev 30 3743ndash3747
httpsdoiorg101016jquascirev201110014
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K Yeager
KM 2016 Different responses of sedimentary δ15N to climatic changes and
116
anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau
J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
117
DISCUSION GENERAL
El Nitroacutegeno (N) es un elemento clave que controla la dinaacutemica y
funcionamiento de ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al
1997) Pese a que el N (N2) es muy abundante en la atmoacutesfera (78) es escaso
en los ecosistemas pues la mayoriacutea de la biota no puede acceder a su forma
molecular (Galloway et al 1995 Zaehle et al 2013) En este sentido la entrada
natural de N en los ecosistemas es viacutea fijacioacuten bioloacutegica del N2 por bacterias que lo
convierten en compuestos bioloacutegicamente disponibles (Battye et al 2017) Debido
a que la fijacioacuten es un proceso energeacuteticamente costoso los ecosistemas
comuacutenmente estaacuten limitados por N (Vitousek and Howarth 1991) Los LUCC
contribuyen al incremento del N disponible y son una de las principales causas de
eutroficacioacuten y contaminacioacuten de los cuerpos de agua (Woodward et al 2012)
En Chile central los LUCC principalmente relacionados con las actividades
agriacutecolas y forestales han causado grandes cambios en el ciclo del N debido al
118
reemplazo de la cobertura natural por un monocultivo y al uso de fertilizantes que
modifican los aportes de MO y N a los cuerpos de agua El programa de
estabilizacioacuten de laderas impulsada por el gobierno (Albert 1900) y la ley forestal
de 1974 han fomentado la forestacioacuten con plantaciones de Pinus radiata y
Eucalyptus globulus y en el caso de Lago Vichuqueacuten estas actualmente cubren maacutes
del 60 de su cuenca (Cap2 Fig 5) Ademaacutes en Laguna Matanzas la
sobreexplotacioacuten del agua en conjunto con el cambio climaacutetico actual el que ha
conducido a una de las megasequiacuteas maacutes severas en las uacuteltimas deacutecadas
(Garreaud et al 2017) generoacute una combinacioacuten letal que secoacute el lago en tan solo
10 antildeos (Cap1 Fig1) Las evidencias obtenidas a partir de estos dos lagos
permiten identificar las huellas del Antropoceno en Chile central basadas en el
registro sedimentario lacustre
La transferencia de N desde los ecosistemas terrestres a acuaacuteticos es un
proceso natural con controles bioacuteticos climaacuteticos y geoloacutegicos El enlace
hidroloacutegico entre los ecosistemas terrestres y acuaacuteticos hace que el estado troacutefico
de los lagos esteacute vinculado al de su cuenca (McLauchlan et al 2013) En Chile
central se sabe poco del impacto de los LUCC sobre el ciclo de nutrientes en los
ecosistemas lacustres aunque al menos desde la colonizacioacuten espantildeola se tienen
registros de influencia humana en las cuencas Durante la colonia espantildeola
Laguna Matanzas fue una hacienda ganadera que exportaba productos caacuternicos al
Peruacute (Contreras-Lopez et al 2014) A su vez en el Lago Vichuqueacuten se realizaban
extensas actividades agriacutecolas hasta comienzos del s XX como por ejemplo
cultivos de vid y trigo ademaacutes de extraccioacuten de madera y mineriacutea de oro (Odone
1997) En las uacuteltimas deacutecadas los LUCC han estado vinculados principalmente con
el desarrollo forestal al compaacutes de una estrategia gubernamental de fomentar con
119
incentivos financieros (Ley de Bosques DFL nordm265 de 1931 y DFL 701 de 1974)
esta actividad El incremento de la superficie forestal es especialmente fuerte en
ambos lagos a partir de la deacutecada de 1980 y hasta 2016 (Laguna de Matanzas 0-
17 Lago Vichuqueacuten 1-66) Este aumento habriacutea sido en detrimento del bosque
nativo y los pastizales (Cap 1 Fig 5 y Cap 2 Fig 8) El incremento de la superficie
forestal generariacutea una disminucioacuten de los aportes de sedimentos y nutrientes al lago
y en este sentido un cambio de estado en los flujos de N (e g tipping points) que
a su vez ha quedado registrado en la sentildeal isotoacutepica (δ15N) y la acumulacioacuten de
MO en los sedimentos lacustres
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central
Los sedimentos lacustres son verdaderos ldquosumiderosrdquo que pueden llegar a
registrar lo que ocurre en cuanto a la historia ambiental de su cuenca En esta tesis
se utilizan los isoacutetopos estables de N (15N y 14N) en sedimentos lacustres para
reconstruir los cambios en la disponibilidad de N en el tiempo y establecer la
magnitud de impacto generado por actividades humanas El fraccionamiento
cineacutetico en los lagos es mediado por procesos bioloacutegicos que mediante la
asimilacioacuten de N genera un producto (eg fitoplancton) maacutes ldquoligerordquo (valores maacutes
bajos de 15N) que su remanente (eg DIN) (Fogel y Cifuentes 1993) Sin embargo
en periacuteodos en que los lagos estaacuten limitados por N el fitoplancton discrimina menos
y la MO resultante queda enriquecida en 15N (Fig 1 a y b) Ademaacutes la
desnitrificacioacuten acentuada en lagos de fondo anoacutexicos (eg Laguna Matanzas
entre 1800 y 1940 Cap1 Fig 6) conduce a un enriquecimiento en 15N de los
sedimentos y de la columna de agua Esta informacioacuten queda reflejada en la MO
120
de los sedimentos lacustres lo que permite conocer la disponibilidad del N en el
tiempo a partir de las variaciones de 15N
En esta tesis analizamos la MO de los sedimentos lacustres para reconstruir
la influencia de los LUCC en los aportes de sedimentos y N al lago En teacuterminos de
asimilacioacuten de N se puede distinguir entre dos grupos principales de productores
primarios que componen el POM (Fig1)
1 Los que asimilan el DIN compuesto por amonio nitratos y nitritos Aquiacute el
δ15N resultante fluctuaraacute en torno a valores maacutes positivos en la medida que
la productividad se incrementa o no hay reposicioacuten del DIN (Fig 1)
2 Los fijadores de N Dado que es un proceso energeacuteticamente costoso en
ambientes que no estaacuten limitados por N muchas veces son excluiacutedas
competitivamente por el resto del fitoplancton Si el DIN queda agotado por
el incremento de la productividad (o bien si el ambiente estaacute limitado ya sea
por N o por otros nutrientes) los fijadores de N pueden florecer y la MO que
se genera a partir de ello tendraacute un δ15N con valores en torno a 0permil
De este modo la MO en los sedimentos lacustres dependeraacute de la
composicioacuten de productores primarios (ie fijadores vs asimiladores del DIN)
ademaacutes de los aportes desde la cuenca tanto de MO como del N de la cuenca que
pasa a formar parte del DIN (eg fertilizantes estieacutercol) (Fig 1)
La MO de los lagos estudiados en esta tesis ha sido analizada a partir de
variables geoquiacutemicas isotoacutepicas frotis y estaacute compuesta principalmente por
diatomeas y MO amorfa A partir de los datos obtenidos en esta tesis los valores
de δ15N aumentan cuando hay una mayor cobertura agriacutecola o pastizales A su vez
tiende a disminuir cuando aumenta la cobertura arboacuterea independiente si esta es
por plantaciones forestales o por bosque nativo
121
Por otra parte la dinaacutemica del lago tambieacuten imprime caracteriacutesticas
especiales en el POM observaacutendose variaciones estacionales en los valores
δ15NPOM Durante la estacioacuten caacutelida y seca se observan valores maacutes bajos que
durante la estacioacuten friacutea y huacutemeda posiblemente relacionado con el incremento de
la contribucioacuten de especies fijadoras de N (Lago Vichuqueacuten Fig 9 Cap 2) durante
el verano En la estacioacuten friacutea y huacutemeda el DIN seriacutea la principal fuente de N Las
mayores entradas de MO y N terrestre debidos a un incremento del lavado de la
cuenca como consecuencia de las precipitaciones y la mineralizacioacuten de la MO
podriacutean estimular la produccioacuten de MO lacustre por crecimiento del fitoplancton
Como consecuencia se observan tendencias decrecientes de los valores de
δ15N en la MO sedimentaria cuando la productividad en el lago estaacute relacionada
con especies fijadoras de N mientras que el δ15N muestra aumentos cuando la
productividad del lago estaacute asociada principalmente al consumo del DIN pero
tambieacuten cuando predominan procesos de perdida de N (eg desnitrificacioacuten Fig
1)
Hemos identificado tres fases en la dinaacutemica del δ15N para ambos lagos
Entre 1800 y 1940 CE la cuenca de laguna de Matanzas estaba dominada por
actividades agriacutecolas y ganaderas (δ15Nsuelo gt 4permil Cap 1 Fig 4 y 6) Las entradas
de sedimentos y MO desde la cuenca son altas (δ13C lt -209permil ZrTi ~029 AlTi
~013) y coincide con un clima maacutes huacutemedo y friacuteo (Von Gunten et al 2009
Christiansen et al 2011) El lago teniacutea un nivel de agua maacutes profundo dominado
por condiciones anoacutexicas (MnFe ~0013) que contribuiriacutean a mayores valores de
δ15N en la columna de agua y por ende de los sedimentos acumulados en el fondo
debido a la peacuterdida diferencial de 14N viacutea desnitrificacioacuten (Fig 7 Cap1) La baja
122
produccioacuten de MO lacustre (BrTi ~002 TOC~17) coincide con un bajo contenido
de NT (~478 μg) y valores maacutes positivos de δ15N (gt 4permil)
La actividad agriacutecola tambieacuten dominaba en la cuenca del Lago Vichuqueacuten
durante este periacuteodo (δ15Nsuelo gt4permil Cap2 Fig 6) El aporte sedimentario desde la
cuenca es alto (AlTi ~029 ZrTi ~032) y fue favorecido por mayores
precipitaciones a las actuales (Christiansen et al 2011) El Lago Vichuqueacuten es un
lago estratificado anoacutexico (MnFe ~002) y profundo (~30 m) por lo que la
desnitrificacioacuten juega un rol importante en el enriquecimiento de la MO
sedimentaria La baja produccioacuten de MO (BrTi ~007 TOC ~12) de origen
lacustre (δ13C ~ -268permil) coincide con un bajo contenido de NT (~309μg +6) y
valores maacutes positivos de δ15N (56permil +03)
Durante esta fase en ambos lagos los aportes de N de la cuenca parecen
ser los principales responsables en las fluctuaciones del δ15N Esta sentildeal podriacutea
estar amplificada por bajas temperaturas (que afectan la productividad lacustre) y
altas precipitaciones que favorecen el lavado de la cuenca y aumentan el aporte de
sedimentos y MO desde la cuenca predominantemente agriacutecola
Entre 1940 y 1980 CE en Laguna Matanzas se observa un incremento en
la acumulacioacuten de MO algal (TOC ~2 +1 δ13C ~-230+09permil) El ambiente
deposicional es ligeramente menos turbulento que el anterior (AlTi ~011+001
ZrTi ~03+001) y el lago estaacute en una fase de transicioacuten hacia condiciones maacutes
oacutexicas (MnFe ~002+0004) y maacutes productivas (BrTi ~004+0007) A su vez δ15N
tiende hacia valores maacutes negativos (δ15N ~30permil+03) mientras que el NT aumenta
oscilando en antifase con el δ15N
En Lago Vichuqueacuten en cambio se observa un ligero incremento en la
acumulacioacuten de MO (TOC ~13 +02) de origen terrestre (δ13C ~-27permil +04) La
123
productividad es similar al periacuteodo anterior (BrTi ~007+003) y el ambiente
deposicional es menos turbulento (AlTi ~02 +009 Zrti ~033 +007) El δ15N y el
NT oscilan en torno a valores ligeramente maacutes bajos (δ15N ~51permil +04 NT ~292μg
+6) Este periacuteodo se caracteriza por ser maacutes caacutelido que el anterior lo que
posibilitariacutea un incremento en la productividad lacustre en Laguna Matanzas pero
que no es observada en el Lago Vichuqueacuten
Entre 1980 y la actualidad en Laguna Matanzas habriacutea un incremento de la
acumulacioacuten de MO (TOC ~59 + 3) asociada a un incremento en la productividad
del lago (BrTi ~009+005) y a los aportes de MO terrestres (δ13C ~-24 + 2permil) El
lago se vuelve maacutes oacutexico (Mn ~0003 +0006) y las entradas de sedimento
disminuyen (AlTi~ 009 +002) El δ15N tiende a valores maacutes negativos (δ15N ~13permil
+1) y el NT aumenta de 20 a 55 μg (NT ~346 + 9 μg) tomando valores sin
precedentes en todo el registro En los sedimentos lacustres del Lago Vichuqueacuten
tambieacuten se observa un incremento de la acumulacioacuten de MO (TOC ~17 + 05)
asociada a un incremento en la productividad del lago (BrTi ~01 + 003) y las
entradas de sedimento desde la cuenca disminuyen (AlTi ~005 + 005) El δ15N
(~43 + 05permil) tiende a valores maacutes negativos y el NT incrementa de 20 a 55 μg (NT
~346 + 9 μg) oscilando en antifase
Durante esta fase en ambos lagos se observa un aumento en la
acumulacioacuten de MO sedimentaria un incremento en el NT y valores maacutes negativos
de δ15N que coincide con el incremento de la superficie forestal de las cuencas
(Laguna Matanzas gt17 Lago Vichuqueacuten gt60)
124
Figura 2 Comparacioacuten de los cambios del ciclo del N entre Laguna Matanzas y
Lago Vichuqueacuten con los procesos biogeoquiacutemicos contrastantes en rojo En L
Matanzas las condiciones oacutexicas podriacutean estar favoreciendo la nitrificacioacuten del
amonio En Vichuqueacuten las condiciones anoacutexicas favorecen un aumento en δ15N de
la MO en los sedimentos por desnitrificacioacuten y mineralizacioacuten
Los ambientes mediterraacuteneos en el que los lagos del presente estudio se
encuentran insertos estaacuten caracterizados por una estacioacuten seca prolongada y las
precipitaciones ocurren en eventos puntuales alcanzando altos montos
pluviomeacutetricos en poco tiempo Las lluvias favorecen un lavado de la cuenca la
perdida de N y otros nutrientes Por lo que es esperable que los sedimentos del
lago reflejen mejor los valores isotoacutepicos de los suelos de la cuenca durante los
periacuteodos con altos montos pluviomeacutetricos En el Lago Vichuqueacuten por ejemplo el
125
POM superficial (2 y 5 m de profundidad) despliega valores isotoacutepicos maacutes
positivos en invierno presumiblemente como resultado de mayores aportes de MO
y N de su cuenca (Fig 1b) En ambos lagos los valores maacutes positivos en los
sedimentos lacustres ocurren durante periacuteodos con mayores montos pluviomeacutetricos
(Cap1 Fig 6 y Cap 2 Fig 12)
Ademaacutes del efecto de la precipitacioacuten en el aporte de nutrientes al lago en
esta tesis hemos encontrado que los mayores aportes de N desde la cuenca se dan
cuando la cobertura vegetal del suelo es menor En Laguna Matanzas el desarrollo
de la actividad ganadera desde la llegada de los espantildeoles a la cuenca habriacutea
incrementado los aportes de N al lago Los valores de δ15N en los sedimentos
lacustres depositados durante este periacuteodo son los maacutes positivos de todo el registro
(~4permil) A su vez en Lago Vichuqueacuten los valores isotoacutepicos maacutes positivos se
registran entre 1300 y 1900 CE que coinciden con el mayor desarrollo de
actividades agriacutecola y de extraccioacuten de maderas de la cuenca (Fig 10 Cap 2)
Los recientes LUCC (a partir de 1975) en las cuencas de Laguna Matanzas
y Lago Vichuqueacuten estaacuten caracterizados por un incremento de la superficie forestal
y agriacutecola en reemplazo de la cobertura de pastizal (Laguna Matanzas) y bosque
nativo (Lago Vichuqueacuten) Los valores de δ15N en los testigos sedimentarios fueron
maacutes positivos cuanto mayor la superficie agriacutecola de la cuenca y maacutes negativos
cuanto mayor la superficie forestal Aunque por los alcances de esta tesis no
podemos ser concluyentes respecto del origen del NT (aloacutectonoautoacutectono) parece
ser que esta maacutes ligado a los aportes de la cuenca pese a la disminucioacuten del aporte
sedimentario observado en ambos lagos Las plantaciones forestales a diferencia
del bosque nativo son fertilizadas con una mezcla de N P y K (Toral et al 2005)
Por lo que pese a la disminucioacuten del aporte sedimentario la concentracioacuten de
126
nutrientes en el aporte al lago puede verse incrementada por el tipo de manejo
forestal con respecto al bosque nativo
Los resultados del primer capiacutetulo demuestran que 1) las plantaciones
forestales yo bosque nativo retiene maacutes sedimentos y MO mientras que el uso de
suelo agriacutecola y los pastizales destinados al desarrollo de la actividad ganadera lo
libera 2) Al parecer la desnitrificacioacuten es el proceso natural maacutes importante de
perdida de N en lagos costeros y favorece el enriquecimiento de 15N tanto en la
columna de agua (ie 15N-DIN) como en los sedimentos una vez enterrados y 3) la
desnitrificacioacuten depende de los cambios en las condiciones REDOX en el lago La
oxigenacioacuten en lagos poco profundos -como Laguna Matanzas- es altamente
fluctuante a escalas decadal-centenal depende de la profundidad de la laacutemina de
agua y de la variabilidad de la precipitacioacuten En este sentido en Laguna Matanzas
habriacutea condiciones anoacutexicas en ambientes cuando el lago alcanza niveles maacutes
altos Estas condiciones se observan entre 1820 y 1940 CE y son coetaacuteneas con
episodios maacutes huacutemedos y friacuteos (von Gunten et al 2009 Christie et al 2009) pero
tambieacuten con una fuerte actividad ganadera en la cuenca
Las fuentes variables de N y su fluctuacioacuten en el tiempo hacen necesario
contar con un anaacutelogo moderno que permite usar al δ15N de los sedimentos
lacustres como un indicador indirecto de los cambios en la disponibilidad de N en
el tiempo Por ello en el segundo capiacutetulo integramos muestras de suelo-
vegetacioacuten y un monitoreo de POM de la columna de agua en verano e invierno La
composicioacuten isotoacutepica de POM estariacutea reflejando cambios de la fuente de N (fijacioacuten
vs consumo del DIN) que variacutea durante el antildeo hidroloacutegico Durante el verano la
mayor temperatura y horas-luz favoreceriacutean las entradas de N al lago viacutea fijacioacuten
bioloacutegica de N2 atmosfeacuterico resultando en un incremento de la MO con un valor
127
isotoacutepico bajo (POM verano~5permil Fig 1b) El N2 (δ15N = 0permil) es fijado praacutecticamente
sin fraccionamiento isotoacutepico generando un δ15N de la OM cercano a 0permil (Leng et
al 2006) En invierno las precipitaciones pueden estar favoreciendo un incremento
en la entrada de nutrientes al lago El DIN de la columna de agua llega a contener
valores de δ15N maacutes altos reflejando estos mayores aportes de la cuenca y el POM
del Lago Vichuqueacuten queda enriquecido en 15N (δ15N ~11permil Cap 2 Fig 9) Estas
variaciones estacionales pueden estar evidenciando un cambio en la composicioacuten
de especies de POM desde especies fijadoras a especies que consumen el N de la
columna de agua Sin embargo para corroborar esta informacioacuten seriacutea deseable
contar con el anaacutelisis de la composicioacuten de especies de las muestras de agua
extraidas
Entre los resultados maacutes importantes estaacuten los valores isotoacutepicos del suelo
y la biomasa representativa de la cuenca que incluye un listado de las especies
nativas de la zona que hasta ahora no se contaba (Cap 2 Tabla en material
suplementario) Las especies colectadas despliegan valores de δ15Nfoliar maacutes
positivos (plantaciones forestales y pastizal ~89permil) que el suelo (35permil) salvo por
las especies nativas las que oscilan en torno a valores cercanos al atmosfeacuterico
(~14permil) La razoacuten isotoacutepica 15N14N refleja la dinaacutemica de intercambio de N entre la
vegetacioacuten y el suelo (Koba et al 2002) La discriminacioacuten es positiva en la mayoriacutea
de los sistemas bioloacutegicos por lo que las plantas tienen valores de δ15N maacutes bajos
que el suelo (Evans et al 2001) como sucede con las especies nativas en Lago
Vichuqueacuten En consecuencia las plantas quedan empobrecidas en 15N y conducen
a un enriquecimiento en 15N del suelo sobretodo si no hay reposicioacuten de N por otras
viacuteas (eg fijacioacuten de N) Por lo tanto los valores maacutes negativos de δsup1⁵Nfoliar de las
especies nativas pueden estar relacionados con el consumo preferencial de 14N del
128
suelo donde la discriminacioacuten isotoacutepica de la vegetacioacuten contra el 15N conduce a
valores del δsup1⁵Nsuelo maacutes altos (Houmlgberg 1997) Por el contrario los valores maacutes
positivos de δsup1⁵Nfoliar de las plantaciones forestales y pastos con respecto al suelo
puede deberse por una parte que el suelo no cuenta con mecanismos naturales de
reposicioacuten de N salvo por la descomposicioacuten de la hojarasca que puede ser maacutes
lenta en Pinus radiata que en bosque nativo (Lusk et al 2001) y por otra por el alto
impacto de los aportes de N (y otros nutrientes) derivado de las actividades
humanas (eg uso de fertilizantes) en el suelo
El alcance maacutes significativo de esta tesis se relaciona con un cambio en la
tendencia maacutes negativa de δsup1⁵N y el incremento del NT a partir de 1940 y a partir
de 1970 CE en ambos lagos La causa maacutes probable de esta tendencia es el
reemplazo de la cobertura natural o preexistente a 1940 CE por plantaciones
forestales
En la figura 2 se observa una siacutentesis de los principales procesos que
afectan la sentildeal isotoacutepica de la MO en los sedimentos lacustres de L Matanzas y
L Vichuqueacuten La entrada de N y MO desde la cuenca depende del tipo de cobertura
Las plantaciones forestales y el bosque nativo retienen maacutes nutrientes y sedimentos
en la cuenca mientras que la agricultura los favorece Sin embargo las praacutecticas
de manejo que incluyen la aplicacioacuten de N (y otros nutrientes) pueden aportar maacutes
nutrientes al lago que la cobertra de bosque nativo Cuando las actividades
forestales cubren una mayor superficie de la cuenca (1940 en adelante) δsup1⁵N oscila
en valores maacutes negativos y el NT tiende a incrementarse en los sedimentos
lacustres Cabe destacar que los valores de δsup1⁵N observados en los testigos
sedimentarios a fines del s XX no tienen precedente durante la conquista y colonia
espantildeola o durante el resto del periodo de la Repuacuteblica
129
Figura 2 Modelo de transferencia de N entre los ecosistemas terrestres y
acuaacuteticos El δ15N de los sedimentos lacustres depende de la interaccioacuten entre los
aportes de la cuenca y porcesos ldquoin-lakerdquo Flechas indican la direccioacuten del flujo de
N En azul las salidasentradas de 14N en rojo las salidasentradas de 15N δ15N de
la interaccioacuten plantas-suelo depende del tipo de cobertura de suelo
130
CONCLUSIONES GENERALES
La transferencia de N entre cuencas y lagos es un factor de control del ciclo
del N en los lagos y queda reflejado en la sentildeal isotoacutepica de los sedimentos
lacustres (Leng et al 2006 McLauchlan et al 2007) En Laguna Matanzas el
suelo de las especies nativas y las plantaciones forestales despliegan valores de
δ15N maacutes bajos (~1permil) que los de Lago Vichuqueacuten (~35permil) Coincidentemente los
sedimentos lacustres de Laguna Matanzas oscilan con valores de δ15N maacutes bajos
(-15 a +45permil Fig 1) que los de Lago Vichuqueacuten (+3 a +8permil)
Por otra parte el estudio de la MO de los sedimentos lacustres ha permitido
reconstruir los cambios en los aportes de MO y N desde la cuenca Aunque no es
posible afirmar queacute tipo de N estaacute entrando al lago (eg Nitroacutegeno orgaacutenico e
inorganico) si podemos decir que los cambios en las fluctuaciones de δ15N son
coincentes con el crecimiento de la actividad forestal (1980 - hasta 2016 L
Matanzas 0-17 y L Vichuqueacuten 1-66) El δ15N fluctuacutea hacia valores maacutes
negativos Ademaacutes se observa un incremento del NT en los sedimentos lacustres
cuanto mayor es la superficie forestal
Entre 1800-1940 las cuencas eran fundamentalmente agriacutecolas y
ganaderas (Cap1 Fig 7 y Cap 2 Fig 12) el δ15N de los sedimentos lacustres
oscila con valores maacutes positivos (L Matanzas +40 plusmn 05permil y L Vichuqueacuten 56 plusmn
033permil Fig 1) hay una baja bioproductividad en el lago (BrTi ~002 TOC~17)
lo que coincide con un bajo contenido de NT (~478 μg) Este es un periacuteodo de altas
precipitaciones (Christie et al 2009) que tienden a favorecer el lavado de la cuenca
131
y con ello el aporte de MO sedimentos y nutrientes desde la cuenca tiende a verse
favorecido Aunque las principales actividades humanas en estas cuencas son
diferentes (ganaderiacutea en Laguna Matanzas Contreras-Lopez et al 2014
agricultura en Lago Vichuqueacuten Odone 1997) en ambos casos hubo un reemplazo
de la cobertura natural de bosques nativo que favorecioacute mayores aportes de N y
sedimentos desde la cuenca en un efecto sumado con el aumento de las
precipitaciones
A partir de 1980 CE en ambos lagos se observa una disminucioacuten en los
valores de δ15N y un incremento del NT que no tiene precedentes en todo el registro
y pese a que ambos lagos son limnologicamente muy diferentes En Lago
Vichuqueacuten el δ15N tiende a oscilar en antifase con el NT (Fig 1) En el caso de
Laguna Matanzas se observa una tendencia hacia valores maacutes negativos y a partir
de 1990s a valores maacutes positivos mientras que el NT tiende a incrementarse de
manera sostenida Ello podriacutea estar relacionado con el crecimiento de la actividad
forestal habriacutea una mayor retencioacuten de N y sedimentos en la cuenca ligado al
incremento de la superficie forestal Ademaacutes en el caso de L Matanzas el
incremento de la actividad agriacutecola (sumado al de las plantaciones forestales)
podriacutea expicar el cambio en la tendencia en la deacutecada de los 1990s
En el contexto de Antropoceno esta tesis nos permite identificar un gran
impacto de las actividades humanas en Chile central a partir de la deacutecada de 1940
y una Gran Aceleracoacuten a partir de la deacutecada de 1970s En el registro sedimentario
de los lagos en estudio se observa una gran acumulacioacuten de MO el δ15N oscila
hacia valores maacutes negativos y un gran incremento del NT y el crecimiento de la
actividad forestal parece ser la principal responsable Ademaacutes la sobreexplotacioacuten
132
del agua junto al cambio climaacutetico global ha generado una combinacioacuten letal para
los lagos costeros de Chile central
Figura 1 Comparacioacuten de las fluctuaciones de δ15N durante los uacuteltimos 300
antildeos en Laguna Matanzas y Lago Vichuqueacuten
133
Referencias
Battye W Aneja VP Schlesinger WH 2017 Earthrsquos Future Is nitrogen the next carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia
UC de V 2014 Elementos de la historia natural del An Mus Hist Natulas Vaplaraiso 27 51ndash67
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Evans RD Evans RD 2001 Physiological mechanisms influencing
plant nitrogen isotope composition 6 121ndash126 Fogel ML Cifuentes LA 1993 Isotope fractionation during primary
production Org Geochem 73ndash98 httpsdoiorg101007978-1-4615-2890-6_3 Galloway JN Schlesinger WH Ii HL Schnoor L Tg N 1995
Nitrogen fixation Anthropogenic enhancement-environmental response reactive N Global Biogeochem Cycles 9 235ndash252
Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J
Christie D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Koba K Koyama LA Kohzu A Nadelhoffer KJ 2003 Natural 15 N
Abundance of Plants Coniferous Forest httpsdoiorg101007s10021-002-0132-6
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos
Transactions American Geophysical Union httpsdoiorg1010292007EO070007 McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE
2007 Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
134
Phytologist N Oct N 2006 Tansley Review No 95 15N Natural Abundance in Soil-Plant Systems Peter Hogberg 137 179ndash203
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Turner BL Kasperson RE Matson PA McCarthy JJ Corell RW
Christensen L Eckley N Kasperson JX Luers A Martello ML Polsky C Pulsipher A Schiller A 2003 A framework for vulnerability analysis in sustainability science Proc Natl Acad Sci U S A httpsdoiorg101073pnas1231335100
Vitousek PM Aber JD Howarth RW Likens GE Matson PA
Schindler DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
Vitousek PM Howarth RW 1991 Nitrogen limitation on land and in the
sea How can it occur Biogeochemistry httpsdoiorg101007BF00002772 von Gunten L Grosjean M Rein B Urrutia R Appleby P 2009 A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 Holocene 19 873ndash881 httpsdoiorg1011770959683609336573
Xu R Tian H Pan S Dangal SRS Chen J Chang J Lu Y Maria
Skiba U Tubiello FN Zhang B 2019 Increased nitrogen enrichment and shifted patterns in the worldrsquos grassland 1860-2016 Earth Syst Sci Data httpsdoiorg105194essd-11-175-2019
Zaehle S 2013 Terrestrial nitrogen-carbon cycle interactions at the global
scale Philos Trans R Soc B Biol Sci 368 httpsdoiorg101098rstb20130125
4
2 Study Site 77 3 Methods 79 4 Results 84
41 Geochemistry and PCA analysis 84 42 Sedimentary units 86 43 Recent seasonal changes of particulate organic matter on water column 88 44 Stable isotope values across the Lake Vichuqueacuten watershed 89 45 Land use and cover change from 1975 to 2014 90
5 Discussion 93 51 Seasonal variability of POM in the water column 93 52 Stable isotope signatures in the Lake Vichuqueacuten watershed 95 53 Recently land use and cover change and its influences on N inputs to the lake 97 54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years 98
6 Conclusions 101
LITERATURE CITED 110
DISCUSION GENERAL 117
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central 119
CONCLUSIONES GENERALES 130
Referencias 133
5
A mis padres Arturo y Malena
A mis hijos Xavi y Panchito
6
AGRADECIMIENTOS
Quiero agradecer a mi tutor y mentor Dr Claudio Latorre por brindarme su
apoyo sin el cual no habriacutea logrado concluir esta tesis de doctorado Claudio tu
apoyo constante incentivo y el fijarme metas que a veces me pareciacutean imposibles
de alcanzar no solo han dado forma a esta tesis sino tambieacuten me ha hecho maacutes
exigente como cientiacutefica Claudio destacas no solo por ser un gran cientiacutefico si no
tambieacuten por tu gran calidad humana eres un gran ejemplo
Quiero agradecer Dr Blas Valero-Garceacutes por nuestras numerosas
conversaciones viacutea Skype que incluiacutean vacaciones y fines de semana para discutir
los resultados de la tesis y que han dado forma a esta investigacioacuten principalmente
al primer capiacutetulo Ademaacutes por haberme acogido como un miembro maacutes en el
laboratorio de Paleoambientes Cuaternarios durante las estancias que he realizado
en el transcurso de estos antildeos Blas eres un ejemplo para miacute conjugas ciencia de
calidad calidez y dedicacioacuten por tus estudiantes
A quienes han financiado mi doctorado la Comisioacuten Nacional de
Investigacioacuten Cientiacutefica y Tecnoloacutegica (CONICYT) con sus becas de manutencioacuten
doctoral (2013) gastos operacionales pasantiacutea (2016) postnatal (2017) y termino
de tesis doctoralrdquo (2013) A FONDECYT a traveacutes del proyecto 1160744 de C
Santoro Al Departamento de InvestigacioacutenAl Instituto de Ecologiacutea y Biodiversidad
(IEB) a traveacutes de del PIA financiamiento basal 170008 la Pontificia Universidad
Catoacutelica de Chile por la beca incentivo para tesis interdisciplinaria para doctorandos
(2015)
Agradezco a mis compantildeeros del laboratorio de Paleoecologiacutea y
Paleoclimatologiacutea Karla Matias Dani Carolina Mauricio y Pancho que han hecho
grato mi tiempo en el laboratorio Agradecimientos especiales a Carolina Matiacuteas y
Leo por acompantildearme a terreno
7
Agradezco a los miembros del laboratorio de Paleoambientes Cuaternarios
(IPE) en especial a Raquel Elena Fernando Ana Penelope Graciela Miguel
Sevilla Mariacutea y Miguel Bartolomeacute
Agradezco a los miembros de la comisioacuten Dr Joseacute Miguel Farintildea Dr Juan
Armesto y Dr Ricardo De Pol-Holz por la gran ayuda durante el desarrollo de mi
doctorado en especial por las correcciones finales de la tesis
Al departamento de Ecologiacutea en especial a Rosario Galaz Edita Luengo
Daniela Mora y Valeria Cavallero por su apoyo
A mis compantildeeros de doctorado Beleacuten Gallardo Pancho Diacuteaz Bryan Bularz
Bian Latorre Renato Borras Juan Carlos Hernaacutendez y Paulina Troncoso con
quienes estudieacute aprendiacute y compartimos muy buenos momentos durante los
primeros antildeos del doctorado
A Pablo Sarricolea mi compantildeero de vida Gracias por tu constante e
incondicional apoyo Sin ti durante este proceso todo habriacutea sido cuesta arriba
A mis padres Arturo y Malena gracias por su carintildeo apoyo y compromiso
Papaacute aunque no esteacutes a mi lado siempre te recuerdo con mucho carintildeo A mi
madre que ha sido un constante apoyo sobre todo en el cuidado de mis hijos xavi
y panchito
A mis hermanos Rodrigo y David por estar presentes durante toda esta
etapa Siempre con carintildeo y hermandad
A Rodrigo David Matiacuteas Jany Paola y Julio por cuidar a mis hijos con carintildeo
siempre que estuve ausente por el doctorado
8
ABREVIATURAS
N Nitroacutegeno (Nitrogen)
DIN Nitroacutegeno Inorgaacutenico Disuelto (Dissolved Inorganic Nitrogen)
C Carbono (Carbon)
TOC Carbono Inorgaacutenico Total (Total Organic Carbon)
TIC Carbono Inorgaacutenico Total (Total inorganic Carbon)
TC Carbono Total (Total Carbon)
TS Azufre Total (Total Sulfur)
LUCC Cambios de UsoCobertura del Suelo (Land Use and Cover Change)
OM Materia Orgaacutenica (Organic Matter)
POM Particulate Organic Matter (materia orgaacutenica particulada)
CE Common Era
BCE Before Common Era
Cal BP Calibrado en antildeos radiocarbono antes de 1950
ie id est (esto es)
e g Exempli gratia (por ejemplo)
9
RESUMEN
El ldquoAntropocenordquo se caracteriza por pertubaciones de origen humano que
conllevan impactos globales Por ejemplo los cambios de usocobertura del suelo
(LUCC) son conocidos por perturbar el ciclo del N a partir de la Revolucioacuten Industrial
pero especialmente desde la Gran Aceleracioacuten (1950 CE) en adelante Sin
embargo existen incertezas asociadas a la magnitud del impacto y su efecto
acoplado con los cambios climaacuteticos actuales (ie megasequiacutea disminucioacuten de las
precipitaciones) y acoplamientos con otros ciclos biogeoquiacutemicos (eg ciclo del
Carbono) Este impacto ha cambiado la disponibilidad de N en los ecosistemas
terrestres y acuaacuteticos y sus enlaces Los sedimentos lacustres contienen
informacioacuten de las condiciones paleoambientales del lago y su cuenca en el
momento en que se depositaron y el anaacutelisis de los isoacutetopos estables de N (δ15N)
en los sedimentos permiten reconstruir los cambios en la disponibilidad de N a
traveacutes del tiempo En este trabajo usamos una aproximacioacuten multiproxy que incluye
10
anaacutelisis sedimentoloacutegicos geoquiacutemicos e isotoacutepicos (δ15N δ13C) de los sedimentos
lacustres columna de agua y suelo-vegetacioacuten de la cuenca y la reconstruccioacuten de
los LUCC entre 1975 y 2016 a partir de imaacutegenes satelitales El objetivo de esta
tesis fue evaluar el rol de LUCC como principal control del ciclo del N en el sistema
cuenca-lago durante las uacuteltimas centurias en Chile central Nuestros principales
resultados muestran que valores maacutes positivos de δ15N en los sedimentos lacustres
estaacuten relacionados con grandes aportes de N de la cuenca que a su vez son
mayores cuanto mayor la superficie agriacutecola yo los pastizales mientras que tanto
las plantaciones forestales como los bosques favorecen la retencioacuten de nutrientes
en la cuenca (lo que se ve reflejado en un δ15N maacutes negativo) Esta tesis plantea
un cambio de estado en la dinaacutemica del N asociado a la introduccioacuten y expansioacuten
de las plantaciones forestales o bosques los que retienen el Nitroacutegeno en las
cuencas mientras que el clima juega un rol secundario
11
ABSTRACT
The Anthropocene is characterized by human disturbances at the global
scale For example changes in land use are known to disturb the N cycle since the
industrial revolution but especially since the Great Acceleration (1950 CE) onwards
This impact has changed N availability in both terrestrial and aquatic ecosystems
However there are some important uncertainties associated with the extent of this
impact and how it is coupled to ongoing climate change (ie megadroughts rainfall
variability and shifting stationarity) or to other nutrient cycles (e g carbon cycle)
Lake sediments contain paleoenvironmental information regarding the conditions of
the watershed and associated lakes and which the respective sediments are
deposited Nitrogen stable isotope analysis (δ15N) on lake sediments allows us to
reconstruct the changes in N availability through time Here we used a multiproxy
approach that uses sedimentological geochemical and isotopic analyses on
lacustrine sediments water column and soilvegetation from the watershed as well
12
as land use and cover change (LUCC) from 1975 to 2016 obtained from satellite
images The goal of this thesis is to evaluate the role of LUCC as the main driver for
N cycling in a coastal watershed system of central Chile over the last centuries Our
main results show that more positive δ15N values in lake sediments are related to
higher N contributions from the watershed which in turn increase with increased
agricultural andor pasture cover whereas either forest plantations or native forests
can favor nutrient retention in the watershed (δ15N more negative) This thesis
proposes that N dynamics are mainly driven by the introduction and expansion of
forest or tree plantations that retain nitrogen in the watershed whereas climate plays
a secondary role
13
INTRODUCCIOacuteN
El N es un elemento esencial para la vida y limita la productividad en
ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al 1997) Las actividades
humanas han tenido un profundo impacto sobre el ciclo del N global principalmente
a partir la revolucioacuten industrial (IPCC 2013 2007) Las concentraciones de N se
han incrementado por la fijacioacuten industrial de N atmosfeacuterico (viacutea el proceso Haber-
Bosch Erisman et al 2008) por el uso de fertilizantes sobre la base de N para
mejorar el rendimiento de los cultivos la produccioacuten ganadera y ademaacutes de los
cambios en el uso del suelo (abreviado en ingleacutes- LUCC) (Robertson y Vitousek
2009 Galloway et al 2008 Battye et al 2017 Xu et al 2019) Estas actividades
contribuyen significativamente al incremento de la disponibilidad bioloacutegica de N
cuyas consecuencias para los ecosistemas incluye la perdida de diversidad
modificacioacuten de la disponibilidad de nutrientes y acidificacioacuten de los suelos entre
otras Sin embargo se desconoce la cantidad destino y el alcance que ha tenido
14
el N movilizado entre los ecosistemas generado por la influencia de las actividades
humanas (Vitousek et al 1997)
La entrada natural de N en los ecosistemas terrestres y acuaacuteticos es viacutea
fijacioacuten atmosfeacuterica la que es llevada a cabo por microorganismos muchos de ellos
en relaciones simbioacuteticas (eg Rhizobium en leguminosas) y las algas (Vitousek et
al 1997 Battye et al 2017) Las principales salidas estaacuten relacionadas con la
desnitrificacioacuten volatilizacioacuten del amoniaco y por escurrimiento superficial y
subterraacuteneo (Galloway et al 2008 Diacuteaz et al 2016) En el caso de los sistemas
lacustres ademaacutes estaacuten la resuspencioacuten de N del fondo del lago los aportes desde
la cuenca (entradas de N) La depositacioacuten de MO en el fondo del lago que marca
la salida de N de la columna de agua Estas relaciones de intercambio de N tienen
un control geograacutefico (eg latitud exposicioacuten relieve etc) climaacutetico y biotico
(McLauchlan et al 2013) La alteracioacuten provocada por los LUCC no solo genera
las peacuterdidas de nutrientes del suelo por erosioacuten y escurrimiento superficial sino que
tambieacuten modifica la disponibilidad y transferencia de N entre los ecosistemas
terrestres y acuaacuteticos (Elbert et al 2012 McLauchlan et al 2013) Por ejemplo el
reemplazo de la vegetacioacuten nativa por la agricultura yo plantaciones forestales
altera el ciclo del N debido al despeje de los terrenos viacutea quema de la vegetacioacuten
pero tambieacuten debido al reemplazo de la vegetacioacuten preexistente por un
monocultivo (eg trigo pino) y el uso ineficiente de fertilizantes para mejorar el
rendimiento agriacutecola Estas actividades favorecen un incremento de los aportes de
N (y otros nutrientes) viacutea sistemas de drenajes a lagos los que actuacutean como
sumideros El incremento del N derivado de las actividades humanas tanto en los
ecosistemas terrestres como acuaacuteticos genera incertezas relacionados con la
trayectoria y la magnitud de la disponibilidad de N en ambos ecosistemas (Batye et
15
al 2017 Mclauchlan et al 2017) Por lo tanto se requiere de una liacutenea base de
largo plazo para saber coacutemo estas perturbaciones impactan la disponibilidad de N
en las uacuteltimas deacutecadassiglos en los ecosistemas lacustres y por lo tanto el alcance
real que los LUCC han tenido en el ciclo del N
Los ecosistemas mediterraacuteneos y el ciclo del N
Los ecosistemas mediterraacuteneos son especialmente sensibles a los LUCC
pues estos se encuentran sometidos a un estreacutes hiacutedrico durante largas temporadas
estivales y las precipitaciones se concentran en eventos puntuales y a veces con
altos montos pluviomeacutetricos Las precipitaciones tienen un alto potencial de arrastre
de sedimentos y nutrientes los que actuacutean como fertilizantes al llegar a los
ecosistemas lacustres A su vez un incremento en la disponibilidad de N puede
generar un desbalance con otros nutrientes (eg Fosforo) con efectos sobre la
productividad y diversidad de los lagos (Pentildeuelas et al 2012 Sardans et al 2012
McLauchlan 2013) En este sentido en Chile central se observa una disminucioacuten
de las precipitaciones especialmente fuerte en las uacuteltimas deacutecadas por lo que se ha
denoacuteminado como Megasequiacutea (Garreaud et al 2017) y el efecto de las
precipitaciones esporaacutedicas pueden generar un arrastre de nutrientes que no ha
sido evaluado
Los ecosistemas mediterraacuteneos concentran el 20 de la biodiversidad global
(Myers et al 2000) pero existe una escasez de conocimiento respecto a los
efectos del incremento de N en los cuerpos de agua como consecuencia de las
actividades humanas en el pasado histoacuterico (Ochoa-Hueso et al 2011) y coacutemo la
disponibilidad de N ha variado en el tiempo Los LUCC han aumentado el flujo de
N en las cuencas hidrograacuteficas viacutea escurrimiento superficial y lixiviacioacuten
16
favoreciendo la peacuterdida de sedimentos y nutrientes del suelo en el mundo entero
(Elser et al 2011 McLauchlan et al 2013 Xu et al 2017 y 2019) Ello ha
contribuido a la acidificacioacuten y eutrofizacioacuten generalizada de riacuteos y lagos
(McLauchlan et al 2013 Schindler et al 2008)
El ecosistema mediterraacuteneo de Chile central es una regioacuten ampliamente
intervenida desde al menos la colonizacioacuten espantildeola (1600-1800 CE) Los LUCC
han tenido efectos negativos en la disponibilidad de agua especialmente
observados con el desarrollo de las actividades minera (Prieto et al 2019) Aunque
se sabe de los impactos en los ecosistemas de bosque en teacuterminos de cobertura
debidos al desarrollo silvo-agropecuarias (Armesto et al 2010) se desconoce el
impacto de estas actividades sobre el ciclo del N en los cuerpos de agua o en queacute
momento cobran fuerza suficiente para alterar el flujo de N hacia los lagos de Chile
Central Por lo tanto en esta tesis nos centramos en entender coacutemo los LUCC han
afectado la disponibilidad de N en los lagos costeros de Laguna Matanzas y Lago
Vichuqueacuten desde la llegada de los espantildeoles a Chile hasta el presente
Los lagos como sensores ambientales
Los sedimentos lacustres son buenos sensores de cambios en los aportes
de nutrientes contaminacioacuten y eutroficacioacuten de los cuerpos de agua ya que son
capaces de preservar sentildeales quiacutemicas y fiacutesicas al momento de su formacioacuten y
ademaacutes con alta resolucioacuten y a largo plazo (Leng et al 2006) Por lo tanto
constituyen una herramienta uacutetil para reconstruir el flujo de N entre ecosistemas
terrestres y acuaacuteticos en el pasado sus consecuencias (eg incremento de la
productividad lacustre) y el rol del clima y los LUCC en el pasado (McLauchlan et
al 2013 von Gunten et al 2009) Por ejemplo el cultivo de maiacutez por parte de los
17
nativos iroqueses en Canadaacute provocoacute la eutrofizacioacuten del Lago Crawford (43degN)
durante los siglos XII y XV Entre las consecuencias registradas se documentoacute un
claro incremento de la productividad primaria y cambios en la estructura comunitaria
de diatomeas foacutesiles (incremento de Stephanodiscus complex y disminucioacuten de
Cyclotella bodanica en Ekdahl et al 2007) Otro ejemplo demuestra como las
actividades agriacutecolas en la cuenca del riacuteo Mississippi modificaron los patrones de
sedimentacioacuten y aporte de nutrientes al lago Pepin EE UU (44degN) a partir del
asentamiento europeo en 1840 CE y principalmente desde 1940 CE (Engstrom et
al 2009) Para Chile von Gunten et al (2009) a partir de indicadores
limnogeoloacutegicos evidencioacute la contaminacioacuten y eutroficacioacuten de cinco lagos cercanos
a la ciudad de Santiago (33ordmS) como consecuencia de la deposicioacuten atmosfeacuterica
de metales pesados (especialmente Cu) quema de combustible foacutesil y flujo de
nutrientes durante los uacuteltimos 200 antildeos
Caracteriacutesticas limnoloacutegicas de los lagos
Los procesos limnoloacutegicos afectan la distribucioacuten y abundancia de los
organismos en los lagos Estaacuten influenciados por forzamientos externos por
ejemplo la precipitacioacuten o la penetracioacuten de la luz y la morfometriacutea del lago En este
sentido el nivel del lago dependeraacute del balance entre las entradas y salidas de agua
(precipitaciones escurrimiento superficial y subterraacuteneo y la evaporacioacuten) y la forma
de la cuenca (profundidad pendiente aacuterea del espejo de agua)
En funcioacuten de la profundidad de penetracioacuten de la luz es posible diferenciar
dos aacutereas que determinan la distribucioacuten de los organismos la zona foacutetica (donde
penetra la luz y permite la fotosiacutentesis) y la zona afoacutetica (subyacente a la zona
foacutetica) Al depender de la radiacioacuten la zona foacutetica variacutea estacionalmente Ademaacutes
18
puede variar por el grado de turbidez la morfometriacutea del lago y la produccioacuten de
materia orgaacutenica en la columna de agua
Otro factor que influye en la productividad es el reacutegimen de mezcla de la
columna de agua pues favorece la oxigenacioacuten y el reciclado de nutrientes La
mezcla ocurre en condiciones de gran inestabilidad usualmente relacionado con el
reacutegimen de viento Por el contrario un lago estratificado resulta de grandes
diferencias en temperatura o salinidad entre la superficie (epilimnion) y el fondo del
lago (hipolimnion) que separa las masas de agua superficial y de fondo por una
termoclina (si la estratificacioacuten estaacute relacionada con diferencias de temperatura de
las masas de agua) o haloclina (diferencias de salinidad) En funcioacuten del reacutegimen
de mezcla los lagos se pueden clasificar en (Lewis 1983)
1 Amiacutecticos no hay mezcla vertical de la columna de agua
2 Monomiacutecticos friacuteos y caacutelidos se mezcla una vez al antildeo
3 Dimiacutecticos la columna de agua se mezcla dos veces al antildeo
4 Oligomiacutecticos Lagos estratificados la mayor parte del tiempo pero se mezclan a
intervalos irregulares mayores a 1 antildeo
5 Polimiacutecticos friacuteos y caacutelidos la columna de agua se mezcla varias veces al antildeo
El ciclo del N en lagos
Al estar presente en aacutecidos nucleicos aminoaacutecidos y proteiacutenas el N es un
nutriente esencial de los organismos Por lo tanto su disponibilidad en la columna
de agua es clave para la produccioacuten composicioacuten y acumulacioacuten de la MO a traveacutes
del tiempo (Olson 1998 Talbot 2002) Aunque el principal reservorio de N estaacute en
19
la atmosfera (N2) solo unos pocos organismos (eg cianobacterias) pueden fijarlo
directamente Asiacute el Nitroacutegeno Inorgaacutenico Disuelto (DIN) representa la principal
fuente de N disponible para la produccioacuten primaria en los ecosistemas acuaacuteticos
(Talbot et al 2002) El DIN estaacute compuesto por nitrato (NO3-) nitrito (N02
-) y amonio
(NH4+) y los cambios en su disponibilidad condicionan el tipo de produccioacuten primaria
(eg fijadores vs asimiladores de N ver Scott y Grant 2013 Scott 2008)
La Figura 1 resume los principales componentes en lagos del ciclo del N y
sus interacciones La principal entrada natural de N es viacutea fijacioacuten de N atmosfeacuterico
y es llevado a cabo principalmente por bacterias y algas verde-azules capaces de
romper el triple enlace entre los dos aacutetomos de N a un alto costo energeacutetico (Torres
et al 2012) El N fijado es reducido a NH3 Tras la muerte de los organismos el N
es liberado de sus tejidos por hongos y bacterias implicados en la descomposicioacuten
de la MO Una parte se deposita en el fondo del lago y la otra queda disponible para
ser asimilada por el fitoplancton como amonio mediante el proceso de
amonificacioacuten (Talbot 2002) La amonificacioacuten resulta de la reduccioacuten bacteriana
del amoniaco y se da preferentemente en condiciones anoacutexicas La volatizacioacuten del
amoniaco es un proceso por medio este se incorpora a la atmoacutesfera Este proceso
se da preferentemente en lagos aacutecidos (pH gt 85) El nitrato es otra forma de N
bioloacutegicamente disponible para el fitoplancton En los lagos el NO3 del DIN estaacute
compuesto por los aportes desde la cuenca hidrograacutefica en la que se circunscriben
por la resuspensioacuten desde el fondo del lago favorecido por procesos de mezcla
(descrito en seccioacuten anterior) y por los nitratos de la columna de agua Estos uacuteltimos
son el resultado de la nitrificacioacuten proceso realizado por bacterias aeroacutebicas
mediante el cual el amonio se oxida para formar nitrito y nitrato El ciclo se completa
20
con la desnitrificacioacuten que es la reduccioacuten bacteriana de nitrato al N atmosfeacuterico
Este proceso se da preferentemente en condiciones anoacutexicas
Figura 1 Materia Orgaacutenica sedimentaria y su relacioacuten con el ciclo del N y las
variaciones de δ15N (elaborado a partir de Talbot et al 2002) En la figura se
representan los factores clave en la acumulacioacuten de la MO sedimentaria y su
relacioacuten con el ciclo del N El δ15N de la MO es resultado de los aportes de MO
desde la cuenca MO de macroacutefitas y de la productividad bioloacutegica La productividad
en ecosistemas lacustres estaacute condicionada por DIN y la fijacioacuten de N atmosfeacuterico
El δ15N del pool del DIN disponible se va enriqueciendo por la asimilacioacuten
preferencial de 14N por parte del fitoplancton Ademaacutes el DIN remanente se va
enriqueciendo por accioacuten microbiana (amonificacioacuten nitrificacioacuten desnitrificacioacuten)
Reconstruyendo el ciclo del N a partir de variaciones en δ15N
La sentildeal isotoacutepica de N (δ15N) en los sedimentos lacustres puede ser usada
para reconstruir los cambios pasados del ciclo N la transferencia de N entre
ecosistemas terrestres y acuaacuteticos y el estado troacutefico del lago (Bruland y Mackenzie
2015 McLauchlan et al 2013 Woodward et al 2012 von Gunten et al 2009
Leng et al 2006 Meyers y Teranes 2001) La Figura 1 resume los principales
procesos y fuentes que afectan la sentildeal isotoacutepica δ15N (total o ldquobulkrdquo) en la MO de
21
los sedimentos lacustres Entre los que destacan las fuentes de MO (aloacutectona vs
autoacutectona) la bioproductividad lacustre y las entradas de N (ie fijacioacuten bioloacutegica
de N2) (Torres et al 2012) Estos procesos conducen a un fraccionamiento
isotoacutepico cineacutetico que favorece la disociacioacuten entre sus isotopos ldquopesadosrdquo (15N) y
ldquoligerosrdquo (14N) Asiacute por ejemplo los valores de δ15N de la MO se enriquecen en 15N
en la medida que los sedimentos lacustres estaacuten sometidos a perdidas de 14N viacutea
desnitrificacioacuten (Bruland y Mackenzie 2015 Diaz et al 2016 Talbot et al 2002)
Por otra parte el fitoplancton en la medida que aumenta su biomasa (eg
durante la estacioacuten caacutelida) puede llegar a asimilar el DIN hasta agotarse En este
caso el N remanente se va empobreciendo en 14N si no hay reposicioacuten de N (eg
aporte de NO3 de la cuenca) y la MO queda enriquecida en 15N Esta disminucioacuten
induce a una limitacioacuten de fitoplancton por N y los organismos diztroacuteficos pueden
verse estimulados por lo que se incrementa la entrada de N viacutea fijacioacuten de N2 (Scott
y Grantsz 2013) como resultado la MO oscila en valores maacutes bajos (en torno a 0permil)
La cantidad de MO que se deposita en el fondo del lago depende del
predominio de contribuciones provenientes desde la cuenca (aloacutectona) versus las
producidas dentro del mismo lago (autoacutectona) (Torres et al 2012) Aunque en
general los lagos reciben permanentemente aportes de sedimentos y MO desde
su cuenca los lagos pequentildeos tienden a reflejar maacutes los procesos que ocurren
solamente en su cuenca directa y reflejan los valores isotoacutepicos de la misma (Gu et
al 2006) Woodward et al (2012) realizaron un estudio en 50 lagos en Irlanda que
les permitioacute correlacionar diferentes patrones de LUCC (bosques de coniacuteferas
agricultura ganaderiacutea entre otros) con los valores de δ15N (y δ13C) en los
sedimentos superficiales de los lagos Encontraron que los valores de δ15N son maacutes
negativos en cuencas poco impactadas y maacutes positivos en aquellas con alto
22
impacto humano (eg usos de suelo agriacutecola y ganadero) McLauchlan et al (2007)
encontraron valores maacutes positivos de δ15N en los sedimentos del Mirror Lake New
Hampshire (29ordm N) a partir de la desforestacioacuten de su cuenca en 1790 CE (inicio
del asentamiento europeo en el lago) para el desarrollo de la actividad agriacutecola
Estos valores se volvieron maacutes negativos hacia valores similares al pre-
asentamiento europeo despueacutes del cese de la agricultura en la cuenca y la
recuperacioacuten del bosque a partir de 1929 CE
El anaacutelisis de δ15N en los sedimentos lacustres es una herramienta casi sin
explorar en Chile central Proponemos reconstruir la dinaacutemica de transferencia de
N cuenca-lago como consecuencia de los LUCC desde la colonizacioacuten espantildeola en
los lagos costeros de Laguna Matanzas (34ordmS) y Lago Vichuqueacuten (36ordmS) Como son
muacuteltiples los procesos que pueden afectar la sentildeal isotoacutepica de N (medido como
δ15N) en los sedimentos lacustres existen muchos problemas para su
interpretacioacuten paleoambiental (Leng et al 2006) Por ello en esta tesis utilizamos
un enfoque multiproxy que consiste en integrar el anaacutelisis limnoloacutegico y geoquiacutemico
de los sedimentos lacustres con los valores isotoacutepicos modernos de la columna de
agua (mediante monitoreo del POM) la relacioacuten vegetacioacuten-suelo de la cuenca y la
reconstruccioacuten de los principales usos de suelo de las cuencas desde 1975 CE
mediante el uso de imaacutegenes satelitales Considerando estas muacuteltiples liacuteneas de
evidencia nos planteamos utilizar las variaciones del δ15N para reconstruir los
cambios en la disponibilidad de N en los lagos costeros de Chile central y establecer
coacutemo y desde cuaacutendo las actividades humanas en la cuenca son lo suficientemente
importantes como para modificar el ciclo del N en los lagos desde la colonizacioacuten
espantildeola (siglo XVII)
23
Como hipoacutetesis planteamos que la dinaacutemica de transferencia de sedimentos
y nutrientes entre la cuenca y el lago tiene controles geoloacutegicos climaacuteticos y
bioloacutegicos que interactuacutean en el tiempo registraacutendose en la acumulacioacuten de MO de
los sedimentos lacustres (Fig1) Bajo este marco los LUCC modifican esta
dinaacutemica alterando la transferencia de N a los lagos ya que las actividades agriacutecolas
y ganaderas tienden a aportar maacutes N (aumentando δ15N) que la actividad forestal
(la que disminuye δ15N)
En el primer capiacutetulo titulado ldquoA combined approach to establishing the timing
and magnitude of anthropogenic nutrient alteration in a mediterranean coastal lake-
watershed systemrdquo se evaluoacute el rol de los LUCC en el control de la descarga de N
y sedimentos a Laguna Matanzas en un sistema cuenca-lago desde el siglo XVII
Para ello se realizoacute un anaacutelisis de la dinaacutemica sedimentaria lacustre mediante el
anaacutelisis de muacuteltiples proxys sedimentoloacutegicos (ie minerales y textura)
geoquiacutemicos (eg geoquiacutemica orgaacutenica e isotoacutepica) y bioloacutegicos (ie diatomeas) de
Laguna Matanzas que cubren los uacuteltimos 200 antildeos Ademaacutes se realizoacute una
reconstruccioacuten de los usos de suelo entre 1975 y 2016 a partir de imaacutegenes de
sateacutelites y se colectaron muestras de suelo de las principales coberturas de la
cuenca a los cuales se midioacute el δ15N
Entre los principales resultados obtenidos se destaca la influencia de la
ganaderiacutea y la denitrificacioacuten lo que explica los altos valores de δ15N evidenciados
por el registro lacustre durante la colonizacioacuten espantildeola e inicios de la repuacuteblica A
partir de la Gran Aceleracioacuten (1950 CE) la desforestacioacuten y el reemplazo de la
ganaderiacutea por plantaciones forestales tienen un correlato en el registro
sedimentario con valores de δ15N maacutes bajos Estos resultados demuestran que los
LUCC son el factor de primer orden para explicar los cambios observados en
24
nuestro registro de δ15N y a su vez implica un rol secundario del clima como posible
control de la disponibilidad de N en Laguna Matanzas Este capiacutetulo fue sometido
a la revista Scientific Reports y estaacute en revisioacuten desde el 26 de marzo de 2019 En
la elaboracioacuten del mauscrito han colaborado Claudio Latorre Blas Valero-Garceacutes
Matiacuteas Frugone Pablo Sarricolea Santiago Giralt Manuel Contreras-Lopez
Ricardo Prego y Patricia Bernardez
El segundo capiacutetulo se titula ldquoStable isotopes track land use and cover
changes in a mediterranean lake in central Chile over the last 600 yearsrdquo Aquiacute
evaluamos la dinaacutemica del N actual en el sistema cuenca-lago considerando los
valores isotoacutepicos modernos tanto de la vegetacioacuten como del suelo ademaacutes de los
cambios estacionales del POM de la columna de agua Esta informacioacuten se utiliza
como un anaacutelogo moderno para conocer el rol de los LUCC en la disponibilidad de
N en los uacuteltimos 600 antildeos Para ello se colectaron muestras filtradas de la columna
de agua a 2 5 y 20 m dos veces por antildeo entre noviembre de 2015 y julio de 2018
y se obtuvo el δsup1⁵N del POM Ademaacutes se colectoacute muestras de vegetacioacuten y suelo
de la cuenca diferenciando entre especies nativas plantaciones forestales y
vegetacioacuten herbaacutecea Esta informacioacuten se analizoacute en conjunto con la reconstruccioacuten
de los LUCC entre 1975 y 2014 y el anaacutelisis limnoloacutegico del lago Ello nos permitioacute
evaluar coacutemo el δ15N de los sedimentos lacustres refleja los valores δ15N de la
cuenca en el Lago Vichuqueacuten y el control que ejerce el clima en la sentildeal isotoacutepica
de POM Este capiacutetulo seraacute sometido a la revista Science of total environmet
proximamente En la elaboracioacuten del mauscrito han colaborado Claudio Latorre
Blas Valero-Garceacutes Matiacuteas Frugone Pablo Sarricolea Leonardo Villacis y M Laura
Carrevedo
25
Entre los principales resultados encontramos que el δ15N en los sedimentos
lacustres es maacutes positivo con presencia de agricultura en la cuenca y maacutes negativo
cuando esta es reemplazada por cobertura arboacuterea bien sea plantaciones
forestales bien sea bosque nativo A su vez durante la estacioacuten seca y caacutelida la
mayor entrada al lago es viacutea fijacioacuten de N atmosfeacuterico (bajos valores de δ15NPOM)
Durante la estacioacuten huacutemeda en cambio prevalecen los aportes de la cuenca con
altos valores de δ15NPOM Ello estariacutea evidenciando un cambio estacional en la
composicioacuten de especies en verano ligado a especies fijadoras de N y en invierno
las algas y microorganismos que consumen el DIN de la columna de agua
Referencias
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea AM Marquet PA 2010 From the Holocene to the Anthropocene A historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the
next carbon Earth rsquo s Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409 httpsdoiorg102134jeq20090005
Diacuteaz FP Frugone M Gutieacuterrez RA Latorre C 2016 Nitrogen cycling in an
extreme hyperarid environment inferred from δ15 N analyses of plants soils and herbivore diet Sci Rep 6 1ndash11 httpsdoiorg101038srep22226
Ekdahl EJ Teranes JL Wittkop CA Stoermer EF Reavie ED Smol JP
2007 Diatom assemblage response to Iroquoian and Euro-Canadian eutrophication of Crawford Lake Ontario Canada J Paleolimnol 37 233ndash246 httpsdoiorg101007s10933-006-9016-7
Elbert W Weber B Burrows S Steinkamp J Buumldel B Andreae MO
Poumlschl U 2012 Contribution of cryptogamic covers to the global cycles of carbon and nitrogen Nat Geosci 5 459ndash462
26
httpsdoiorg101038ngeo1486 Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506
httpsdoiorg101126science1215567 ARTICLE Engstrom DR Almendinger JE Wolin JA 2009 Historical changes in
sediment and phosphorus loading to the upper Mississippi River Mass-balance reconstructions from the sediments of Lake Pepin J Paleolimnol 41 563ndash588 httpsdoiorg101007s10933-008-9292-5
Erisman J W Sutton M A Galloway J Klimont Z amp Winiwarter W (2008)
How a century of ammonia synthesis changed the world Nature Geoscience 1(10) 636
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892
httpsdoiorg101126science1136674 Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J Christie
D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592 Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
IPCC 2013 Climate Change 2013 The Physical Science BasisContribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change Cambridge University Press Cambridge United Kingdom and New York NY USA
httpsdoiorg101017CBO9781107415324 Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007 Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32iumliquestfrac12S 3050iumliquestfrac12miumliquestfrac12asl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470
27
httpsdoiorg101073pnas0701779104 McLauchlan KK Gerhart LM Battles JJ Craine JM Elmore AJ Higuera
PE Mack MC McNeil BE Nelson DM Pederson N Perakis SS 2017 Centennial-scale reductions in nitrogen availability in temperate forests of the United States Sci Rep 7 1ndash7 httpsdoiorg101038s41598-017-08170-z
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Myers N Mittermeler RA Mittermeler CG Da Fonseca GAB Kent J
2000 Biodiversity hotspots for conservation priorities Nature httpsdoiorg10103835002501
Pentildeuelas J Poulter B Sardans J Ciais P Van Der Velde M Bopp L
Boucher O Godderis Y Hinsinger P Llusia J Nardin E Vicca S Obersteiner M Janssens IA 2013 Human-induced nitrogen-phosphorus imbalances alter natural and managed ecosystems across the globe Nat Commun 4 httpsdoiorg101038ncomms3934
Prieto M Salazar D Valenzuela MJ 2019 The dispossession of the San
Pedro de Inacaliri river Political Ecology extractivism and archaeology Extr Ind Soc httpsdoiorg101016jexis201902004
Robertson GP Vitousek P 2010 Nitrogen in Agriculture Balancing the Cost of
an Essential Resource Ssrn 34 97ndash125 httpsdoiorg101146annurevenviron032108105046
Sardans J Rivas-Ubach A Pentildeuelas J 2012 The CNP stoichiometry of
organisms and ecosystems in a changing world A review and perspectives Perspect Plant Ecol Evol Syst 14 33ndash47 httpsdoiorg101016jppees201108002
Schindler DW Howarth RW Vitousek PM Tilman DG Schlesinger WH
Matson PA Likens GE Aber JD 2007 Human Alteration of the Global Nitrogen Cycle Sources and Consequences Ecol Appl 7 737ndash750 httpsdoiorg1018901051-0761(1997)007[0737haotgn]20co2
Scott E and EM Grantzs 2013 N2 fixation exceeds internal nitrogen loading as
a phytoplankton nutrient source in perpetually nitrogen-limited reservoirs Freshwater Science 2013 32(3)849ndash861 10189912-1901
28
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Talbot M R (2002) Nitrogen isotopes in palaeolimnology In Tracking
environmental change using lake sediments (pp 401-439) Springer Dordrecht
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable
isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K
Yeager KM 2016 Different responses of sedimentary δ15N to climatic changes and anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
29
CAPIacuteTULO 1 A COMBINED APPROACH TO ESTABLISHING THE TIMING
AND MAGNITUDE OF ANTHROPOGENIC NUTRIENT ALTERATION IN A
MEDITERRANEAN COASTAL LAKE- WATERSHED SYSTEM
30
A combined approach to establishing the timing and magnitude of anthropogenic
nutrient alteration in a mediterranean coastal lake- watershed system
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugoneabc Pablo
Sarricolead Santiago Giralte Manuel Contreras-Lopezf Ricardo Prego g Patricia
Bernaacuterdez g Blas Valero-Garceacutesch
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Institute of Earth Sciences Jaume Almera (ICTJA-CSIC) CLluis Solegrave Sabaris sn Barcelona E-
08028 Spain
f Facultad de Ingenieriacutea y Centro de Estudios Avanzados Universidad de Playa Ancha Traslavintildea
450 Vintildea del Mar Chile
g Instituto de Investigaciones Marinas (CSIC) CEduardo Cabello 6 36208 Vigo Spain
h Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding author
E-mail address
clatorrebiopuccl magdalenafuentealbagmailcom
Abstract
Since the industrial revolution and especially during the Great Acceleration (1950
CE) human activities have profoundly altered the global nutrient cycle through land
use and cover changes (LUCC) However the timing and intensity of recent N
variability together with the extent of its impact in terrestrial and aquatic ecosystems
and coupled effects of regional LUCC and climate are not well understood Here
we used a multiproxy approach (sedimentological geochemical and isotopic
31
analyses historical records climate data and satellite images) to evaluate the role
of LUCC as the main control for N cycling in a coastal watershed system of central
Chile during the last few centuries The largest changes in N dynamics occurred in
the mid-1970s associated with the replacement of native forests and grasslands for
livestock grazing by plantations of introduced Monterey pine (Pinus radiata) and
eucalyptus (Eucalyptus globulus) Lake productivity increased and is evidenced by
an increase trend in δ15N values Our study shows that anthropogenic land
usecover changes are key in controlling nutrient supply and N availability in
Mediterranean watershed ndash lake systems and that large-scale forestry
developments during the mid-1970s likely caused the largest changes in central
Chile
Keywords Anthropocene Organic geochemistry watershedndashlake system Stable
Isotope Analyses Land usecover change Nitrogen cycle Mediterranean
ecosystems central Chile
1 INTRODUCTION
Human activities have become the most important driver of the nutrient cycles in
terrestrial and aquatic ecosystems since the industrial revolution (Gruber and
Galloway 2008 Galloway et al 2008 Holtgrieve et al 2011 Fowler et al 2013
Goyette et al 2016) Among these N is a common nutrient that limits productivity
in terrestrial and aquatic ecosystems (Vitousek and Howarth 1991 McLauchlan et
al 2013) With the advent of the Haber-Bosch industrial N fixation process in the
early 20th century total N fluxes have surpassed previous planetary boundaries
32
(Galloway et al 2008 Elser 2011) reaching unprecedented values (ie tipping
points) in the Earth system especially during what is now termed the Great
Acceleration (which began in the 1950s) which intensified in the 1970s (Howarth
2004 Steffen et al 2015) Yet dramatic changes in LUCC have occurred in the last
few centuries mainly linked to forestry and agro-pastoral activities (Poraj-Goacuterska et
al 2017 Ge et al 2019 Cheng et al 2019) These have had large impacts on N
(and Carbon -C-) availability in terrestrial and aquatic ecosystems but the synergic
effect with climate change and global N dynamics has not been established
(Vitousek et al 1997 Mclauchlan et al 2007 2013 Pongratz et al 2010
Woodward et al 2012 Mclauchlan et al 2017)
The onset of the Anthropocene poses significant challenges in mediterranean
regions that have a strong seasonality of hydrological regimes and an annual water
deficit (Stocker et al 2013) Mediterranean climates occur in all continents
(California central Chile Australia South Africa circum-Mediterranean regions)
providing a unique opportunity to investigate global change processes during the
Anthropocene in similar climate settings but with variable geographic and cultural
contexts The effects of global change in mediterranean watersheds have been
analyzed from different perspectives hydrology (Giorgi 2006 Arnell and Gosling
2013 Prudhomme et al 2014) vegetation dynamics (Lenihan et al 2003 Vicente-
Serrano et al 2013 Matesanz and Valladares 2014) sediment dynamics (Garciacutea-
Ruiz et al 2013 Garciacutea-Ruiz 2010 Syvitski et al 2005) changes in
biogeochemical cycles (Thomas et al 2013 Fowler et al 2013 Frank et al 2015)
carbon storage (Muntildeoz-Rojas et al 2015) and biodiversity (Hooper et al 2012) A
recent review showed an extraordinarily high variability of erosion rates in
mediterranean watersheds positive relationships with slope and annual
33
precipitation and the paramount effect of land use (Garciacutea-Ruiz et al 2015)
However the temporal context and effect of LUCC on nutrient supply to
mediterranean lakes has not been analyzed in much detail
Major LUCC in central Chile occurred during the Spanish Colonial period
(17th-18th centuries) (Armesto 1994 Gurnell et al 2001 Figueroa et al 2004
Martiacuten-Foreacutes et al 2012 Contreras-Loacutepez et al 2014) with the onset of
industrialization and mostly during the mid to late 20th century (von Gunten et al
2009 Gayo et al 2012) Recent copper pollution caused by 20th century mining
and industrial smelters has been documented in cores throughout the Andes
(Laguna el Ocho and Laguna Ensuentildeo von Gunten et al 2009) and also from our
surveys in the central valley (Batuco wetland) coastal range (Cordillera de Name)
and along the coast itself (Bucalemito and Colejuda) (Valero-Garceacutes et al 2010
unpublished data)
Paleolimnological studies have shown how these systems respond to
climate LUCC and anthropogenic impacts during the last millennia (Jenny et al
2002 2003 Villa Martiacutenez et al 2002 Frugone-Aacutelvarez et al 2017 Contreras et
al 2018) Furthermore changes in sediment and nutrient cycles have also been
identified in associated terrestrial ecosystems dating as far back as the Spanish
Conquest and related to fire clearance and wood extraction practices of the native
forests (Armesto 1994 Contreras-Loacutepez et al 2014) Nevertheless pollen and
limnological evidence argue for a more recent timing of the largest anthropogenic
impacts in central Chile For example paleo records show that during the mid-20th
century increased soil erosion followed replacement of native forest by Pinus
radiata and Eucalyptus globulus plantations at Laguna Matanzas Aculeo and
34
Vichuqueacuten lakes (Jenny et al 2002 Villa-Martiacutenez et al 2002 2003 Frugone-
Aacutelvarez et al 2017)
Lakes are a central component of the global carbon cycle Lakes act as a
sink of the carbon cycle both by mineralizing terrestrially derived organic matter and
by storing substantial amounts of organic carbon (OC) in their sediments (Anderson
et al 2009) Paleolimnological studies have shown a large increase in OC burial
rates during the last century (Heathcote et al 2015) however the rates and
controls on OC burial by lakes remain uncertain as do the possible effects of future
global change and the coupled effect with the N cycle LUCC intensification of
agriculture and associated nutrient loading together with atmospheric N-deposition
are expected to enhance OC sequestration by lakes Climate change has been
mainly responsible for the increased algal productivity since the end of the 19th
century and during the late 20th century in lakes from both the northern (Ruumlhland et
al2015) and southern hemispheres (Michelutti et al 2015 Carrevedo et al 2015)
but many studies suggest a complex interaction of global warming and
anthropogenic influences and it remains to be proven if climate is indeed the only
factor controlling these transitions (Catalaacuten et al 2013) Alternative causes for
recent N (Galloway et al 2008) increases in high altitude lakes such as catchment
mediated processes cannot be ruled out (Camarero and Catalan 2012 Fritz and
Anderson 2013) Few lake-watershed systems have robust enough chronologies of
recent changes to compare variations in C and N with regional and local processes
and even fewer of these are from the southern hemisphere (McLauchlan et al
2007 Holtgrieve et al 2011)
In this paper we present a multiproxy lake-watershed study including N and
C stable isotope analyses on a series of short cores from Laguna Matanzas in
35
central Chile focused in the last 200 years We complemented our record with land
use surveys satellite and aerial photograph studies Our major objectives are 1) to
reconstruct the dynamics among climate human activities and changes in the N
cycle over the last two centuries 2) to evaluate how human activities have altered
the N cycle during the Great Acceleration (since the mid-20th century)
2 STUDY SITE
Laguna Matanzas (33ordm45rsquoS 71ordm 40acuteW 7 m asl Fig 1) is a coastal lake located
in central Chile near to a large populated area (Santiago gt6106 inhabitants) The
lake has a surface area of 15 km2 with a max depth of 3 m and a watershed of 30
km2 The lake basin is emplaced over Pleistocene-Holocene aeolian and alluvial fan
deposits (SERNAGEOMIN 2003) A recent phase of dune activity occurred from the
mid to late Holocene which mostly sealed off the basin from the ocean (Villa-
Martinez 2002) Climate in Laguna Matanzas is characterized by cool-wet winters
and hot-dry summers with annual precipitation of ~510 mm and a mean annual
temperature of 12ordmC Central Chile is in the transition zone between the southern
hemisphere mid-latitude westerlies belt and the South Pacific Anticyclone (or SPA)
(Garreaud et al 2009) In winter precipitation is modulated by the north-west
displacement of the SPA the northward shift of the westerlies wind belt and an
increased frequency of storm fronts stemming off the Southern Hemisphere
Westerly Winds (SWW) (Frugone-Aacutelvarez et al 2017) Austral summers are
typically dry and warm as a strong SPA blocks the northward migration of storm
tracks stemming off the SWW
36
Historic land cover changes started after the Spanish conquest with a Jesuit
settlement in 1627 CE near El Convento village and the development of a livestock
ranch that included the Matanzas watershed After the Jesuits were expelled from
South America in 1778 CE the farm was bought by Pedro Balmaceda and had more
than 40000 head of cattle around 1800 CE (Contreras-Lopez et al 2014) The first
Pinus radiata and Eucalyptus globulus trees were planted during the second half of
the 19th century and were mostly used for dune stabilization (Albert 1900 Gibson
1972) However the main plantation phase occurred 60 years ago (Villa-Martinez
2002) as a response to the application of Chilean Forestry Laws promulgated in
1931 and 1974 and associated state subsidies
Major land cover changes occurred recently from 1975 to 2008 as shrublands
were replaced by more intensive land uses practices such as farmland and tree
plantations (Schulz et al 2010) Laguna Matanzas is part of the Reserva Nacional
Humedal El Yali protected area (Ramsar Ndeg 878) Despite this protected status the
lake and its watershed have been heavily affected by intense agricultural and
farming activities during the last decades The main inlet ldquoLas Rosasrdquo has been
diverted for crop irrigation causing a significant loss of water input to the lake
Consequently the flooded area of the lake has greatly decreased in the last couple
of decades (Fig 1b) Exotic tree species cover a large surface area of the
watershed Recently other activities such as farms for intensive chicken production
have been emplaced in the watershed
37
Figure 1 Laguna Matanzas a) Digital Elevation Model showing the watershed and
the surface hydrologic connection with Colejuda and Cabildo lakes b) Climograph
depicting the warm dry season in austral summer c) Annual precipitation from
1965-2015 (arrow shows start of the most recent ldquomega-droughtrdquo- see Garreaud et
al 2017) d) A decline of lake area occurs between 2007 and 2018 Lake surface
area decreased first along the western sector (in 2007) followed by more inland
areas (in 2018)
38
3 RESULTS
31 Age Model
The age model for the Matanzas sequence was developed using Bacon software
to establish the deposition rates and age uncertainty (Blaauw and Christen 2011)
It is based on 210Pb and 137Cs dates and two 14C dates (Table 2) According to this
age model the lake sequence spans the last 1000 years (Fig 2) A major breccia
layer (unit 3b) was deposited during the early 18th century which agrees with
historic documents indicating that a tsunami impacted Laguna Matanzas and its
watershed in 1730 CE (Contreras-Loacutepez et al 2014) Here we focus in the last 200
years were the most important changes occurred in terms of LUCC (after the
sedimentary hiatus caused by the tsunami) The Spanish colonial period (17th -18th
century) brought new forms of territorial management along with an intensification
of watershed use which remained relatively unchanged until the 1900s
39
Figure 2 a) Age-depth model obtained for the Laguna de Matanzas sedimentary
sequence A hiatus occurs at 80 cm depth (unit 3b) The section of core used for our
analysis is highlighted in a red rectangle b) Close up of the age model used for
analysis of recent anthropogenic influences on the N cycle c) Information regarding
the 14C dates used to construct age model
Lab code Sample ID
Depth (cm) Material Fraction of modern C
Radiocarbon age
Pmc Error BP Error
D-AMS 021579
MAT11-6A 104-105 Bulk Sediment
8843 041 988 37
D-AMS 001132
MAT11-6A 1345-1355
Bulk Sediment
8482 024 1268 21
POZ-57285
MAT13-12 DIC Water column 10454 035 Modern
Table 2 Laguna Matanzas radiocarbon dates
32 The sediment sequence
Laguna Matanzas sediments consist of massive to banded mud with some silt
intercalations They are composed of silicate minerals (plagioclase quartz and clay
minerals) with relatively high TOC content (Fig 3) Pyrite is a common mineral
indicating dominant anoxic conditions in the lake sediments whereas aragonite
occurs only in the uppermost section Mineralogical analyses visual descriptions
texture and geochemical composition were used to characterize five main facies
(Fig 3) F1 (organic-rich mud) represents baseline sedimentation in a shallow well-
mixed brackish highly productive lake and F1acute (dark orange) is a less organic facies
than F1 (more details see table in the supplementary material) F2 (massive to
banded silty mud) indicates periods of higher clastic input into the lake but finer
(mostly clay minerals) likely from suspension deposition associated with flooding
40
events Aragonite (up to 15 ) occurs in both facies but only in samples from the
uppermost 30 cm and it is interpreted as endogenic linked to higher MgFe waters
and elevated biologic productivity
Figure 3 Sedimentary facies and units mineralogy grain size elemental geochemical
and isotopic composition of Laguna Matanzas core Values highlighted in gray indicate
that these are above average
The banded to laminated fining upward silty clay layers (F3) reflect
deposition by high energy turbidity currents The presence of aragonite suggests
that littoral sediments were incorporated by these currents Non-graded laminated
coarse silt layers (F4) do not have aragonite indicating a dominant watershed
41
sediment source Both facies are interpreted as more energetic flood deposits but
with different sediment sources A unique breccia layer with coarse silt matrix and
cm-long soft clasts (F5) is interpreted as a high-energy event (ie a tsunami)
capable of eroding the littoral zone and depositing coarse clastic material in the
distal zone of the lake Similar coarse breccia layers have been found at several
coastal sites in Chile and interpreted as tsunami-related deposits (Vargas et al
2005 Le Roux et al 2008)
33 Sedimentary units
Three main units and six subunits have been defined (Fig 3) based on
sedimentary facies and sediment composition We use ZrTi as an indicator of the
mineral fraction transported from the watershed (Marzecoacuteva et al 2011) as higher
ZrTi (F3 and F4) is commonly associated with coarser sediments (Cuven et al
2010) A high correlation among Br BrTi and TOC (r = 046-087 p value=0)
supports the use of BrTi as an indicator of lake productivity (Marzecoacuteva et al 2011
Frugone-Aacutelvarez et al 2017) The MnFe ratio is indicative of lake bottom
oxygenation (Naecher et al 2013) as under reducing conditions Mn mobilizes more
than Fe leading to a decreased MnFe ratio (Marzecoacuteva et al 2011) SrTi indicates
periods of increased aragonite formation as Sr is preferentially included in the
aragonite mineral structure (Veizer et al 1971) (See supplementary material)
The basal unit (3c 129 to 99 cm) is relatively organic-rich (TOC mean=26
BioSi mean=5) and composed by F1 (without aragonite) with some coarser F4
flood layers (ZrTi mean=025) Unit 3b (98 to 80 cm) is interpreted as a tsunami or
storm surge deposit (breccia F5 grading into massive to banded silt F2) Unit 3a
(79 to 73 cm) is characterized by relatively low productivity (TOC=2 BrTi=002
42
BioSi=4) under anoxic conditions (MnFe=001) Unit 2a (72 to 31cm) has
relatively less organic content and more intercalated clastic facies F3 and F4 The
top of this unit (43-30 cm) has elevated TS values The Subunit 1b (30-20 cm)
shows increasing TOC BioSi and BrTi values (TOC mean = 29 BioSi mean =
54 BrTi mean=004) and the upper subunit 1a (19-0 cm) has the highest TOC
(mean = 64) and BioSi (mean = 56) high BrTi mean = 010 and the presence
of aragonite More frequent anoxic conditions (MnFe lower than 001) during units
3 and 2 shifted towards more oxic episodes (MnFe = 003) in unit 1 (Fig 3)
34 Isotopic signatures
Figure 4 shows the isotopic signature from soil samples of the major land
usescover present in the Laguna Matanzas used as an end member in comparison
with the lacustrine sedimentary units δ15N from cropland samples exhibit the
highest values whereas grassland and soil samples from lake shore areas have
intermediate values (Fig 4) Tree plantations and native forests have similarly low
δ15N values (+11 permil SD=24) All samples (except those from the lake shore)
exhibit low δ13C values (from ndash285permil to ndash298permil) CNmolar from agriculture land
lakeshore area and non-vegetation areas samples display the lowest values (about
18) CNmolar from tree plantations and native forest have the highest values (383
and 267 respectively)
43
Figure 4 C-N stable isotope plot showing a comparison of lake sediments grouped
by sedimentary units (MAT11-6A) with the soil end members of present-day (lake
shore and land usecover) from Laguna Matanzas
The δ15N values from sediment samples (MAT11-6A) range from ndash15 and
+53permil (mean= +35 SD=05) δ13C values range from ndash266permil to ndash202permil (mean=
ndash240 SD=14) In unit 3c δ15N and δ13C show relatively high values (mean=
+41permil and ndash233permil respectively) δ15N and δ13C from unit 3b and 3a fluctuate at
slightly lower values than in 3c (mean δ15N from 3a= +38permil and 3b mean= +39permil
mean δ13C from 3a= ndash242permil and 3b mean= ndash244permil) In unit 2a δ15N values are
relatively high (mean= +38permil) but show a slightly decreasing trend (from +52 to
+24permil at 66 cm and 36 cm respectively) δ13C also decreases (mean= ndash242permil)
reaching minimum values at 45 cm (ndash266permil) with a sharp increase towards the top
of this unit with maximum values ca ndash210permil Unit 1 exhibits the lowest δ15N values
(1a mean= +11permil 1b mean= +28permil) with negative values in the uppermost
44
sediments (ndash04permil at 14 cm) δ13C values show a decreasing trend over most of
subunit 1b and increase only near the very top of this unit
35 Recent land use changes in the Laguna Matanzas watershed
Major LUCC from 1975 (unit 1b) to 2016 CE in the Laguna Matanzasacutes
watershed is summarized in Figure 5 The watershed has a surface area of 30 km2
of which native forest (36) and grassland areas (44) represented 80 of the
total surface in 1975 The area occupied by agriculture was only 02 and tree
plantations were absent Isolated burned areas (33) were located mostly in the
northern part of the watershed By 1989 tree plantations surface area had increased
to 5 burned areas to 17 and agricultural fields to 9 (a 45-fold increase) and
native forest and grassland sectors decreased to 23 and 27 respectively By
2016 agricultural land and tree plantations have increased to 17 of the total area
whereas native forests decreased to 21
45
Figure 5 Land uses and cover changes from 1975 to 2016 in Laguna Matanzas
watershed from natural cover and areas for livestock grazing (grassland) to the
expansion of agriculture and forest plantation
4 DISCUSSION
41 N and C dynamics in Laguna Matanzas
Small lakes with relatively large watersheds such as Laguna Matanzas would
be expected to have relatively high contributions of allochthonous C to the sediment
OM (Gu et al 2006) Terrestrial C3 plants (δ13C =ndash26 to ndash28permil Meyers and Terranes
2001 Ku et al 2007) are dominant in the Laguna Matanzas watershed Likewise
our soil samples ranged across similar although slightly more negative values
46
(δ13C = ndash30 to ndash28permil Fig 4) to those proposed by Meyers and Terranes (2001)
and are used here as terrestrial end members oil samples were taken from the lake
shore (δ13C = ndash22permil) and POM from surface water (δ13C = ndash24permil) were more
positive than the terrestrial end member and are used as lacustrine end members
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from terrestrial vegetation and more positive δ13C values have increased
aquatic OM (Gu et al 2006 Torres et al 2012) Phytoplankton preferentially uptake
12C leaving the DIC pool enriched in 13C (Gu et al 2006) especially when there are
no important external sources of C (eg decreased C input from the watershed)
Therefore during events of elevated primary productivity the phytoplankton uptakes
12C until its depletion and are then obligated to use the heavier isotope resulting in
an increase in δ13C Changes in lake productivity thus greatly affect the C isotope
signal (Torres et al 2012) with high productivity leading to elevated δ13C values
(Torres et al 2012 Gu et al 2006)
In a similar fashion the N isotope signatures in Laguna Matanzas reflect a
combination of factors including different N sources (autochthonousallochthonous)
and lake processes such as productivity isotope fractionation in the water column
and sediment denitrification Elevated δ15N values from a POM sample (+22permil) and
average values from the lake shore (mean=+34permil SD=028) are used as aquatic
end members whereas terrestrial samples have values from +10 +24 (tree species)
to +77permil +35 (agriculture) which we use as terrestrial end members (Fig 4)
Autochthonous OM in aquatic ecosystems typically displays low δ15N values
when the OM comes from N-fixing species Atmospheric fixation of N2 by
cyanobacteria ends in OM with δ15N values close to 0permil (Leng et al 2006)
Phytoplankton preferentially uptake 14N from Dissolved Inorganic Nitrogen (DIN) in
47
the water column and derived OM typically have δ15N values lower than DIN values
When productivity increases the remaining DIN becomes depleted in 14N which in
turn increases the δ15N values of phytoplankton over time especially if the N not
replenished (Torres et al 2012) Thus high POM δ15N values from Laguna
Matanzas reflect elevated phytoplankton productivity with a 14N- depleted DIN In
addition N-watershed inputs also contribute to high δ15N values Heavily impacted
watersheds by human activities are often reflected in isotope values due to land use
changes and associated modified N fluxes For example the input of N runoff
derived from the use of inorganic fertilizers leads to the presence of elevated δ15N
(between ndash4 to +4permil) values in the water bodies (Elliott and Brush 2006 Diebel and
Vander Zanden 2009) Widory et al (2004) reported a direct relationship between
elevated δ15N values and increased nitrate concentration from manure in the
groundwater of Brittany (France) Terranes and Bernasconi (2000) found a good
correlation between augmented nutrient loading and a progressive increase in δ15N
values of sedimentary OM related to agricultural land use
Post-depositional diagenetic processes can further affect C and N isotope
signatures Several studies have shown a decrease in δ13C values of OM in anoxic
environments particularly during the first years of burial related to the selective
preservation of OM depleted in 13C (Hollander and Smith 2001 Lehmann et al
2002 Gӓlman et al 2009 Torres et al 2012) Diagenetic processes can also lead
to post-burial N isotope enrichment of the sediments Indeed 14N is consumed more
rapidly than 15N by denitrifying bacteria which intensifies under anoxic conditions
(Diebel and Vander Zanden 2009) Thus the remaining OM pool will be enriched
in 15N leaving to elevated δ15N values (Granger et al 2008 Menzel et al 2013)
48
In summary the relatively high δ15N values in sediments of Laguna Matanzas
reflect N input from an agriculturegrassland watershed with positive synergetic
effects from increased lake productivity enrichment of DIN in the water column and
most likely denitrification The increase of algal productivity associated with
increased N terrestrial input andor recycling of lake nutrients (and lesser extent
fixing atmospheric N) and denitrification under anoxic conditions can all increase
δ15N values (Fig 3) In addition elevated lake productivity without C replenishing
(eg by terrestrial C input) produces shifts towards positive δ13C values whereas C
input from the watershed generates more negative δ13C values
42 Recent evolution of the Laguna Matanzas watershed
Sedimentological compositional and geochemical indicators show three
depositional phases in the lake evolution under the human influence in the Laguna
Matanzas over the last two hundred years Although the record is longer (around
1000 years) we analyzed the last two centuries (unit 2 and 1) to provide a recent
historical context for the large changes detected during the 20th century
The first phase lasted from the beginning of the 19th century until ca 1940
(unit 2a Fig 3 4) and was characterized by moderate productivity with elevated
sediment input from the watershed as indicated by our geochemical proxies (BrTi
= 004 AlTi gt 01 Fig 6) Higher lake levels and dominant anoxic bottom conditions
(MnFe = 001 Fig3) can be related to increased precipitation (mean= 557 mmyr)
and lower temperatures (summer annual temperature lt19ordmC) During the Spanish
colonial period the Laguna Matanzas watershed was used as a livestock farm
(Jesuits 1627-1780 unit 3 followed by the Pedro Balmaceda era 1780-1810- unit
2a) (Contreras-Loacutepez et al 2014) The main buildings and farm facilities at the El
49
Convento village During this period livestock grazing and lumber extraction for
mining would have involved extensive deforestation and loss of native vegetation
(eg Armesto et al 1994 2010) However the Matanzas pollen record does not
show any significant regional deforestation during this period (Villa Martiacutenez 2002)
suggesting that the impact may have been highly localized
Lake productivity sediment input and elevated precipitation (Fig 6) all
suggest that N availability was related to this increased input from the watershed
The N from cow manure and soil particles would have led to higher δ15N values
(Elliot and Brush 2016) In addition anoxic lake bottom conditions would lead to
even further enrichment of buried sediment N The δ13C values lend further support
to our interpretation of increased sediment input -and N- from the watershed
Decreased δ13C values reach their lowest values in the entire record (ndash270permil) at
ca 1910 CE (Fig 4 6)
During most of the 19th century human activities in Laguna Matanzas were
similar to those during the Spanish Colonial period However the appearance of
Pinus radiata and Eucalyptus globulus pollen ca 1890 CE (Gibson 1972) the dune
stabilization-afforestation program which began ca 1900 CE (Albert 1900) and the
application of the Chilean forestry law of 1931 (DFL nordm265) contributed to an
increased capacity of the surrounding vegetation to retain nutrients and sediments
The law subsidized forest plantations in areas devoid of vegetation and prohibited
the cutting of forest on slopes greater than 45ordm These land use changes were coeval
with decreased sediment inputs (AlTi trend) from the watershed slightly increased
lake productivity (BrTi from 001 to 003) and a decrease in annual precipitation
(Fig 6) N isotope values become more negative during this period although they
remained high (from +49permil to +37permil) whereas the δ13C trend towards more
50
positive values reflects changes in the N source from watershed to in-lake dynamics
(e g increased endogenic productivity)
The second phase started after 1940 and is clearly marked by an abrupt
change in the general trend of δ15N that oscillates around +41permil (SD = 04permil) during
the first phase to +24permil (SD = 14permil) during the second These δ15N trends reflect
the lowest watershed nutrient and sediment inputs (based on the AlTi record)
decreased precipitation (mean = 318 mm year) and a slight increase in lake
productivity (increased BrTI) Depositional dynamics in the lake likely crossed a
threshold as human activity intensified throughout the watershed and lake levels
decreased
During the Great Acceleration δ15N values shifted towards higher values to
ca 3permil with an increase in δ13C values that are not reflected either in lake
productivity or lake level As the sediment input from the watershed increased and
precipitation remained as low as the previous decade δ15N values during this period
are likely related to watershed clearance which would have increased both nutrient
and sediment input into the lake
The δ13C trend to more positive values reaching the peaks in the 1960s (ndash
212permil) at the same time as the peaks in temperature (196 ordmC mean annual) with a
downward trend in precipitation A shift in OM origin from macrophytes and
watershed input influences to increased lake productivity could explain this trend
(Fig 4 1b)
In the 1970s the Laguna Matanzasacute watershed was mostly covered by native
forest (36) and grassland areas were intended for livestock grazing (44 Fig 5)
Both soils have δ15NSoil lt 5permil (Fig 4) Farming fields occupied a very small area and
tree plantations were almost nonexistent The decreasing trend in δ15N values seen
51
in our record is interrupted by several large peaks that occurred between ca 1975
and ca 1989 when the native forest and grassland areas fell by 23 and 27
respectively largely due to fires affecting 17 of the forests Agriculture fields
increased by 4 and forest plantations increased by 9 (unit 1a) Concomitantly
sediment ndash and likely N - inputs from the watershed decreased (as indicated by the
trend in AlTi) the precipitation was still relatively low (Fig 6) These changes are
likely related to the increase of vegetation cover especially of tree plantations (which
have more negative δ15N values) The small increase in productivity in the lake could
have been favored by increased temperature (von Gunten et al 2009) After 1989
the increase in agriculture land (17 in 2016) correlates with increasing δ15N δ13C
and TOC trends in spite of declining rainfall The increase of forest plantations was
mostly in response to the implementation of the Law Decree of Forestry
Development (DL 701 of 1974) that subsidized forest plantation After 1989 the
increase in agricultural land (17 in 2016) is synchronous with increasing δ15N
δ13C TOC and MnFe trends despite the decline in rainfall and overall lower lake
levels as more water is used for irrigation
The third phase started c 1990 CE (unit 1a) when OM accumulation rates
increase and δ13C δ15N decreased reaching their lowest values in the sequence
around 2000 CE Afterward during the 21st century δ13C and δ15N values again
began to increase The onset of unit 1 is marked by increased lake productivity and
decreased sediment input (AlTi lt 02) synchronous with intensive farming replacing
forestry and extensive agriculture (Fig 5 6)
A change in the general trend of δ15N values which decreased until 1990
(unit 1a) as the native forest and grassland area fell to 23 and 27 respectively
is most likely due to deforestation and fires Agriculture surface increased to 4 and
52
forest plantations increased by 9 (Fig 5) Concomitantly sediment ndash and likely N
ndash inputs from the watershed decreased probably related to the low precipitation (Fig
1b) and the increase of vegetation cover in the watershed in particularly by tree
plantations (with more negative δ15N Fig 4)
At present agriculture and tree plantations occupy around 34 of the
watershed surface whereas native forests and grassland cover 21 and 25
respectively Lake productivity as indicated by BrTi (Fig6) is higher and generates
OM with lower δ15N and higher δ13C values (about +2permil and ndash20permil ca 2000 CE
respectively) due to in-lake processes (ie biological N fixation and nutrient
recycling) and driven by changes in the arboreal cover which diminishes nutrient
flux into the lake (Fig6)
53
Figure 6 Anthropogenic and climatic forcing and lake dynamics response
(productivity sediment input N and C cycles) at Matanzas Lake over the last two
54
centuries Mean annual precipitation reconstructed and temperatures (von Gunten
et al 2009) Vertical gray bars indicate mega-droughts
5 CONCLUSIONS
Human activities have been the main factor controlling the N and C cycle in
the Laguna Matanzas during the last two centuries The N isotope signature in the
lake sediments reflects changes in the watershed fluxes to the lake but also in-lake
processes such as productivity and post-depositional changes Denitrification could
have been a dominant process during periods of increased anoxic conditions which
were more frequent prior to Great Acceleration (1950 CE) Higher δ15N and lower
δ13C values are associated with increased nutrient input from the watershed due to
increased livestock grazing and agriculture pressure (phase 1 Fig 7a) whereas
lower isotope values occurred during periods of increased forest plantations (phase
3 Fig 7c) During periods of increased lake productivity - such as in the last few
decades - δ15N values increased significantly
The most important change in C and N dynamics in the lake occurred after the
1950s (Phase 2 and 3) as tree plantations dominated the watershed Recent
changes in N dynamics can be explained by the higher nutrient contribution
associated with intensive agriculture (i e fertilizers) since the 1990s Although the
replacement of livestock activities with forestry and farming seems to have reduced
nutrient and soil export from the watershed to the lake the inefficient use of fertilizer
(by agriculture) can be the ultimate responsible for lake productivity increase during
the last decades
55
Figure 7 Schematic diagrams illustrating the main factors controlling the
isotope N signal in sediment OM of Laguna Matanzas N input from watershed
depends on human activities and land cover type Agriculture practices and cattle
(grassland development) contribute more N to the lake than native forest and
plantations Periods of higher productivity tend to deplete the dissolved inorganic N
in 14N resulting in higher δ15N (OM) The denitrification processes are more effective
in anoxic conditions associated with higher lake levels
6 METHODS
Short sediment cores were recovered from Laguna Matanzas using an Uwitec
gravity piston corer in 2011 and 2013 (Fig 1a) Sediment cores (MAT11-6A 129 cm
MAT13-2A 149 cm MAT13-3A 33 cm MAT13-4A 99 cm) were split
photographed sub-sampled and stored at the Pyrenean Institute of Ecology (IPE-
CSIC Spain) Core MAT11-6A was obtained from the central sector of the lake and
56
was selected for detailed multiproxy analyses (including elemental geochemistry C
and N isotope analyses XRF and 14C dating)
The isotope analyses (δ13C and δ15N) were performed at the Laboratory of
Biogeochemistry and Applied Stable Isotopes (LABASI-PUC Chile) using a Delta
V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via a
Conflo IV interface Isotope results are expressed in standard delta notation (δ) in
per mil (permil) relative to the standards Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Sediment samples
for δ13Corg were pre-treated with 50 ml of HCl and 50 ml of deionized water and
dried at 60ordmC for 4 hr to remove carbonates (Harris et al 2011)
Total Carbon (TC) Total Organic Carbon (TOC) Total Inorganic Carbon (TIC)
and Total Sulphur (TS) were measured every cm with a LECO SC 144 DR at IPE-
CSIC XRF measurements were carried out every 4 mm in MAT11-1A-1G core using
an AVAATECH X-ray Fluorescence II core scanner at the University of Barcelona
(Spain) Results are expressed as element intensities in counts per second (cps)
Tube voltage was operated at 30 kV and 10 kV to obtain the abundances of 15
elements (Al Si S Cl K Ca Ti V Mn Fe Br Rb Sr Y Zr) with an average at
least of 1600 cps (less for Br=1000)
Biogenic silica content mineralogy and grain size were measured every 4
cm Biogenic silica was measured following Mortlock and Froelich (1989) and
Bernaacuterdez et al (2005) using an Auto Analyzer Technicon AAII for dissolved silicate
analysis Mineralogy was analyzed with a Siemens D-500 X-ray diffractometer (Cu
kα 40 kV 30 mA graphite monochromator) at the ICTJA-CSIC (Spain) Grain size
analyses were performed in a Beckmann Coulter LS 13 320 Particle Size Analyzer
57
at the IPE-CSIC The samples were classified according to textural classes as
follows clay (lt2 μm) silt (20-2 microm) and sand (gt2 microm) fractions
The age-depth model for the Laguna Matanzas sedimentary sequence was
constructed using 210Pb and 137Cs dating techniques (MAT13-4A) as well as two 14C
AMS dates on bulk sediment samples (MAT11-6A fig 2) We dated the dissolved
inorganic carbon (DIC) in the water column and no significant reservoir effect is
present in the modern-day water column (10454 + 035 pcmc Table 2) An age-
depth model was obtained with the Bacon R package to estimate the deposition
rates and associated age uncertainties along the core (Blaauw and Christen 2011)
To estimate LUCC of Laguna Matanzasacutes watershed we use satellite images
Landsat MSS for 1975 Landsat TM for 1989 and Landsat OLI for 2016 all taken in
summer or autumn (Table 1) We performed supervised classification of land uses
(maximum likelihood algorithm) for each year (1975 1989 and 2016) and the results
were mapped using software ArcGIS 102 in 2017
Satellite Images Acquisition Date
Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat OLI 20160404 30 m
Table 1 Landsat imagery
Surface water samples were filtered for obtained particulate organic matter In
addition soil samples from the main land usecover present in the Laguna Matanzas
watershed were collected Elemental C N and their corresponding isotopes from
POM and soil were obtained at the LABASI and used here as end members
Daily precipitation at Santo Domingo (33ordm39`S 71ordm3 6`W)- the nearest weather
station to Laguna Matanzasndash was compiled using the redPrec R package (Fig1 d
Serrano-Notivoli et al 2017 Sarricolea et al 2019) To extend our precipitation
58
reconstruction back to 1824 we correlated this dataset with that available for
Santiago The Santiago data was compiled from data published in the Anales of
Universidad of Chile (1850) for the years 1824 to 1849 Almeyda (1940) for the years
1849 to 1864 and the Quinta Normal series from 1866 to the present (Direccioacuten
Meteoroloacutegica de Chile) We generated a linear regression model between the
presentday Santo Domingo station and the compiled Santiago data with a Pearson
coefficient of 087 and p-valuelt 001
Acknowledgments This research was funded by grants CONICYT AFB170008
to the Institute of Ecology and Biodiversity (IEB) and FONDECYT 1150763 (to CL)
Doctoral grant Becas Chile 21150224 MEDLANT (Spanish Ministry of Economy
and Competitiveness grant CGL2016-76215-R) Additional funding was provided
by the Laboratorio Internacional de Cambio Global (LINCGlobal PUC-CSIC) We
thank R Lopez E Royo and M Gallegos for help with sample analyses We thank
the Laboratory of Biogeochemistry and Applied Stable Isotopes (LABASI) of the
Department of Ecology (PUC) for sample analyses
References
Albert F 1900 Las dunas o sean arenas volantes voladeros arenas muertas invasion de las arenas playas I meacutedanos del centro de Chile comprendiendo el litoral desde el liacutemite norte de la provincia de Aconcagua hasta el liacutemite sur de la de Arauco Memorias cientiacuteficas I Lit
Anderson NJ W D Fritz SC 2009 Holocene carbon burial by lakes in SW
Greenland Glob Chang Biol httpsdoiorg101111j1365-2486200901942x
Armesto J Villagraacuten C Donoso C 1994 La historia del bosque templado
chileno Ambient y Desarro 66 66ndash72 httpsdoiorg101007BF00385244
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A
59
historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Arnell NW Gosling SN 2013 The impacts of climate change on river flow
regimes at the global scale J Hydrol 486 351ndash364 httpsdoiorg101016JJHYDROL201302010
Bernaacuterdez P Prego R Franceacutes G Gonzaacutelez-Aacutelvarez R 2005 Opal content in
the Riacutea de Vigo and Galician continental shelf Biogenic silica in the muddy fraction as an accurate paleoproductivity proxy Cont Shelf Res httpsdoiorg101016jcsr200412009
Blaauw M Christen JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474 httpsdoiorg10121411-BA618
Brush GS 2009 Historical land use nitrogen and coastal eutrophication A
paleoecological perspective Estuaries and Coasts 32 18ndash28 httpsdoiorg101007s12237-008-9106-z
Camarero L Catalan J 2012 Atmospheric phosphorus deposition may cause
lakes to revert from phosphorus limitation back to nitrogen limitation Nat Commun httpsdoiorg101038ncomms2125
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego
R Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and environmental change from a high Andean lake Laguna del Maule central Chile (36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Catalan J Pla-Rabeacutes S Wolfe AP Smol JP Ruumlhland KM Anderson NJ
Kopaacuteček J Stuchliacutek E Schmidt R Koinig KA Camarero L Flower RJ Heiri O Kamenik C Korhola A Leavitt PR Psenner R Renberg I 2013 Global change revealed by palaeolimnological records from remote lakes A review J Paleolimnol httpsdoiorg101007s10933-013-9681-2
Chen Y Y Huang W Wang W H Juang J Y Hong J S Kato T amp
Luyssaert S (2019) Reconstructing Taiwanrsquos land cover changes between 1904 and 2015 from historical maps and satellite images Scientific Reports 9(1) 3643 httpsdoiorg101038s41598-019-40063-1
Contreras S Werne J P Araneda A Urrutia R amp Conejero C A (2018)
Organic matter geochemical signatures (TOC TN CN ratio δ 13 C and δ 15 N) of surface sediment from lakes distributed along a climatological gradient on the western side of the southern Andes Science of The Total Environment 630 878-888 httpsdoiorg101016jscitotenv201802225
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia UC
60
de V 2014 ELEMENTOS DE LA HISTORIA NATURAL DEL An Mus Hist Natulas Vaplaraiso 27 51ndash67
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-010-9453-1
Diebel M Vander Zanden MJ 201209 Nitrogen stable isotopes in streams
stable isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19 1127ndash1134 httpsdoiorg10189008-03271
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506 Espizua LE Pitte P 2009 The Little Ice Age glacier advance in the Central
Andes (35degS) Argentina Palaeogeogr Palaeoclimatol Palaeoecol 281 345ndash350 httpsdoiorg101016JPALAEO200810032
Figueroa JA Castro S A Marquet P A Jaksic FM 2004 Exotic plant
invasions to the mediterranean region of Chile causes history and impacts Invasioacuten de plantas exoacuteticas en la regioacuten mediterraacutenea de Chile causas historia e impactos Rev Chil Hist Nat 465ndash483 httpsdoiorg104067S0716-078X2004000300006
Fowler D Coyle M Skiba U Sutton MA Cape JN Reis S Sheppard
LJ Jenkins A Grizzetti B Galloway N Vitousek P Leach A Bouwman AF Butterbach-bahl K Dentener F Stevenson D Amann M Voss M 2013 The global nitrogen cycle in the twenty- first century Philisophical Trans R Soc B httpsdoiorghttpdxdoiorg101098rstb20130164
Frank D Reichstein M Bahn M Thonicke K Frank D Mahecha MD
Smith P van der Velde M Vicca S Babst F Beer C Buchmann N Canadell JG Ciais P Cramer W Ibrom A Miglietta F Poulter B Rammig A Seneviratne SI Walz A Wattenbach M Zavala MA Zscheischler J 2015 Effects of climate extremes on the terrestrial carbon cycle Concepts processes and potential future impacts Glob Chang Biol httpsdoiorg101111gcb12916
Fritz SC Anderson NJ 2013 The relative influences of climate and catchment
processes on Holocene lake development in glaciated regions J Paleolimnol httpsdoiorg101007s10933-013-9684-z
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
61
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS) implications for past sea level and environmental variability J Quat Sci 32 830ndash844 httpsdoiorg101002jqs2936
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892 httpsdoiorg101126science1136674
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924 httpsdoiorg104319lo20095430917
Garciacutea-Ruiz JM 2010 The effects of land uses on soil erosion in Spain A
review Catena httpsdoiorg101016jcatena201001001
Garciacutea-Ruiz JM Loacutepez-Moreno JI Lasanta T Vicente-Serrano SM
Gonzaacutelez-Sampeacuteriz P Valero-Garceacutes BL Sanjuaacuten Y Begueriacutea S Nadal-Romero E Lana-Renault N Goacutemez-Villar A 2015 Los efectos geoecoloacutegicos del cambio global en el Pirineo Central espantildeol una revisioacuten a distintas escalas espaciales y temporales Pirineos httpsdoiorg103989Pirineos2015170005
Garciacutea-Ruiz JM Nadal-Romero E Lana-Renault N Begueriacutea S 2013
Erosion in Mediterranean landscapes Changes and future challenges Geomorphology 198 20ndash36 httpsdoiorg101016jgeomorph201305023
Garreaud RD Vuille M Compagnucci R Marengo J 2009 Present-day
South American climate Palaeogeogr Palaeoclimatol Palaeoecol httpsdoiorg101016jpalaeo200710032
Gayo EM Latorre C Jordan TE Nester PL Estay SA Ojeda KF
Santoro CM 2012 Late Quaternary hydrological and ecological changes in the hyperarid core of the northern Atacama Desert (~21degS) Earth-Science Rev httpsdoiorg101016jearscirev201204003
Ge Y Zhang K amp Yang X (2019) A 110-year pollen record of land use and land
cover changes in an anthropogenic watershed landscape eastern China Understanding past human-environment interactions Science of The Total Environment 650 2906-2918 httpsdoiorg101016jscitotenv201810058
Goyette J Bennett EM Howarth RW Maranger R 2016 Global
Biogeochemical Cycles 1000ndash1014 httpsdoiorg1010022016GB005384Received
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and
oxygen isotope fractionation during dissimilatory nitrate reduction by
62
denitrifying bacteria 53 2533ndash2545 httpsdoiorg10230740058342
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
Gurnell AM Petts GE Hannah DM Smith BPG Edwards PJ Kollmann
J Ward J V Tockner K 2001 Effects of historical land use on sediment yield from a lacustrine watershed in central Chile Earth Surf Process Landforms 26 63ndash76 httpsdoiorg1010021096-9837(200101)261lt63AID-ESP157gt30CO2-J
Heathcote A J et al Large increases in carbon burial in northern lakes during the
Anthropocene Nat Commun 610016 doi 101038ncomms10016 (2015) Hollander DJ Smith MA 2001 Microbially mediated carbon cycling as a
control on the δ13C of sedimentary carbon in eutrophic Lake Mendota (USA) New models for interpreting isotopic excursions in the sedimentary record Geochim Cosmochim Acta httpsdoiorg101016S0016-7037(00)00506-8
Holtgrieve GW Schindler DE Hobbs WO Leavitt PR Ward EJ Bunting
L Chen G Finney BP Gregory-Eaves I Holmgren S Lisac MJ Lisi PJ Nydick K Rogers LA Saros JE Selbie DT Shapley MD Walsh PB Wolfe AP 2011 A coherent signature of anthropogenic nitrogen deposition to remote watersheds of the Northern Hemisphere Science (80) 334 1545ndash1548 httpsdoiorg101126science1212267
Hooper DU Adair EC Cardinale BJ Byrnes JEK Hungate BA Matulich
KL Gonzalez A Duffy JE Gamfeldt L Connor MI 2012 A global synthesis reveals biodiversity loss as a major driver of ecosystem change Nature httpsdoiorg101038nature11118
Horvatinčić N Sironić A Barešić J Sondi I Krajcar Bronić I Borković D
2016 Mineralogical organic and isotopic composition as palaeoenvironmental records in the lake sediments of two lakes the Plitvice Lakes Croatia Quat Int httpsdoiorg101016jquaint201701022
Howarth RW 2004 Human acceleration of the nitrogen cycle Drivers
consequences and steps toward solutions Water Sci Technol httpsdoiorg1010382Fscientificamerican0490-56
Jenny B Valero-Garceacutes BL Urrutia R Kelts K Veit H Appleby PG Geyh
M 2002 Moisture changes and fluctuations of the Westerlies in Mediterranean Central Chile during the last 2000 years The Laguna Aculeo record (33deg50primeS) Quat Int 87 3ndash18 httpsdoiorg101016S1040-
63
6182(01)00058-1
Jenny B Wilhelm D Valero-Garceacutes BL 2003 The Southern Westerlies in
Central Chile Holocene precipitation estimates based on a water balance model for Laguna Aculeo (33deg50primeS) Clim Dyn 20 269ndash280 httpsdoiorg101007s00382-002-0267-3
Leng M J Lamb A L Heaton T H Marshall J D Wolfe B B Jones M D
amp Arrowsmith C 2006 Isotopes in lake sediments In Isotopes in palaeoenvironmental research (pp 147-184) Springer Dordrecht
Le Roux JP Nielsen SN Kemnitz H Henriquez Aacute 2008 A Pliocene mega-
tsunami deposit and associated features in the Ranquil Formation southern Chile Sediment Geol 203 164ndash180 httpsdoiorg101016jsedgeo200712002
Lehmann M Bernasconi S Barbieri a McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66 3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Lenihan JM Drapek R Bachelet D Neilson RP 2003 Climate change
effects on vegetation distribution carbon and fire in California Ecol Appl httpsdoiorg101890025295
McLauchlan K K Gerhart L M Battles J J Craine J M Elmore A J
Higuera P E amp Perakis S S (2017) Centennial-scale reductions in nitrogen availability in temperate forests of the United States Scientific reports 7(1) 7856 httpsdoi101038s41598-017-08170-z
Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32ordmS 3050 masl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
Martiacuten-Foreacutes I Casado MA Castro I Ovalle C Pozo A Del Acosta-Gallo
B Saacutenchez-Jardoacuten L Miguel JM De 2012 Flora of the mediterranean basin in the chilean ESPINALES Evidence of colonisation Pastos 42 137ndash160
Marzecovaacute A Mikomaumlgi a Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54 httpsdoiorg103176eco2011105
Matesanz S Valladares F 2014 Ecological and evolutionary responses of
Mediterranean plants to global change Environ Exp Bot httpsdoiorg101016jenvexpbot201309004
64
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Menzel P Gaye B Wiesner MG Prasad S Stebich M Das BK Anoop A
Riedel N Basavaiah N 2013 Influence of bottom water anoxia on nitrogen isotopic ratios and amino acid contributions of recent sediments from small eutrophic Lonar Lake central India Limnol Oceanogr 58 1061ndash1074 httpsdoiorg104319lo20135831061
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Michelutti N Wolfe AP Cooke CA Hobbs WO Vuille M Smol JP 2015
Climate change forces new ecological states in tropical Andean lakes PLoS One httpsdoiorg101371journalpone0115338
Muntildeoz-Rojas M De la Rosa D Zavala LM Jordaacuten A Anaya-Romero M
2011 Changes in land cover and vegetation carbon stocks in Andalusia Southern Spain (1956-2007) Sci Total Environ httpsdoiorg101016jscitotenv201104009
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Poraj-Goacuterska A I Żarczyński M J Ahrens A Enters D Weisbrodt D amp
Tylmann W (2017) Impact of historical land use changes on lacustrine sedimentation recorded in varved sediments of Lake Jaczno northeastern Poland Catena 153 182-193 httpdxdoiorg101016jcatena201702007
Pongratz J Reick CH Raddatz T Claussen M 2010 Biogeophysical versus
biogeochemical climate response to historical anthropogenic land cover change Geophys Res Lett 37 1ndash5 httpsdoiorg1010292010GL043010
Prudhomme C Giuntoli I Robinson EL Clark DB Arnell NW Dankers R
Fekete BM Franssen W Gerten D Gosling SN Hagemann S Hannah DM Kim H Masaki Y Satoh Y Stacke T Wada Y Wisser D 2014 Hydrological droughts in the 21st century hotspots and uncertainties from a global multimodel ensemble experiment Proc Natl Acad Sci httpsdoiorg101073pnas1222473110
65
Ruumlhland KM Paterson AM Smol JP 2015 Lake diatom responses to
warming reviewing the evidence J Paleolimnol httpsdoiorg101007s10933-
015-9837-3
Sarricolea P Meseguer-Ruiz Oacute Serrano-Notivoli R Soto M V amp Martin-Vide
J (2019) Trends of daily precipitation concentration in Central-Southern Chile Atmospheric research 215 85-98 httpsdoiorg101016jatmosres201809005
Sernageomin 2003 Mapa geologico de chile version digital Publ Geol Digit Serrano-Notivoli R de Luis M Begueriumlia S 2017 An R package for daily
precipitation climate series reconstruction Environ Model Softw httpsdoiorg101016jenvsoft201611005
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Stine S 1994 Extreme and persistent drought in California and Patagonia during
mediaeval time Nature 369 546ndash549 httpsdoiorg101038369546a0
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL
Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Syvitski JPM Voumlroumlsmarty CJ Kettner AJ Green P 2005 Impact of humans
on the flux of terrestrial sediment to the global coastal ocean Science 308(5720) 376-380 httpsdoiorg101126science1109454
Thomas RQ Zaehle S Templer PH Goodale CL 2013 Global patterns of
nitrogen limitation Confronting two global biogeochemical models with observations Glob Chang Biol httpsdoiorg101111gcb12281
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Urbina Carrasco MX Gorigoitiacutea Abbott N Cisternas Vega M 2016 Aportes a
la historia siacutesmica de Chile el caso del gran terremoto de 1730 Anuario de Estudios Americanos httpsdoiorg103989aeamer2016211
Vargas G Ortlieb L Chapron E Valdes J Marquardt C 2005 Paleoseismic
inferences from a high-resolution marine sedimentary record in northern Chile
66
(23degS) Tectonophysics httpsdoiorg101016jtecto200412031 399(1-4) 381-398httpsdoiorg101016jtecto200412031
Valero-Garceacutes BL Latorre C Morelliacuten M Corella P Maldonado A Frugone M Moreno A 2010 Recent Climate variability and human impact from lacustrine cores in central Chile 2nd International LOTRED-South America Symposium Reconstructing Climate Variations in South America and the Antarctic Peninsula over the last 2000 years Valdivia p 76 (httpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-yearshttpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-years
Vicente-Serrano SM Gouveia C Camarero JJ Begueria S Trigo R
Lopez-Moreno JI Azorin-Molina C Pasho E Lorenzo-Lacruz J Revuelto J Moran-Tejeda E Sanchez-Lorenzo A 2013 Response of vegetation to drought time-scales across global land biomes Proc Natl Acad Sci httpsdoiorg101073pnas1207068110
Villa Martiacutenez R P 2002 Historia del clima y la vegetacioacuten de Chile Central
durante el Holoceno Una reconstruccioacuten basada en el anaacutelisis de polen sedimentos microalgas y carboacuten PhD
Villa-Martiacutenez R Villagraacuten C Jenny B 2003 Pollen evidence for late -
Holocene climatic variability at laguna de Aculeo Central Chile (lat 34o S) The Holocene 3 361ndash367
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Human alteration of the global nitrogen cycle sources and consequences Ecological applications 7(3) 737-750 httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Rein B Urrutia R amp Appleby P (2009) A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 The Holocene 19(6) 873-881 httpsdoiorg1011770959683609336573
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Widory D Kloppmann W Chery L Bonnin J Rochdi H Guinamant JL
2004 Nitrate in groundwater An isotopic multi-tracer approach J Contam Hydrol 72 165ndash188 httpsdoiorg101016jjconhyd200310010
67
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
68
Supplementary material
Facie Name Description Depositional Environment
F1 Organic-rich
mud
Massive to banded black
organic - rich (TOC up to 14 )
mud with aragonite in dm - thick
layers Slightly banded intervals
contain less OM (TOClt4) and
aragonite than massive
intervals High MnFe (oxic
bottom conditions) High CaTi
BrTi and BioSi (up to 5)
Distal low energy environment
high productivity well oxygenated
and brackish waters and relative
low lake level
F2 Massive to
banded silty clay
to fine silt
cm-thick layers mostly
composed by silicates
(plagioclase quartz cristobalite
up to 65 TOC mean=23)
Some layers have relatively high
pyrite content (up to 25) No
carbonates CaTi BrTi and
BioSi (mean=48) are lower
than F1 higher ZrTi (coarser
grain size)
Deposition during periods of
higher sediment input from the
watershed
69
F3 Banded to
laminated light
brown silty clay
cm-thick layers mostly
composed of clay minerals
quartz and plagioclase (up to
42) low organic matter
(TOC mean=13) low pyrite
and BioSi content
(mean=46) and some
aragonite
Flooding events reworking
coastal deposits
F4 Laminated
coarse silts
Thin massive layers (lt2mm)
dominated by silicates Low
TOC (mean=214 ) BrTi
(mean=002) MnFe (lt02)
TIC (lt034) BioSi
(mean=46) and TS values
(lt064) and high ZrTi
Rapid flooding events
transporting material mostly
from within the watershed
F5 Breccia with
coarse silt
matrix
A 17 cm thick (80-97 cm
depth) layer composed by
irregular mm to cm-long ldquosoft-
clastsrdquo of silty sediment
fragments in a coarse silt
matrix Low CaTi BrTi and
MnFe ratios and BioSi
Rapid high energy flood
events
70
(mean=43) and high ZrTi
(gt018)
Table Sedimentological and compositional characteristics of Laguna Matanzas
facies
71
CAPIacuteTULO 2 STABLE ISOTOPES TRACK LAND USE AND COVER
CHANGES IN A MEDITERRANEAN LAKE IN CENTRAL CHILE OVER THE
LAST 600 YEARS
72
Stable isotopes track land use and cover changes in a mediterranean lake in
central Chile over the last 600 years
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugone-Aacutelvarezabc Pablo
Sarricolead Leonardo Villaciacutese M Laura Carrevedob Blas Valero-Garceacutescg
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Departamento de Ecologiacutea Universidad de ChileLas Palmeras 3425 Nuntildeoa Santiago Chile
f Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding authors E-mail addresses clatorrebiopuccl magdalenafuentealbagmailcom
Key words Anthropocene lake sediment δsup1⁵N δsup1sup3C Great Acceleration organic
geochemistry watershedndashlake system Stable Isotope Analyses land usecover
change Nitrogen cycle mediterranean ecosystems central Chile
73
Abstract
Nutrient fluxes in many aquatic ecosystems are currently being overridden by
anthropic controls especially since the industrial revolution (mid-1800s) and the
Great Acceleration (mid-1900s) Land use and cover changes (LUCC) alter the
availability and fluxes of nutrients such as nitrogen that are transferred via runoff
and groundwater into lakes By altering lake productivity and trophic status these
changes are often preserved in the sedimentary record Here we use
geolimnological proxies and stable isotope analyses (δsup1⁵N δsup13C) on lake sediments
to reconstruct the lake-watershed nutrient transfer in a central Chilean lake (Lago
Vichuqueacuten) over the past 600 years We compare our multiproxy record from recent
lake sediments to the soilvegetation relationship across the watershed as well as
land usecover changes from 1975 to 2014 derived from satellite images Our results
show that lake sediment δsup1⁵N values increased with meadow cover but decreased
with tree plantations suggesting increased nitrogen retention when trees dominate
the watershed Our δsup1⁵N record from central Chile can thus be interpreted as a proxy
for nutrient availability over the last 600 years mainly derived from land use changes
coupled with climate drivers Although variable sources of organic matter and in situ
fractionation often hinder straightforward environmental interpretations of stable N
isotope signatures our study shows that lake sediment δsup1⁵N can be a useful tool for
assessing the contribution of past human activities in nutrient and nitrogen cycling
1 Introduction
Nitrogen (N) is a limiting nutrient in terrestrial and aquatic ecosystems (Vitousek
et al 1997) Changes in its availability can drive eutrophication and increase
pollution in these ecosystems (McLauchlan et al 2013) Although recent human
74
impacts on the global N cycle have been significant the consequences of increased
anthropogenic N on these ecosystems are unclear (Gerhart and Mclauchlan 2014
Battye et al 2017) Isotopic signatures are key to understand N dynamics in lakes
nevertheless in situ andor diagenetic fractionation along with multiple sources of
organic matter (OM) often hinder straightforward environmental interpretations from
isotopes Monitoring δ15N and δ13C values as components of the N cycle
specifically those related to the link between terrestrial and aquatic ecosystems can
help differentiate between effects from processes versus sources in stable isotope
values (eg from Particulate Organic Matter -POM- soil and vegetation) and
improve how we interpret variations in δ15N (and δ13C) values at longer temporal
scales
The main processes controlling stable N isotopes in bulk lake OM are source
lake productivity diagenesis and denitrification (Talbot 2001 Leng et al 2006
Fariacuteas et al 2009 Torres et al 2012 Brahney et al 2014) N sources depend on
contributions from the watershed (ie soil and biomass) the transfer of atmospheric
N to the lake (ie atmospheric N2 fixation) and nutrient recycling (Leng et al 2006)
Atmospheric N2 (isotopically defined as δ15N =0) is fixed by cyanobacteria with
minimal fractionation resulting in δ15NOM values that fluctuate from -2 to +2permil (Fogel
and Cifuentes 1993 Talbot 2001 Elliot and Brush 2006) Besides N2 fixation by
cyanobacteria phytoplankton species assimilate dissolved inorganic nitrogen (DIN)
and values typically range from +7 to +10permil (Xu et al 2016 Meyers et al 2003) In
addition seasonal changes in POM occur in the lake water column Gu et al (2006)
sampled the water column of Lake Wauberg (Florida USA) during a hydrologic year
and found a higher development of N fixing species during the summer A major
factor behind this increase are human activities in the watershed which control the
75
inputs of the sediment and nutrients to the lake (Woodward et al 2011) Some
studies have shown higher δ15N values in lake sediments from watersheds that are
highly affected by human activities (Bruland and Mackenzie 2010 Woodward et al
2011) Syntenic fertilizers displayed δ15N values between -4 to +4permil cattle manure
around +8 to +22permil and surface and groundwater OM between 0 to +10permil (Elliott
and Brush 2006 Leng et al 2006) Although relatively low δ15N values from
fertilizers constitute major N input to human-altered watersheds the elevated loss
of 14N via volatilization of ammonia and denitrification leaves the remaining total N
input enriched in 15N (Bruland and Mackenzie 2010)
In addition to the different sources and variations in lake productivity early
diagenesis at the sedimentndashwater interface in the sediment can further alter
sediment OM isotope values (Brahney et al 2014 Galman et al 2009) During
diagenetic loss of N in lacustrine systems kinetic fractionation occurs and the
remaining N is enriched in 15N leading to higher δ15N values (Galman et al 2006
Talbot 2001) Denitrification tends to enrich lake sediments in 15N due to the
assimilation of nitrates by denitrifying bacteria (Granger et al 2008) and is more
prevalent in anoxic environments (Brahney et al 2016 Lehman et al 2002)
Carbon isotopes in lake sediments can also provide useful information about
paleoenvironmental changes OM origin and depositional processes (Meyers et al
2003) Allochthonous organic sources (high CN ratios) produce isotope values
similar to values from catchment vegetation Autochthonous organic matter (low CN
ratio) is influenced by fractionation both in the lake and the watershed leading up to
carbon assimilation by lacustrine biota (Woodward et al 2011) Changes in
productivity greatly affect δ13COM values (Torres et al 2012) as during C uptake
plankton preferentially assimilate 12C leaving the dissolved inorganic carbon (DIC)
76
pool enriched in 13C (Gu et al 2006) with phytoplankton averaging ca 20 permil lower
than the respective δ13CDIC values (Leng et al 2006) Under conditions of low to
moderate primary productivity plankton preferentially uptake the lighter 12C
resulting in low δ13C values displayed by the OM (Torres et al 2012) Conversely
during high primary productivity phytoplankton will uptake 12C until its depletion and
is then forced to assimilate the heavier isotope resulting in an increase in δ13C
values Higher productivity in C-limited lakes due to slow water-atmosphere
exchange of CO2 also results in high δ13C values (Galman et al 2009) In these
cases algae are forced to uptake dissolved bicarbonate with δ13C values between
7 to 10permil higher than those of the atmospheric CO2 dissolved in water (Sun et al
2016 Torres et al 2012 Galman et al 2009)
Stable isotope analyses from lake sediments are thus useful tools to
reconstruct shifts in lake-watershed dynamics caused by changes in limnological
parameters and LUCC Our knowledge of the current processes that can affect
stable isotope signals in a watershed-lake system is limited however as monitoring
studies are scarce Besides in order to use stable isotope signatures to reconstruct
past environmental changes we require a multiproxy approach to understand the
role of the different variables in controlling these values Hence in this study we
carried out a detailed survey of current N dynamics in a coastal central Chilean lake
(Lago Vichuqueacuten) and a multiproxy study of a sediment record spanning the last
600 years The characterization of the recent changes in the watershed since 1970s
is based on satellite images to compare recent changes in the lake and assess how
these are related with climate variability and an ever increasing human footprint
(Lara et al 2012 Gallardo et al 2015 Steffen et al 2015) Our main goal is to
investigate how stable isotope values from lake sediment reflect changes in the lake
77
ndash watershed system during periods of high watershed disruption (eg Spanish
Conquest late XIX century Great Acceleration) and recent climate change (eg
Little Ice Age and current global warming)
2 Study Site
Lago Vichuqueacuten (34ordm49`S 72ordm04W 4 m asl 30 m of depth Fig 1) is a
mesotrophic to eutrophic lake with dominant anoxic bottom conditions that is
stratified year around (Frugone-Aacutelvarez et al 2017) The main inlets are the
Vichuqueacuten and Huintildee rivers and the only outlet is the Llico estuary that drains into
the Pacific Ocean High tides can sporadically shift the flow direction of the Llico
estuary which increases the marine influence in the lake Dune accretion gradually
limited ocean-lake connectivity until the estuary was almost completely closed off
by ca 10 ky BP and the current lake regime formed (Frugone-Aacutelvarez et al 2017)
The area is characterized by a mediterranean climate with cold-wet winters and
hot-dry summers and an annual precipitation of ~650 mm and a mean annual
temperature of 15ordmC During the austral winter months (June - August) precipitation
is modulated by north-west shifts of the South Pacific Anticyclone (SPA) followed by
an increased frequency of storm fronts stemming off the South Westerly Winds
(SWW) A strengthened SPA during austral summers (December - March) which
are typically dry and warm blocks the northward migration of storm tracks stemming
off the SWW (Fig1 Frugone-Aacutelvarez et al 2017)
78
Figure 1 a) The location of the Lago Vichuqueacuten in central Chile b) Map of land
uses and cover in 2016 c) Ombrothermic diagram mediterranean climates are
characterized by cold-wet winters with surplus moisture from June to August and
hot-dry summers d) Lake bathymetry showing location of cores and water sampling
sites used in this study
Although major land cover changes in the area have occurred since 1975 to the
present as the native forests were replaced by tree (Monterey pine and eucalyptus)
plantations the region was settled before the Spanish conquest (Frugone-Alvarez
et al 2018) Historical documents indicate that Vichuqueacuten was occupied by a
Mitimaes colony as part of the Inka empire (Odone 1998) Unlike other Andean
areas during the Inka period the introduction of corn and quinoa in the Vichuqueacuten
watershed do not seem to have intensified land use The Spanish colonial period in
Chile lasted from 1542 CE to the independence in 1810 CE The first historical
document (1550 CE) shows that the areas around Vichuqueacuten were settled by the
Spanish early on as the Vichuqueacuten watershed was given as an ldquoencomiendardquo
system to Juan Cuevas The ldquoencomiendardquo system provided the owners with land
79
and indigenous people to work but also the introduction of wheat wine cattle
grazing and logging of native forests for lumber extraction and increasing land for
agricultural activities (Odone et al 1998 Armesto et al 2010) During the 19th
century (the Republic) the export of wheat to Australia and Canada generated
intensive changes in land cover use The town of Vichuqueacuten became the regional
capital and plans to build a maritime port in the bay of Vichuqueacuten were drawn
However the fall of international markets in 1880 paralyzed these plans During the
20th century the plantation of trees (Pinus radiata and Eucalyptus globulus) in areas
cleared of native forests was subsidized by two national laws DFL nordm265 (1931) and
DFL nordm 701 (1974) both of which provided funds for such plantations During the last
decades the urbanization with summer vacation homes along the shorelines of
Lago Vichuqueacuten has greatly increased Extensive lake eutrophication is currently a
large environmental problem (EULA 2008)
3 Methods
Coring campaigns were organized from 2011 to 2015 (Fig 1d) We recovered
12 short cores and 4 long cores (up to 14 m long) at several sites using a hammer-
modified UWITEC gravity corer Sediment cores (VIC11-2A 170 cm VIC13-2B 170
cm VIC13-2D 1200 cm and VIC15-1A 119 cm) were split photographed sub-
sampled and stored in the Pyrenean Institute of Ecology (IPE-CSIC Spain) Core
VIC13-2B was selected for detailed multiproxy analyses (including elemental
geochemistry C and N isotopes XRF and 14C dating) Stable isotope analyses
(δ13C δ15N) were performed at the Laboratory of Biogeochemistry and Applied
Stable Isotopes (LABASI-PUC Chile) Sediment samples for δ13Corg were pre-
treated with 50 ml of HCl and 50 ml of deionized water and dried at 60ordmC for 4 hr to
remove carbonates (Harris et al 2011) Isotope analyses were conducted using a
80
Delta V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via
a Conflo IV interface Isotope results are expressed in standard delta notation (δ)
and expressed in per mil (permil) relative to the Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Total carbon (TC)
Total Organic Carbon (TOC) Total Inorganic Carbon (TIC) and Total Sulfur (TS)
were measured every 2 cm with a LECO SC 144 DR at the IPE-CSIC
An AVAATECH X-Ray Fluorescence II core scanner with a Rh X-ray tube from
the University of Barcelona was used to obtain XRF logs every 4 mm of resolution
Results are expressed as element intensities in counts per second (cps) Tube
voltage was operated at 30 kV and 10 kV to obtain the abundances of 18 elements
(Pb Al Si S Cl K Ca Ti V Mn Fe Co Zn Br Rb Sr Y Zr) with an average of
at least of 1800 cps (less for Pb=437 and Zn=934 cps) Pb was included due to
similar behavior with Co and Fe Element ratios were calculated to describe changes
in redox conditions (MnFe) endogenic productivity (BrTi) carbonate formation
(SrCa and CaTi) and sediment input (ZrTi) (Barreiro-Lostres et al 2014
Carrevedo et al 2015 Frugone-Aacutelvarez et al 2017 Kylander et al 2011 Moreno
et al 2007a)
Several campaigns were carried out to sample the POM from the water column
two per hydrologic year from November 2015 to August 2018 A liter of water was
recovered in three sites through to the lake two are from the shallower areas (with
samples taken at 2 and 5 m depth at each site) and one in the deeper central portion
(at 2 5 and 20 m depth) of the lake (Fig 1d) The samples were filtered using glass
fiber filters (25 m) and then frozen to obtain the stable carbon and nitrogen isotope
signal of lacustrine POM Additionally soil and vegetation samples from the
following communities native species meadow hydrophytic vegetation and
81
Monterrey pine (Pinus radiata) were taken for stable isotope analyses (see in
supplementary material)
The age model for the complete Lago Vichuqueacuten sedimentary sequence is
based on Frugone-Aacutelvarez et al (2017) with the last 1000 years based on
210Pb137Cs dating techniques in VIC15-1A and 14C AMS dates on bulk sediment
samples (Supplementary Table S1) The 14C measurements of lake water DIC show
a moderate reservoir effect of 180 plusmn 25 14C yr) The revised age-depth model used
here includes three more 14C AMS dates performed with the program Bacon to
establish the deposition rates along the core (Blaauw and Christen 2011 Fig 2)
The age-depth model indicates that average resolution between 0 to 87 cm is lt2
cm per year and from 88 to 170 cm it is lt47 cm per year
82
Figure 2 Chronological age-depth model for the Lago Vichuqueacuten sedimentary
sequence spanning the last 1000 yr (see also Frugone-Alvarez et al 2017)
To estimate land use changes in the watershed we use Landsat MSS images
for 1975 and Landsat TM for 1989 and 2014 all taken in either summer or autumn
(Table 1) We performed supervised classification of land uses (maximum likelihood
83
algorithm) for each year (1975 1989 and 2014) and results were mapped using
ArcGIS 102
Table 1 Images using for LUCC reconstruction
Source of LUCC
Acquisition
Date Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat TM 19991226 30 m
CONAF 2009 30 m
Land cover Chile 2014 30 m
CONAF 2016 30 m
Previous Work on Lago Vichuqueacuten sedimentary sequence
The sediments are organic-poor dark brown to brown laminated silt with some
intercalated thin coarser clastic layers Lacustrine facies have been classified
according to elemental composition (TOC TS TIC and TN) grain size and
sedimentary textures (Frugone-Aacutelvarez et al 2017) Four main types of lacustrine
facies were identified in this short core Facies L1 is a laminated (1cm) black to dark
brown diatom-poor silt with relatively higher TOC percentages (TOC about 185)
TS (about of 18) and TOCTN (about 92) Facies L2 is characterized by a
homogeneous black organic-poor silty clay with low TOC (mean= 13) TS (mean=
13) TOCTN (mean= 88) values Facies L3 is a laminated (1cm) black organic-
poor (TOC mean 14) silts with higher TS (mean= 21) and TOCTN ratios
(mean= 88) Facies L3 is interpreted as ldquobaselinerdquo deposition on the distal areas
84
of Vichuqueacuten lake Mineralogy is dominated by mica (muscovite) clay minerals
(chlorite) and plagioclase (albite) Quartz and calcite occur at the top and pyrite
occurs in the lower part of the sequence Facies T is composed by massive banded
sediments 4 cm thick and coarse grain size It is interpreted as an instantaneous
depositional event (for more detail see Frugone-Aacutelvarez et al 2017) In this work
we identified four subunits based on geochemical and stable isotope signals
4 Results
41 Geochemistry and PCA analysis
High positive correlations exist between Al Si K and Ti (r = 078 ndash 096
supplementary Fig S1) and between Zn Zr Rb and Y (r = 072 - 096) which reflect
the silicate content of the watershed-derived sediments (Marzecoacuteva et al 2011) Zr
is commonly associated with minerals more abundant in coarser deposits Thus the
ZrTi profile could reflect grain size (Cuven et al 2010) showing higher variability
in the upper part of the Lake Vichuqueacuten sequence and in the alternation between
facies L1 and L2 Another group of elements made up of Co Fe and Pb displayed
positive correlations (r = 067ndash 097) and represents the input of heavy metals Br
Cl and Ca show a positive correlation (r = 049 ndash 06 p-value = 00001) BrTi ratio
is interpreted as a productivity indicator due to Br having a strong affinity with humic
and organic compounds (Marzecoacuteva et al 2011 Frugone-Aacutelvarez et al 2017) In
our Lago Vichuqueacuten sequence BrTi values are high from 170 to 132 cm and from
36 cm to the top of core where it shows several sharps peaks (Fig 3) The MnFe
ratio is indicative of lake bottom oxygenation (Naecher et al 2013) as under
reducing conditions Mn tends to become more mobile than Fe leading to a
decreased MnFe (Marzecoacuteva et al 2011) Several sharp peaks in MnFe occurred
85
from 127 to 120 cm and from 57 to 30 cm (MgFegt03) The silicate group and the
Br Cl Ca Mn group are negatively correlated (r= -012 and -066)
Principal Component Analysis (PCA) was undertaken on the XRF
geochemical data to investigate the main factors controlling sediment deposition in
Lago Vichuqueacuten (Fig 3) The four main eigenvectors explain 801 of the variance
(supplementary material Table S2) The principal component (PC1) explain 437
of the total of variance and grouped elements are associated with terrigenous input
to the lake Positive values of the biplot have been attributed to higher heavy metals
deposition (Pb Co and Fe) and negative values to higher silicate input (Al K Ti and
Si) in a high energy catchment environment (Zr) The PC2 explains 198 of the
total of variance and highlights the endogenic productivity in the lake The positive
loading is compound by Br Cl Ca and Sr and to a lesser extent by Y Zn Rb and
Mn The PC2 may explain both the precipitation of carbonates (Ca Sr) and biological
production (Br)
86
Figure 3 Principal Component Analysis of XRF geochemical measurements in
VIC13-2B Lago Vichuqueacuten lake sediments
42 Sedimentary units
Based on geochemical and stable isotope analysis we identified four
lithological subunits in the short core sedimentary sequence Our PCA analyses and
Pearson correlations pointed out which variables were better for characterizing the
subunits (Fig 4) in terms of endogenic productivity heavy metals and terrestrial
input The unit 2c (170-131 cm) is characterized by the occurrence of some clastic
layers (L2 from 160 and 150 cm) high δsup1⁵N values (mean= +59 plusmn 04permil) with
Magnetic Susceptibility (MS) BrTi ZrTi and SrCa values increase towards the top
Also it has relatively low δsup1sup3C values (mean= -265 plusmn 11permil) and CNmolar ratios
(mean=78 plusmn 04) The ZrTi show upwards increasing values reaching the highest
values of the sequence at the top of this unit suggesting a coarsening upward trend
and relatively higher depositional energy The MS trend also indicates higher
erosion in the watershed and enhanced delivery of ferromagnetic minerals likely
from soil particles particularly from 150 to 160 cm (Frugone-Aacutelvarez at al 2017)
The subunit 2b (130-118 cm) is also composed of black silts but it has the
lowest MS values of the whole sequence and its onset is marked by a sharp
decrease in ZrTi This subunit is marked by sharp peaks of MnFe (at 126 and 120
cmgt027) SrCa (at 125 and 120 cmgt033) CaTi (at 124 and 120 cmgt199) TOC
(about 26 at 130 124 122 and 118 cm) and δsup1⁵N (gt+67permil at 126 and 120 cm)
BrTi oscillates around low values (about 01) whereas the δsup1sup3C values range
between -262 and -282permil
87
The unit 2a (58-117 cm) shows increasing and then decreasing MS values
and displayed sharps peaks of δsup1⁵N (gt64 at 102 to 99 87 and 75 cm) and CN
(about 10 at 86 74 66 and 60 cm) SrCa (mean = 04 plusmn 013) BrTi (mean = 008
plusmn 002) MnFe (mean = 002 plusmn 001) and δsup1sup3C (mean = -260 plusmn 07permil) oscillate in
low values The upper part of this unit (72 ndash 57 cm) shows an increase of SrCa
(from 03 to 05)
The upper unit 1a (0 - 57 cm) starts with a fine clastic layer (T at 57 to 54
cm) interpreted as deposition during a high-energy event It is characterized by
lower BrTi (mean = 003 plusmn 002) TOC (mean = 12 plusmn 03) and δsup1sup3C (mean= -
266 plusmn 06permil) higher δsup1⁵N (mean = +55 plusmn 08 permil) This unit is made of alternating
fine (lt 1cm) black and dark brown laminae with carbonate (L1 facies) Recently
deposited sediments show decreasing MS values increasing TOC (mean= 15 plusmn
04 ) sharp peaks of BrTi (gt015 at 33 21 5 and 1 cm) and the lowest δsup1⁵N values
of the entire record (mean= 45 plusmn 06 permil) and δsup1sup3C values (mean= -277 plusmn 09 permil)
Heavy metal content increases abruptly in the upper section of the core (2 - 20 cm)
(peaks of FeTi CoTi and PbTi)
88
Figure 4 Sedimentary units of Lago Vichuqueacuten sedimentary sequence Selected
variables illustrate watershed input (Magnetic Susceptibility ZrTi CoTi and MnFe)
endogenic productivity (SrCa CaTi TS) organic matter source (BrTi TOC
CNmolar and stable isotope records (δ13Corg and δ15Nbulk)
43 Recent seasonal changes of particulate organic matter on water column
The average of δ15NPOM from the three sites across to the lake was +86 plusmn 58
permil oscillating widely from -14 to +193permil (Fig 5) Important seasonal differences
occur at 2 and 5 m depth with more negative values during summer (~53 plusmn 52 permil)
than winter (~122 plusmn 40permil) At 20 m depth δ15NPOM values exhibit small seasonal
ranges (mean = +10permil for winter and +87permil for summer) The δ13CPOM average was
-296 plusmn 33permil with slightly seasonal and water column depth differences However
more positive δ13CPOM values occurred during winter (mean=-290 plusmn 41permil) than in
summer (mean= -301 plusmn 19 permil) Like the δ15NPOM values CNPOM (mean= 83 plusmn 17)
displayed important seasonal and water depth differences Lower CNPOM ratios
89
occurred during summer at 2 and 5 m depth (mean= 95 plusmn 17) and higher and more
constant values during winter (mean = 73 plusmn 09) at the same depth At 20 m CNPOM
shows similar values in both winter (70) and summer (74)
Figure 5 Seasonal POM averages from samples taken of the Lago Vichuqueacuten
water column Samples were taken from 2015 to 2018 at 2 (n=16) 5 (n=14) and 20
(n=8) meters depth
44 Stable isotope values across the Lake Vichuqueacuten watershed
Figure 6 shows modern vegetation soil and sediment isotope values found for
the watershed Huintildee tributary and Vichuqueacuten lake The δsup1⁵NFoliar values from
meadow plantations and macrophytes have similar range values with a mean of
+89 plusmn50 permil (n=18) +74 plusmn 75 permil (n=5) +82 plusmn 36 permil (n=6) respectively Native
vegetation has overall lower δsup1⁵NFoliar values with a mean of 14 plusmn 21 permil (n=28 see
Table 2 in supplementary material) The δsup1sup3CFoliar from terrestrial samples exhibit
similar values across the different plant communities (tree plantation mean=-274 plusmn
13permil meadow mean = -275 plusmn 23 permil native species mean=-272 plusmn 14permil) whereas
macrophytes display slightly more negative values with a mean of -287 plusmn 23permil
Macrophytes substrate displayed the highest δsup1⁵N values with a mean of +60 plusmn
14permil (n=3) Similar and relatively high δsup1⁵N values for soil of plantation (mean= +54
plusmn 13permil n=4) meadow (mean= +49 plusmn 04permil n=8) and surface river sediment
90
(mean= +52 plusmn 17permil n=10) Native forest soils oscillate around slightly more
negative values with a mean of +45 plusmn 12permil (n=4) Slightly more positive soil δsup1sup3C
values occur both underneath native forests and in tree plantations with means of -
284 plusmn 23permil and -298 plusmn 13permil respectively compared to either meadow soils
(mean= -302 plusmn 11permil) aquatic sediments underneath macrophytes (-311 plusmn 10permil)
or from surface river sediments (mean= -312 plusmn 10permil)
Figure 6 Stable isotope content of modern soil river sediment and foliar vegetation
used as end members in the sedimentary sequence of Lago Vichuqueacuten a)
Scatterplot of foliar δsup1⁵N and δsup13C values for vegetation from Lago Vichuqueacuten
watershed (plantation meadow and native species) and macrophytes on Lake
Vichuqueacuten See supplementary material for more detail of vegetation types b)
Scatterplot of soil δsup1⁵N and δsup13C values across the watershed the sediments of the
Huintildee and Vichuqueacuten rivers and the fluvial sediment corresponding to the
macrophyte vegetation
45 Land use and cover change from 1975 to 2014
Major land use changes between 1975 CE and 2016 CE in the Lago
Vichuqueacuten watershed are summarized in Figure 7 The watershed has a surface
area of 535329 km2 of which native vegetation (26) and shrublands (53)
represent 79 of the total surface in 1975 Meadows are confined to the valley and
91
represent 17 of watershed surface Tree plantations initially occupied 1 of the
watershed and were first located along the lake periphery By 1989 the areas of
native forests shrublands and meadows had decreased to 22 31 and 14
respectively whereas tree plantations had expanded to 30 These trends
continued almost invariably until 2016 when shrublands and meadows reached 17
and 5 of the total areas while tree plantations increased to 66 Native forests
had practically disappeared by 1989 and then increased up to 7 of the total area
in 2016 (Fig 6) preferentially occupying the southeastern sector of the watershed
Figure 7 Changes in land use and cover between 1975 and 2016 in the Lago
Vichuqueacuten watershed as measured from satellite images The major change is
represented by the replacement of native forest shrubland and meadows by
plantations of Monterrey pine (Pinus radiata)
Figure 8 shows correlations between lake sediment stable isotope values and
changes in the soil cover from 1975 to 2013 Positive relationships occurred
between sediment δsup1⁵N and δsup13C values from core VIC13-2B (unit 1) and the
92
percentage of native forest (r=089 for δsup1⁵N 082 for δsup13C) shrublands (r = 071 for
δsup1⁵N 057 for δsup13C) and meadow cover (r = 088 for δsup1⁵N 081 for δsup13C) All of these
correlations are significant (p value lt 0001) In contrast significant negative
correlations (p lt0001) occurred between tree plantation cover and lake sediment
stable isotope values (δsup1⁵N r = -082 and δsup13C r = -067) native forest (r = -097)
meadows (r = -086) and shrubland (r =-093)
Figure 8 Correlation plots of land use and cover change versus lake sediment
stable isotope values The δsup1⁵N values are positively correlated with native forests
agricultural fields and meadow cover across the watershed Total Plantation area
increases are negatively correlated with native forest meadow and shrubland total
area Significance levels are indicated by the symbols p-values (0 0001 001
005 01 1) lt=gt symbols ( )
93
5 Discussion
51 Seasonal variability of POM in the water column
The stable isotope values of POM can vary during the annual cycle due to
climate and biologic controls namely temperature and length of the photoperiod
which affect phytoplankton growth rates and isotope fractionation in the water
column (Gu et al 2006 Leng et al 2006) POM from Lago Vichuqueacuten surface
samples (2 and 5 m depth) displayed lower δ13CPOM values during the summer than
in winter During C uptake phytoplankton preferentially utilize 12C leaving the
DICpool enriched in 13C Therefore as temperature increases during the summer
phytoplankton growth generates OM enriched in 12C until this becomes depleted
and then the biomas come to enriched u At the onset of winter the DICpool is now
enriched in 13C and despite an overall decrease in phytoplankton production the
OM produced will also be enriched in 13C In contrast δ13CPOM values at 20 m depth
did not reflect these seasonal differences probably due to water-column
stratification that maintains similar temperatures and biological activity throughout
the year
Lake N availability depends on N sources including inputs from the
watershed and the atmosphere (ie deposition of N compounds and fixation of
atmospheric N2) which varies during the hydrologic year The fixation of atmospheric
N2 is an important natural source of N to the lake occurring mainly during the
summer season associated with higher temperature and light (Gu et al 2006)
Because the δ15NPOM of N2 fixers is close to 0permil with almost no overall isotope
fractionation (Leng et al 2006) when N2 is the major N source δ15NPOM values are
typically low However when DIN concentrations are high or alternatively when little
94
N2 fixation occurs δ15NPOM values will be higher (Gu et al 2006) δ15NPOM values
from summer Lago Vichuqueacuten samples were lower than those from winter with large
differences between the seasons (~5permil Fig 5 and 9) Furthermore δ15NPOM values
were high when monthly average temperature was low and monthly precipitation
was high (Fig 9) This pattern is consistent with the dominance of high N2 fixation
by cyanobacteria associated with increased summer temperatures This correlation
of δ15NPOM values with temperature further suggests a functional group shift i e
from N fixers to phytoplankton that uptake DIN The correlation between wetter
months and higher δ15NPOM values could be caused by increased N input from the
watershed due to increased runoff during the winter season The lack of data of the
δ15NPOM between the 30 mm and the 160 mm of precipitation is due to the
mediterranean-type climate that concentrates precipitations in the winter months
Finally Lago Vichuqueacuten is characterized by pronounced seasonality leading to
higher phytoplankton biomass in summer characterized by low δ15NPOM In winter
low biomass production and increased input from watershed is associated to high
δ15NPOM
95
Figure 9 Seasonal changes can drive δ15NPOM values in Lago Vichuqueacuten Data
correspond to average monthly temperature and total monthly precipitation for the
months when the water samples were taken (years 2015 - 2018) P-valuelt005
52 Stable isotope signatures in the Lake Vichuqueacuten watershed
The natural abundance of 15N14N isotopes of soil and vegetation samples
from the Lago Vichuqueacuten watershed appear to result from a combination of factors
isotope fractionation different N sources for plants and soil microorganisms (eg N2
fixation Nitrates) depth in the soil where N uptake by vegetation occurs and N loss
mechanisms (ie denitrification leaching and ammonia volatilization Hogberg
1997) The lowest δsup1⁵Nfoliar values are associated with native species and are
probably due to the contribution of N-fixing species (eg Sophora) (Fig6 and for
more detail see Table S3 in supplementary material) The number native N-fixers
species present in the Chilean mediterranean vegetation are not well known
however so there could be other N-fixing species in this group Alternatively δsup1⁵Nfoliar
values reflect soil N uptake (Kahmen et al 2008) In environments limited by N
plants have δsup1⁵N values similar to the soil δsup1⁵N values however the denitrification
and volatilization of ammonia can lead to the remain N of soil to come enriched in
15N (Cifuentes et al 1989 Craine et al 2009 Bruland and Mckenzie 2010) Soil N
isotope samples from native species communities tends to display relatively high
δsup1⁵N values respect to foliar samples due to loss of N-soil
The higher foliar and soil δsup1⁵N values obtained from samples of meadows
aquatic macrophytes and tree plantations can be attributed to the presence of
greater amounts of nitrate available for plant uptake (Fig 6) Houmlgberg (1997)
suggests that the availability of different N sources in soils (ie nitrates versus
96
ammonia) with different residence times can also explain these δsup1⁵NFoliar values
Indeed Feigin et al (1974) described differences of up to 20permil between ammonia
and nitrates sources Denitrification and nitrification discriminate much more against
15N than N2 fixation does (Craine et al 2009) leading to soils (and plants after
uptake) enriched in 14N
In general multiple processes that affect the isotopic signal result in similar
δsup1⁵N values between the soil of the watershed and the sediments of the river
However POM isotope fluctuations allow to say that more negative δsup1⁵N values are
associated to lake productivity while more positive δsup1⁵N values are associated with
N input from the watershed
δ13C values in Lago Vichuqueacuten watershed reflect CO2 gas exchange between
C3 plants and algae with the atmosphere During photosynthesis plants
discriminate against the heavier stable isotope of carbon (13C) in favor of the lighter
isotope 12C which results in δ13CFoliar values between -220permil and -320permil (Browman
and Cook 2002 Liu et al 2007) Likewise the δ13CFoliar values from Lago Vichuqueacuten
oscillate around -27permil (Fig 6) Plants transform atmospheric CO2 into soil organic
carbon (C) which in turn reflects this initial discrimination against 13C during C
uptake and post photosynthetic fractionation (Farquhar et al 1982 1989 Badeck
et al 2005 Pausch and Kuzyakov 2017) Slightly more positive δ13CFoliar values
(about 15permil) were measured in comparison with their δ13CSoil values This may be
reflecting the C transference from plants to the soil but also a soil-atmosphere
interchange The preferential assimilation of the light isotopes (12C) during soil
respiration carried by the roots and the microbial biomass that is associated with the
decomposition of litter roots and soil organic matter explain this differential
(Berhardt et al 2006 Blagodatskaya et al 2010 Midwood and Millard 2011)
97
In general the δ13CSoil values from the Lago Vichuqueacuten watershed oscillated
around -290permil and did not vary with our plant classification types Here we use
these values as terrestrial-end members to track changes in source OM (Fig 6)
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from the terrestrial watershed By the other hand more positive δ13C
values most likely reflect an increased aquatic OM component as indicated by POM
isotope fluctuations (Fig 9)
53 Recently land use and cover change and its influences on N inputs to the lake
Tree plantations in the Lago Vichuqueacuten watershed have increased greatly in
the last few decades (from 1 in 1975 to 66 in 2016 Fig 7) replacing previous
native forests (from 26 to 7) meadows (17 to 5) and shrublands (53 to
17) In 1975 tree plantations were confined to the lake perimeter with discrete
patches distributed throughout the watershed The ldquoForest Lawrdquo (DFL 701- passed
in 1974) allocated state funding to afforestation efforts and management of tree
plantations which greatly favored the replacement native forests by introduced trees
This increase is marked by a sharp and steady decrease in lake sediment δ15N and
δ13C values because tree plantations function as a nutrient sink whereas other land
uses such as meadows favor nutrient loss from the watershed (Fig8) Bruland and
Mackenzie (2014) noted a decrease in wetland δ15N values when watershed
forested cover increased and concluded that N inputs to the wetlands are lower from
the forested areas as they generally do not export as much N as agricultural lands
A positive correlation between native vegetation and δ15Ncore values can be
explained by the relatively scarce arboreal cover in the watershed in 1975 when
native forest occupied just 26 of the watershed surface whereas shrublands and
98
meadows occupied more than the 70 of the surface of the watershed with the
concomitant elevated loss of N (Fig 7 and Fig 8)
54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years
Sedimentological compositional and geochemical indicators all show changes in
the N cycle associated to LUCC lake dynamics and climate changes (Fig 10) From
the pre-Columbian indigenous settlement including the Spanish colonial period up
to the start of the Republic (1300 - 1800 CE) the introduction of crops such as
quinoa and wheat but also the clearing of land for extensive agriculture would have
favored the entry of N into the lake Conversely major changes observed during the
last century were characterized by a sharp decrease of N input that were coeval
with the increase of tree plantations
From 1365 until 1550 CE (unit 2c 2b and half of 2a Fig 3 and Fig 10) pre-
Columbian agricultural societies planted mostly crops of corn and quinoa (Ramirez
and Vidal 1985) This period is characterized by large peaks seen in the δsup1⁵N record
(eg 1403 1452 1476 and 1505) with overall high δsup1⁵N values (up to 5permil) indicating
that N input from watershed was elevated and oscillating to the beat of the NT These
positive δsup1⁵N peaks could be due to several causes including a) the clearing of land
for farming b) N loss via denitrification which would be generally augmented in
anoxic environments (Diebel et al 2009) such as Lago Vichuqueacuten (low MnFe
values about 0001 Fig 4) or c) climate especially cold-wet winters or hot-dry
summers can also exert control on the δsup1⁵N record Indeed tree-ring records and
summer temperature reconstructions show overall wetcold conditions during this
period (Christie et al 2009 Von Gunten et al 2009 Fig 10) Increased
precipitation would bring more sediment (and nutrients) from the watershed into the
99
lake and increase lake productivity which is also detected by the geochemical
proxies for productivity (peaks of BrTi SrCa and TOC Fig 4 and Fig 10) (see also
Frugone-Alvarez et al 2017)
Figure 10 Changes in the N availability during the last six centuries in Lago
Vichuqueacuten a) lake sediment δsup1⁵N values displayed more positive values during the
prehistoric period Spanish Colony and the starting 19th century which is associated
with enhanced N input from the watershed by extensive clearing and crop
plantations The inset shows this relationship between sediment δsup1⁵N and
100
percentage of meadow cover over the last 30 years b) Summer temperature
reconstruction from central Chile (von Gunten et al 2009) showing a
correspondence with cold phases and high δsup1⁵N values at Vichuqueacuten except for the
last 100 years c) Palmer Drought Severity Index (PDSI) is an interannual moisture
variability reconstruction for late springndashearly summer during the last six centuries
(Christie et al 2009) Grey shadow indicating higher precipitation periods
From 1550 and 1810 CE (unit 2a) and during the Colonial period several peaks
of δ15N could be further related not only to high precipitation (Fig 10 grey shadow)
but also pulses of enhanced N input from the watershed linked to human land use
In 1550 CE Juan Cuevas was granted lands and indigenous workers under the
encomienda system for agricultural and mining development of the Vichuqueacuten
village (Ramiacuterez and Vidal 1985) Historical documents show that around 1579 CE
the Vichuqueacuten watershed was occupied by indigenous communities dedicated to
wheat plantations and vineyards wood extraction and gold mining (Odone 1998)
The introduction of the Spanish agricultural system implied not just a change in the
types of crops used (from quinoa to vineyards and wheat) but also a clearing of
native species for the continuous increase of agricultural surface and wood
extraction During the Colonial period wheat from Vichuqueacuten was exported to Peru
(Ramiacuterez and Vidal 1985) Armesto et al (2009) indicate that during the XVIII and
XIX centuries the extraction of wood for mining operations was important enough to
cause extensive loss of native forests The independence and instauration of the
Chilean Republic did not change this prevailing system Increases in the
contributions of N to the lake during the second half of the XIX century (peaks in
δsup1⁵N ca 1870 CE) could be due to expanding farm land dedicated for wheat
101
production and increased commercial trade with California and Canada (Ramiacuterez
and Vidal 1985)
In contrast LUCC in the last century are clearly related to the development of
large-scale tree plantations (Fig 7) The δsup1⁵N record reaches the lowest values of
the entire sequence in the last few decades (Fig 10) A marked increase in lake
productivity NT concentration and decreasing sediment input is synchronous (unit
1 Fig 4) with trees replacing meadows shrublands and areas with native forests
(Fig 7 and 8) The implementation of the lsquoForest Lawrsquo in 1974 had a stronger impact
on the landscape and lake ecosystem dynamics than the impacts of ongoing climate
change in the region which is much more recent (Garreaud et al 2018) although
the prevalence of hot dry summers seen over the last decade would also be
associated with low δsup1⁵N values and increase of NT (Fig 10) Heavy metals ratios
(FeTi CoTi and PbTi) show notable increases in lake sediments from 1992 to 2011
CE (Fig 4) Although this could be related to mining in the El Maule region the
closest mines are 60 Km away (Pencahue and Romeral) so local factors related to
shoreline urbanization for the summer homes and an increase in tourist activity
could also be a major factor
6 Conclusions
The N isotope signal in the watershed depends on the rates of exchange
between vegetation and soil cover (Fig 11) Plants preferentially uptake 14N and the
underlying soils become enriched in 15N especially when the terrestrial ecosystem
is N-limited andor significant N loss occurs (ie denitrification andor ammonia
volatilization) LUCC in the Lago Vichuqueacuten watershed have heavily modified the
links between terrestrial and aquatic ecosystems with agriculture practices
102
contributing more N to the lake than tree plantations or native forests In situ lake
processes can also fractionate N isotopes An increase of N-fixing species results
in OM depleted in 15N which results in POM with lower δsup1⁵N values during these
periods During winter phytoplankton is typically enriched in 15N due to the
decreased abundance of N-fixing species and increased N input from the
watershed This in turn generates the highest δsup1⁵NPOM values in Lago Vichuqueacuten
Periods of elevated productivity tend to deplete the dissolved inorganic N of 14N
resulting in even higher δ15N values
Our study illustrates the usefulness of δsup1⁵N as a tool for reconstructing the past
influence of LUCC on N availability in lake ecosystems To constrain the relative
roles of the diverse forcing mechanisms that can alter N cycling in mediterranean
ecosystems all main components of the N cycle should be monitored seasonally
(or monthly) including the measurements of δ15N values in land samples
(vegetation-soil) as well as POM
103
Figure 11 Summary of human and environmental factors controlling the δ15N
values of lake sediments Particulate organic matter(POM) δ15N values in
mediterranean lakes are driven by N input from the watershed that in turn depend
on land use and cover changes (ie forest plantation agriculture) andor seasonal
changes in N sources andor lake ecosystem processes (ie bioproductivity redox
condition denitrification) Arrows indicate the direction of N fluxes (inputoutput from
the N cycle) N cycle processes that deplete lake sediments of 15N are shown in
blue whereas those that enrich sediments in 15N are shown in red
104
Supplementary material
Figure S1 Pearson correlate coefficient between geochemical variables in core
VIC13-2B Positive and large correlations are in blue whereas negative and small
correlations are in red (p valuelt0001)
Figure S2 Principal Component Analysis of geochemical elements from core
VIC13-2B
105
Table S1 Lago Vichuqueacuten radiocarbon samples
RADIOCARBON
LAB CODE
SAMPLE
CODE
DEPTH
(m)
MATERIAL
DATED
14C AGE ERROR
D-AMS 029287
VIC13-2B-
1 043 Bulk 1520 24
D-AMS 029285
VIC13-2B-
2 085 Bulk 1700 22
D-AMS 029286
VIC13-2B-
2 124 Bulk 1100 29
Poz-63883 Chill-2D-1 191 Bulk 945 30
D-AMS 001133
VIC11-2A-
2 201 Bulk 1150 44
Poz-63884
Chill-2D-
1U 299 Bulk 1935 30
Poz-64089
VIC13-2D-
2U 463 Bulk 1845 30
Poz-64090
VIC13-2A-
3U 469 Bulk 1830 35
D-AMS 010068
VIC13-2D-
4U 667 Bulk 2831 25
Poz-63886
VIC13-2D-
4U 719 Bulk 3375 35
106
D-AMS 010069
VIC13-2D-
5U 775 Bulk 3143 27
Poz-64088
VIC13-2D-
5U 807 Bulk 3835 35
D-AMS-010066
VIC13-2D-
7U 1075 Bulk 6174 31
Poz-63885
VIC13-2D-
7U 1197 Bulk 6440 40
Poz-5782 VIC13-15 DIC 180 25
Table S2 Loadings of the trace chemical elements used in the PCA
Elementos PC1 PC2 PC3 PC4
Zr 0922 0025 -0108 -0007
Zn 0913 -0124 -0212 0001
Rb 0898 -0057 -0228 0016
K 0843 0459 0108 0113
Ti 0827 0497 0060 -0029
Al 0806 0467 0080 0107
Si 0803 0474 0133 0136
Y 0784 -0293 -0174 0262
V 0766 0455 0090 -0057
Br 0422 -0716 -0045 0226
Ca 0316 -0429 0577 0489
Sr 0164 -0420 0342 -0182
Cl 0151 -0781 -0397 0162
107
Mn -0121 -0091 0859 0095
S -0174 -0179 -0051 0714
Pb -0349 0414 -0282 0500
Fe -0700 0584 -0023 0280
Co -0704 0564 -0107 0250
Table S3 Stable Isotope values from vegetation of Lago Vichuqueacutenacutes watershed
Taxa Classification δsup1⁵N δsup1sup3C CN
molar
Poaceae Meadow 1216 -2589 3602
Juncacea Meadow 1404 -2450 3855
Cyperaceae Meadow 1031 -2596 1711
Taraxacum
officinale Meadow 836 -2400 2035
Poaceae Meadow 660 -2779 1583
Poaceae Meadow 453 -2813 1401
Poaceae Meadow 966 -2908 4010
Juncus Meadow 1247 -2418 3892
Poaceae Meadow 747 -3177 6992
Poaceae Meadow 942 -2764 3147
Poaceae Meadow 1479 -2634 2895
Poaceae Meadow 1113 -2776 1795
Poaceae Meadow 2215 -2737 7971
Poaceae Meadow 1121 -2944 2934
Poaceae Meadow 638 -3206 1529
108
Macrophytes Macrophytes 886 -3044 2286
Macrophytes Macrophytes 1056 -2720 2673
Macrophytes Macrophytes 769 -3297 1249
Macrophytes Macrophytes 967 -2763 1442
Macrophytes Macrophytes 959 -2670 2105
Macrophytes Macrophytes 334 -2728 1038
Acacia dealbata
Introduced
species 656 -2696 1296
Acacia dealbata
Introduced
species 487 -2941 1782
Acacia dealbata
Introduced
species 220 -2611 3888
Luma apiculata Native species 433 -2542 4135
Luma apiculata Native species 171 -2664 7634
Luma apiculata Native species -001 -2736 6283
Luma apiculata Native species 029 -2764 6425
Azara sp Native species 159 -2868 8408
Azara sp Native species 101 -2606 2885
Baccharis concava Native species 104 -2699 5779
Baccharis concava Native species 265 -2488 4325
Baccharis concava Native species 287 -2562 7802
Baccharis concava Native species 427 -2781 5204
Baccharis linearis Native species 190 -2610 4414
Baccharis linearis Native species 023 -2825 5647
109
Peumus boldus Native species 042 -2969 6327
Peumus boldus Native species 205 -2746 4110
Peumus boldus Native species 183 -2743 6293
Chusquea quila Native species 482 -2801 4275
Poaceae meadow 217 -2629 7214
Lobelia sp Native species 224 -2645 3963
Lobelia sp Native species -091 -2565 4538
Aristotelia chilensis Native species -035 -2785 5247
Aristotelia chilensis Native species -305 -2889 2305
Aristotelia chilensis Native species 093 -2836 5457
Chusquea quila Native species 173 -2754 3534
Chusquea quila Native species 045 -2950 6739
Quillaja saponaria Native species 223 -2838 9385
Scirpus meadow 018 -2820 7115
Sophora sp Native species -184 -2481 2094
Sophora sp Native species -181 -2717 1721
Pinus radiata
Introduced
trees 1581 -2602 3679
Pinus radiata
Introduced
trees 1431 -2784 4852
Pinus radiata
Introduced
trees -091 -2708 9760
Pinus radiata
Introduced
trees 153 -2568 3470
110
Salix sp
Introduced
trees 632 -2878 1921
LITERATURE CITED
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A historical
framework for land cover change in southwestern South America in the past 15000
years Land use policy 27 148ndash160
httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the next
carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014
httpsdoiorg101002eft2235
Blaauw M Christeny JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474
httpsdoiorg10121411-BA618
Bowman DMJS Cook GD 2002 Can stable carbon isotopes (δ13C) in soil
carbon be used to describe the dynamics of Eucalyptus savanna-rainforest
boundaries in the Australian monsoon tropics Austral Ecol
httpsdoiorg101046j1442-9993200201158x
Brahney J Ballantyne AP Turner BL Spaulding SA Otu M Neff JC 2014
Separating the influences of diagenesis productivity and anthropogenic nitrogen
deposition on sedimentary δ15N variations Org Geochem 75 140ndash150
httpsdoiorg101016jorggeochem201407003
111
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409
httpsdoiorg102134jeq20090005
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego R
Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and
environmental change from a high Andean lake Laguna del Maule central Chile
(36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the
Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from
tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Craine JM Elmore AJ Aidar MPM Dawson TE Hobbie EA Kahmen A
Mack MC Kendra K Michelsen A Nardoto GB Pardo LH Pentildeuelas J
Reich B Schuur EAG Stock WD Templer PH Virginia RA Welker JM
Wright IJ 2009 Global patterns of foliar nitrogen isotopes and their relationships
with cliamte mycorrhizal fungi foliar nutrient concentrations and nitrogen
availability New Phytol 183 980ndash992 httpsdoiorg101111j1469-
8137200902917x
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty
Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-
010-9453-1
Diebel M Vander Zanden MJ 2012 Nitrogen stable isotopes in streams stable
isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19
1127ndash1134 httpsdoiorg10189008-03271
112
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in freshwater
wetlands record long-term changes in watershed nitrogen source and land use SO
- Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash
2916
Fariacuteas L Castro-Gonzales M Cornejo M Charpentier J Faundez J
Boontanon N Yoshida N 2009 Denitrification and nitrous oxide cycling within the
upper oxycline of the oxygen minimum zone off the eastern tropical South Pacific
Limnol Oceanogr 54 132ndash144
Farquhar GD Ehleringer JR Hubick KT 1989 Carbon Isotope Discrimination
and Photosynthesis Annu Rev Plant Physiol Plant Mol Biol
httpsdoiorg101146annurevpp40060189002443
Farquhar GD OrsquoLeary MH Berry JA 1982 On the relationship between
carbon isotope discrimination and the intercellular carbon dioxide concentration in
leaves Aust J Plant Physiol httpsdoiorg101071PP9820121
Fogel ML Cifuentes LA 1993 Isotope fractionation during primary production
Org Geochem httpsdoiorg101007978-1-4615-2890-6_3
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A
Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-
resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS)
implications for past sea level and environmental variability J Quat Sci 32 830ndash
844 httpsdoiorg101002jqs2936
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924
httpsdoiorg104319lo20095430917
113
Gerhart LM McLauchlan KK 2014 Reconstructing terrestrial nutrient cycling
using stable nitrogen isotopes in wood Biogeochemistry 120 1ndash21
httpsdoiorg101007s10533-014-9988-8
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and oxygen
isotope fractionation during dissimilatory nitrate reduction by denitrifying bacteria 53
2533ndash2545 httpsdoiorg10230740058342
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater
eutrophic lake Limnol Oceanogr 51 2837ndash2848
httpsdoiorg104319lo20065162837
Harris D Horwaacuteth WR van Kessel C 2001 Acid fumigation of soils to remove
carbonates prior to total organic carbon or CARBON-13 isotopic analysis Soil Sci
Soc Am J 65 1853 httpsdoiorg102136sssaj20011853
Houmlgberg P 1997 Tansley review no 95 natural abundance in soil-plant systems
New Phytol httpsdoiorg101046j1469-8137199700808x
Kylander ME Ampel L Wohlfarth B Veres D 2011 High-resolution X-ray
fluorescence core scanning analysis of Les Echets (France) sedimentary sequence
New insights from chemical proxies J Quat Sci 26 109ndash117
httpsdoiorg101002jqs1438
Lara A Solari ME Prieto MDR Pentildea MP 2012 Reconstruccioacuten de la
cobertura de la vegetacioacuten y uso del suelo hacia 1550 y sus cambios a 2007 en la
ecorregioacuten de los bosques valdivianos lluviosos de Chile (35o - 43o 30acute S) Bosque
(Valdivia) 33 03ndash04 httpsdoiorg104067s0717-92002012000100002
Lehmann M Bernasconi S Barbieri A McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during
114
simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66
3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007
Liu G Li M An L 2007 The Environmental Significance Of Stable Carbon
Isotope Composition Of Modern Plant Leaves In The Northern Tibetan Plateau
China Arctic Antarct Alp Res httpsdoiorg1016571523-0430(07-505)[liu-
g]20co2
Marzecovaacute A Mikomaumlgi A Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54
httpsdoiorg103176eco2011105
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash
1643 httpsdoiorg1011770959683613496289
Meyers PA 2003 Application of organic geochemistry to paleolimnological
reconstruction a summary of examples from the Laurention Great Lakes Org
Geochem 34 261ndash289 httpsdoiorg101016S0146-6380(02)00168-7
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland
Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Odone C 1998 El Pueblo de Indios de Vichuqueacuten siglos XVI y XVII Rev Hist
Indiacutegena 3 19ndash67
Pausch J Kuzyakov Y 2018 Carbon input by roots into the soil Quantification of
rhizodeposition from root to ecosystem scale Glob Chang Biol
httpsdoiorg101111gcb13850
115
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98
httpsdoiorg1011772053019614564785
Sun W Zhang E Jones RT Liu E Shen J 2016 Biogeochemical processes
and response to climate change recorded in the isotopes of lacustrine organic
matter southeastern Qinghai-Tibetan Plateau China Palaeogeogr Palaeoclimatol
Palaeoecol httpsdoiorg101016jpalaeo201604013
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of
different trophic status J Paleolimnol 47 693ndash706
httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl
httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M 2009 High-resolution quantitative climate
reconstruction over the past 1000 years and pollution history derived from lake
sediments in Central Chile Philos Fak PhD 246
Woodward C Chang J Zawadzki A Shulmeister J Haworth R Collecutt S
Jacobsen G 2011 Evidence against early nineteenth century major European
induced environmental impacts by illegal settlers in the New England Tablelands
south eastern Australia Quat Sci Rev 30 3743ndash3747
httpsdoiorg101016jquascirev201110014
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K Yeager
KM 2016 Different responses of sedimentary δ15N to climatic changes and
116
anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau
J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
117
DISCUSION GENERAL
El Nitroacutegeno (N) es un elemento clave que controla la dinaacutemica y
funcionamiento de ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al
1997) Pese a que el N (N2) es muy abundante en la atmoacutesfera (78) es escaso
en los ecosistemas pues la mayoriacutea de la biota no puede acceder a su forma
molecular (Galloway et al 1995 Zaehle et al 2013) En este sentido la entrada
natural de N en los ecosistemas es viacutea fijacioacuten bioloacutegica del N2 por bacterias que lo
convierten en compuestos bioloacutegicamente disponibles (Battye et al 2017) Debido
a que la fijacioacuten es un proceso energeacuteticamente costoso los ecosistemas
comuacutenmente estaacuten limitados por N (Vitousek and Howarth 1991) Los LUCC
contribuyen al incremento del N disponible y son una de las principales causas de
eutroficacioacuten y contaminacioacuten de los cuerpos de agua (Woodward et al 2012)
En Chile central los LUCC principalmente relacionados con las actividades
agriacutecolas y forestales han causado grandes cambios en el ciclo del N debido al
118
reemplazo de la cobertura natural por un monocultivo y al uso de fertilizantes que
modifican los aportes de MO y N a los cuerpos de agua El programa de
estabilizacioacuten de laderas impulsada por el gobierno (Albert 1900) y la ley forestal
de 1974 han fomentado la forestacioacuten con plantaciones de Pinus radiata y
Eucalyptus globulus y en el caso de Lago Vichuqueacuten estas actualmente cubren maacutes
del 60 de su cuenca (Cap2 Fig 5) Ademaacutes en Laguna Matanzas la
sobreexplotacioacuten del agua en conjunto con el cambio climaacutetico actual el que ha
conducido a una de las megasequiacuteas maacutes severas en las uacuteltimas deacutecadas
(Garreaud et al 2017) generoacute una combinacioacuten letal que secoacute el lago en tan solo
10 antildeos (Cap1 Fig1) Las evidencias obtenidas a partir de estos dos lagos
permiten identificar las huellas del Antropoceno en Chile central basadas en el
registro sedimentario lacustre
La transferencia de N desde los ecosistemas terrestres a acuaacuteticos es un
proceso natural con controles bioacuteticos climaacuteticos y geoloacutegicos El enlace
hidroloacutegico entre los ecosistemas terrestres y acuaacuteticos hace que el estado troacutefico
de los lagos esteacute vinculado al de su cuenca (McLauchlan et al 2013) En Chile
central se sabe poco del impacto de los LUCC sobre el ciclo de nutrientes en los
ecosistemas lacustres aunque al menos desde la colonizacioacuten espantildeola se tienen
registros de influencia humana en las cuencas Durante la colonia espantildeola
Laguna Matanzas fue una hacienda ganadera que exportaba productos caacuternicos al
Peruacute (Contreras-Lopez et al 2014) A su vez en el Lago Vichuqueacuten se realizaban
extensas actividades agriacutecolas hasta comienzos del s XX como por ejemplo
cultivos de vid y trigo ademaacutes de extraccioacuten de madera y mineriacutea de oro (Odone
1997) En las uacuteltimas deacutecadas los LUCC han estado vinculados principalmente con
el desarrollo forestal al compaacutes de una estrategia gubernamental de fomentar con
119
incentivos financieros (Ley de Bosques DFL nordm265 de 1931 y DFL 701 de 1974)
esta actividad El incremento de la superficie forestal es especialmente fuerte en
ambos lagos a partir de la deacutecada de 1980 y hasta 2016 (Laguna de Matanzas 0-
17 Lago Vichuqueacuten 1-66) Este aumento habriacutea sido en detrimento del bosque
nativo y los pastizales (Cap 1 Fig 5 y Cap 2 Fig 8) El incremento de la superficie
forestal generariacutea una disminucioacuten de los aportes de sedimentos y nutrientes al lago
y en este sentido un cambio de estado en los flujos de N (e g tipping points) que
a su vez ha quedado registrado en la sentildeal isotoacutepica (δ15N) y la acumulacioacuten de
MO en los sedimentos lacustres
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central
Los sedimentos lacustres son verdaderos ldquosumiderosrdquo que pueden llegar a
registrar lo que ocurre en cuanto a la historia ambiental de su cuenca En esta tesis
se utilizan los isoacutetopos estables de N (15N y 14N) en sedimentos lacustres para
reconstruir los cambios en la disponibilidad de N en el tiempo y establecer la
magnitud de impacto generado por actividades humanas El fraccionamiento
cineacutetico en los lagos es mediado por procesos bioloacutegicos que mediante la
asimilacioacuten de N genera un producto (eg fitoplancton) maacutes ldquoligerordquo (valores maacutes
bajos de 15N) que su remanente (eg DIN) (Fogel y Cifuentes 1993) Sin embargo
en periacuteodos en que los lagos estaacuten limitados por N el fitoplancton discrimina menos
y la MO resultante queda enriquecida en 15N (Fig 1 a y b) Ademaacutes la
desnitrificacioacuten acentuada en lagos de fondo anoacutexicos (eg Laguna Matanzas
entre 1800 y 1940 Cap1 Fig 6) conduce a un enriquecimiento en 15N de los
sedimentos y de la columna de agua Esta informacioacuten queda reflejada en la MO
120
de los sedimentos lacustres lo que permite conocer la disponibilidad del N en el
tiempo a partir de las variaciones de 15N
En esta tesis analizamos la MO de los sedimentos lacustres para reconstruir
la influencia de los LUCC en los aportes de sedimentos y N al lago En teacuterminos de
asimilacioacuten de N se puede distinguir entre dos grupos principales de productores
primarios que componen el POM (Fig1)
1 Los que asimilan el DIN compuesto por amonio nitratos y nitritos Aquiacute el
δ15N resultante fluctuaraacute en torno a valores maacutes positivos en la medida que
la productividad se incrementa o no hay reposicioacuten del DIN (Fig 1)
2 Los fijadores de N Dado que es un proceso energeacuteticamente costoso en
ambientes que no estaacuten limitados por N muchas veces son excluiacutedas
competitivamente por el resto del fitoplancton Si el DIN queda agotado por
el incremento de la productividad (o bien si el ambiente estaacute limitado ya sea
por N o por otros nutrientes) los fijadores de N pueden florecer y la MO que
se genera a partir de ello tendraacute un δ15N con valores en torno a 0permil
De este modo la MO en los sedimentos lacustres dependeraacute de la
composicioacuten de productores primarios (ie fijadores vs asimiladores del DIN)
ademaacutes de los aportes desde la cuenca tanto de MO como del N de la cuenca que
pasa a formar parte del DIN (eg fertilizantes estieacutercol) (Fig 1)
La MO de los lagos estudiados en esta tesis ha sido analizada a partir de
variables geoquiacutemicas isotoacutepicas frotis y estaacute compuesta principalmente por
diatomeas y MO amorfa A partir de los datos obtenidos en esta tesis los valores
de δ15N aumentan cuando hay una mayor cobertura agriacutecola o pastizales A su vez
tiende a disminuir cuando aumenta la cobertura arboacuterea independiente si esta es
por plantaciones forestales o por bosque nativo
121
Por otra parte la dinaacutemica del lago tambieacuten imprime caracteriacutesticas
especiales en el POM observaacutendose variaciones estacionales en los valores
δ15NPOM Durante la estacioacuten caacutelida y seca se observan valores maacutes bajos que
durante la estacioacuten friacutea y huacutemeda posiblemente relacionado con el incremento de
la contribucioacuten de especies fijadoras de N (Lago Vichuqueacuten Fig 9 Cap 2) durante
el verano En la estacioacuten friacutea y huacutemeda el DIN seriacutea la principal fuente de N Las
mayores entradas de MO y N terrestre debidos a un incremento del lavado de la
cuenca como consecuencia de las precipitaciones y la mineralizacioacuten de la MO
podriacutean estimular la produccioacuten de MO lacustre por crecimiento del fitoplancton
Como consecuencia se observan tendencias decrecientes de los valores de
δ15N en la MO sedimentaria cuando la productividad en el lago estaacute relacionada
con especies fijadoras de N mientras que el δ15N muestra aumentos cuando la
productividad del lago estaacute asociada principalmente al consumo del DIN pero
tambieacuten cuando predominan procesos de perdida de N (eg desnitrificacioacuten Fig
1)
Hemos identificado tres fases en la dinaacutemica del δ15N para ambos lagos
Entre 1800 y 1940 CE la cuenca de laguna de Matanzas estaba dominada por
actividades agriacutecolas y ganaderas (δ15Nsuelo gt 4permil Cap 1 Fig 4 y 6) Las entradas
de sedimentos y MO desde la cuenca son altas (δ13C lt -209permil ZrTi ~029 AlTi
~013) y coincide con un clima maacutes huacutemedo y friacuteo (Von Gunten et al 2009
Christiansen et al 2011) El lago teniacutea un nivel de agua maacutes profundo dominado
por condiciones anoacutexicas (MnFe ~0013) que contribuiriacutean a mayores valores de
δ15N en la columna de agua y por ende de los sedimentos acumulados en el fondo
debido a la peacuterdida diferencial de 14N viacutea desnitrificacioacuten (Fig 7 Cap1) La baja
122
produccioacuten de MO lacustre (BrTi ~002 TOC~17) coincide con un bajo contenido
de NT (~478 μg) y valores maacutes positivos de δ15N (gt 4permil)
La actividad agriacutecola tambieacuten dominaba en la cuenca del Lago Vichuqueacuten
durante este periacuteodo (δ15Nsuelo gt4permil Cap2 Fig 6) El aporte sedimentario desde la
cuenca es alto (AlTi ~029 ZrTi ~032) y fue favorecido por mayores
precipitaciones a las actuales (Christiansen et al 2011) El Lago Vichuqueacuten es un
lago estratificado anoacutexico (MnFe ~002) y profundo (~30 m) por lo que la
desnitrificacioacuten juega un rol importante en el enriquecimiento de la MO
sedimentaria La baja produccioacuten de MO (BrTi ~007 TOC ~12) de origen
lacustre (δ13C ~ -268permil) coincide con un bajo contenido de NT (~309μg +6) y
valores maacutes positivos de δ15N (56permil +03)
Durante esta fase en ambos lagos los aportes de N de la cuenca parecen
ser los principales responsables en las fluctuaciones del δ15N Esta sentildeal podriacutea
estar amplificada por bajas temperaturas (que afectan la productividad lacustre) y
altas precipitaciones que favorecen el lavado de la cuenca y aumentan el aporte de
sedimentos y MO desde la cuenca predominantemente agriacutecola
Entre 1940 y 1980 CE en Laguna Matanzas se observa un incremento en
la acumulacioacuten de MO algal (TOC ~2 +1 δ13C ~-230+09permil) El ambiente
deposicional es ligeramente menos turbulento que el anterior (AlTi ~011+001
ZrTi ~03+001) y el lago estaacute en una fase de transicioacuten hacia condiciones maacutes
oacutexicas (MnFe ~002+0004) y maacutes productivas (BrTi ~004+0007) A su vez δ15N
tiende hacia valores maacutes negativos (δ15N ~30permil+03) mientras que el NT aumenta
oscilando en antifase con el δ15N
En Lago Vichuqueacuten en cambio se observa un ligero incremento en la
acumulacioacuten de MO (TOC ~13 +02) de origen terrestre (δ13C ~-27permil +04) La
123
productividad es similar al periacuteodo anterior (BrTi ~007+003) y el ambiente
deposicional es menos turbulento (AlTi ~02 +009 Zrti ~033 +007) El δ15N y el
NT oscilan en torno a valores ligeramente maacutes bajos (δ15N ~51permil +04 NT ~292μg
+6) Este periacuteodo se caracteriza por ser maacutes caacutelido que el anterior lo que
posibilitariacutea un incremento en la productividad lacustre en Laguna Matanzas pero
que no es observada en el Lago Vichuqueacuten
Entre 1980 y la actualidad en Laguna Matanzas habriacutea un incremento de la
acumulacioacuten de MO (TOC ~59 + 3) asociada a un incremento en la productividad
del lago (BrTi ~009+005) y a los aportes de MO terrestres (δ13C ~-24 + 2permil) El
lago se vuelve maacutes oacutexico (Mn ~0003 +0006) y las entradas de sedimento
disminuyen (AlTi~ 009 +002) El δ15N tiende a valores maacutes negativos (δ15N ~13permil
+1) y el NT aumenta de 20 a 55 μg (NT ~346 + 9 μg) tomando valores sin
precedentes en todo el registro En los sedimentos lacustres del Lago Vichuqueacuten
tambieacuten se observa un incremento de la acumulacioacuten de MO (TOC ~17 + 05)
asociada a un incremento en la productividad del lago (BrTi ~01 + 003) y las
entradas de sedimento desde la cuenca disminuyen (AlTi ~005 + 005) El δ15N
(~43 + 05permil) tiende a valores maacutes negativos y el NT incrementa de 20 a 55 μg (NT
~346 + 9 μg) oscilando en antifase
Durante esta fase en ambos lagos se observa un aumento en la
acumulacioacuten de MO sedimentaria un incremento en el NT y valores maacutes negativos
de δ15N que coincide con el incremento de la superficie forestal de las cuencas
(Laguna Matanzas gt17 Lago Vichuqueacuten gt60)
124
Figura 2 Comparacioacuten de los cambios del ciclo del N entre Laguna Matanzas y
Lago Vichuqueacuten con los procesos biogeoquiacutemicos contrastantes en rojo En L
Matanzas las condiciones oacutexicas podriacutean estar favoreciendo la nitrificacioacuten del
amonio En Vichuqueacuten las condiciones anoacutexicas favorecen un aumento en δ15N de
la MO en los sedimentos por desnitrificacioacuten y mineralizacioacuten
Los ambientes mediterraacuteneos en el que los lagos del presente estudio se
encuentran insertos estaacuten caracterizados por una estacioacuten seca prolongada y las
precipitaciones ocurren en eventos puntuales alcanzando altos montos
pluviomeacutetricos en poco tiempo Las lluvias favorecen un lavado de la cuenca la
perdida de N y otros nutrientes Por lo que es esperable que los sedimentos del
lago reflejen mejor los valores isotoacutepicos de los suelos de la cuenca durante los
periacuteodos con altos montos pluviomeacutetricos En el Lago Vichuqueacuten por ejemplo el
125
POM superficial (2 y 5 m de profundidad) despliega valores isotoacutepicos maacutes
positivos en invierno presumiblemente como resultado de mayores aportes de MO
y N de su cuenca (Fig 1b) En ambos lagos los valores maacutes positivos en los
sedimentos lacustres ocurren durante periacuteodos con mayores montos pluviomeacutetricos
(Cap1 Fig 6 y Cap 2 Fig 12)
Ademaacutes del efecto de la precipitacioacuten en el aporte de nutrientes al lago en
esta tesis hemos encontrado que los mayores aportes de N desde la cuenca se dan
cuando la cobertura vegetal del suelo es menor En Laguna Matanzas el desarrollo
de la actividad ganadera desde la llegada de los espantildeoles a la cuenca habriacutea
incrementado los aportes de N al lago Los valores de δ15N en los sedimentos
lacustres depositados durante este periacuteodo son los maacutes positivos de todo el registro
(~4permil) A su vez en Lago Vichuqueacuten los valores isotoacutepicos maacutes positivos se
registran entre 1300 y 1900 CE que coinciden con el mayor desarrollo de
actividades agriacutecola y de extraccioacuten de maderas de la cuenca (Fig 10 Cap 2)
Los recientes LUCC (a partir de 1975) en las cuencas de Laguna Matanzas
y Lago Vichuqueacuten estaacuten caracterizados por un incremento de la superficie forestal
y agriacutecola en reemplazo de la cobertura de pastizal (Laguna Matanzas) y bosque
nativo (Lago Vichuqueacuten) Los valores de δ15N en los testigos sedimentarios fueron
maacutes positivos cuanto mayor la superficie agriacutecola de la cuenca y maacutes negativos
cuanto mayor la superficie forestal Aunque por los alcances de esta tesis no
podemos ser concluyentes respecto del origen del NT (aloacutectonoautoacutectono) parece
ser que esta maacutes ligado a los aportes de la cuenca pese a la disminucioacuten del aporte
sedimentario observado en ambos lagos Las plantaciones forestales a diferencia
del bosque nativo son fertilizadas con una mezcla de N P y K (Toral et al 2005)
Por lo que pese a la disminucioacuten del aporte sedimentario la concentracioacuten de
126
nutrientes en el aporte al lago puede verse incrementada por el tipo de manejo
forestal con respecto al bosque nativo
Los resultados del primer capiacutetulo demuestran que 1) las plantaciones
forestales yo bosque nativo retiene maacutes sedimentos y MO mientras que el uso de
suelo agriacutecola y los pastizales destinados al desarrollo de la actividad ganadera lo
libera 2) Al parecer la desnitrificacioacuten es el proceso natural maacutes importante de
perdida de N en lagos costeros y favorece el enriquecimiento de 15N tanto en la
columna de agua (ie 15N-DIN) como en los sedimentos una vez enterrados y 3) la
desnitrificacioacuten depende de los cambios en las condiciones REDOX en el lago La
oxigenacioacuten en lagos poco profundos -como Laguna Matanzas- es altamente
fluctuante a escalas decadal-centenal depende de la profundidad de la laacutemina de
agua y de la variabilidad de la precipitacioacuten En este sentido en Laguna Matanzas
habriacutea condiciones anoacutexicas en ambientes cuando el lago alcanza niveles maacutes
altos Estas condiciones se observan entre 1820 y 1940 CE y son coetaacuteneas con
episodios maacutes huacutemedos y friacuteos (von Gunten et al 2009 Christie et al 2009) pero
tambieacuten con una fuerte actividad ganadera en la cuenca
Las fuentes variables de N y su fluctuacioacuten en el tiempo hacen necesario
contar con un anaacutelogo moderno que permite usar al δ15N de los sedimentos
lacustres como un indicador indirecto de los cambios en la disponibilidad de N en
el tiempo Por ello en el segundo capiacutetulo integramos muestras de suelo-
vegetacioacuten y un monitoreo de POM de la columna de agua en verano e invierno La
composicioacuten isotoacutepica de POM estariacutea reflejando cambios de la fuente de N (fijacioacuten
vs consumo del DIN) que variacutea durante el antildeo hidroloacutegico Durante el verano la
mayor temperatura y horas-luz favoreceriacutean las entradas de N al lago viacutea fijacioacuten
bioloacutegica de N2 atmosfeacuterico resultando en un incremento de la MO con un valor
127
isotoacutepico bajo (POM verano~5permil Fig 1b) El N2 (δ15N = 0permil) es fijado praacutecticamente
sin fraccionamiento isotoacutepico generando un δ15N de la OM cercano a 0permil (Leng et
al 2006) En invierno las precipitaciones pueden estar favoreciendo un incremento
en la entrada de nutrientes al lago El DIN de la columna de agua llega a contener
valores de δ15N maacutes altos reflejando estos mayores aportes de la cuenca y el POM
del Lago Vichuqueacuten queda enriquecido en 15N (δ15N ~11permil Cap 2 Fig 9) Estas
variaciones estacionales pueden estar evidenciando un cambio en la composicioacuten
de especies de POM desde especies fijadoras a especies que consumen el N de la
columna de agua Sin embargo para corroborar esta informacioacuten seriacutea deseable
contar con el anaacutelisis de la composicioacuten de especies de las muestras de agua
extraidas
Entre los resultados maacutes importantes estaacuten los valores isotoacutepicos del suelo
y la biomasa representativa de la cuenca que incluye un listado de las especies
nativas de la zona que hasta ahora no se contaba (Cap 2 Tabla en material
suplementario) Las especies colectadas despliegan valores de δ15Nfoliar maacutes
positivos (plantaciones forestales y pastizal ~89permil) que el suelo (35permil) salvo por
las especies nativas las que oscilan en torno a valores cercanos al atmosfeacuterico
(~14permil) La razoacuten isotoacutepica 15N14N refleja la dinaacutemica de intercambio de N entre la
vegetacioacuten y el suelo (Koba et al 2002) La discriminacioacuten es positiva en la mayoriacutea
de los sistemas bioloacutegicos por lo que las plantas tienen valores de δ15N maacutes bajos
que el suelo (Evans et al 2001) como sucede con las especies nativas en Lago
Vichuqueacuten En consecuencia las plantas quedan empobrecidas en 15N y conducen
a un enriquecimiento en 15N del suelo sobretodo si no hay reposicioacuten de N por otras
viacuteas (eg fijacioacuten de N) Por lo tanto los valores maacutes negativos de δsup1⁵Nfoliar de las
especies nativas pueden estar relacionados con el consumo preferencial de 14N del
128
suelo donde la discriminacioacuten isotoacutepica de la vegetacioacuten contra el 15N conduce a
valores del δsup1⁵Nsuelo maacutes altos (Houmlgberg 1997) Por el contrario los valores maacutes
positivos de δsup1⁵Nfoliar de las plantaciones forestales y pastos con respecto al suelo
puede deberse por una parte que el suelo no cuenta con mecanismos naturales de
reposicioacuten de N salvo por la descomposicioacuten de la hojarasca que puede ser maacutes
lenta en Pinus radiata que en bosque nativo (Lusk et al 2001) y por otra por el alto
impacto de los aportes de N (y otros nutrientes) derivado de las actividades
humanas (eg uso de fertilizantes) en el suelo
El alcance maacutes significativo de esta tesis se relaciona con un cambio en la
tendencia maacutes negativa de δsup1⁵N y el incremento del NT a partir de 1940 y a partir
de 1970 CE en ambos lagos La causa maacutes probable de esta tendencia es el
reemplazo de la cobertura natural o preexistente a 1940 CE por plantaciones
forestales
En la figura 2 se observa una siacutentesis de los principales procesos que
afectan la sentildeal isotoacutepica de la MO en los sedimentos lacustres de L Matanzas y
L Vichuqueacuten La entrada de N y MO desde la cuenca depende del tipo de cobertura
Las plantaciones forestales y el bosque nativo retienen maacutes nutrientes y sedimentos
en la cuenca mientras que la agricultura los favorece Sin embargo las praacutecticas
de manejo que incluyen la aplicacioacuten de N (y otros nutrientes) pueden aportar maacutes
nutrientes al lago que la cobertra de bosque nativo Cuando las actividades
forestales cubren una mayor superficie de la cuenca (1940 en adelante) δsup1⁵N oscila
en valores maacutes negativos y el NT tiende a incrementarse en los sedimentos
lacustres Cabe destacar que los valores de δsup1⁵N observados en los testigos
sedimentarios a fines del s XX no tienen precedente durante la conquista y colonia
espantildeola o durante el resto del periodo de la Repuacuteblica
129
Figura 2 Modelo de transferencia de N entre los ecosistemas terrestres y
acuaacuteticos El δ15N de los sedimentos lacustres depende de la interaccioacuten entre los
aportes de la cuenca y porcesos ldquoin-lakerdquo Flechas indican la direccioacuten del flujo de
N En azul las salidasentradas de 14N en rojo las salidasentradas de 15N δ15N de
la interaccioacuten plantas-suelo depende del tipo de cobertura de suelo
130
CONCLUSIONES GENERALES
La transferencia de N entre cuencas y lagos es un factor de control del ciclo
del N en los lagos y queda reflejado en la sentildeal isotoacutepica de los sedimentos
lacustres (Leng et al 2006 McLauchlan et al 2007) En Laguna Matanzas el
suelo de las especies nativas y las plantaciones forestales despliegan valores de
δ15N maacutes bajos (~1permil) que los de Lago Vichuqueacuten (~35permil) Coincidentemente los
sedimentos lacustres de Laguna Matanzas oscilan con valores de δ15N maacutes bajos
(-15 a +45permil Fig 1) que los de Lago Vichuqueacuten (+3 a +8permil)
Por otra parte el estudio de la MO de los sedimentos lacustres ha permitido
reconstruir los cambios en los aportes de MO y N desde la cuenca Aunque no es
posible afirmar queacute tipo de N estaacute entrando al lago (eg Nitroacutegeno orgaacutenico e
inorganico) si podemos decir que los cambios en las fluctuaciones de δ15N son
coincentes con el crecimiento de la actividad forestal (1980 - hasta 2016 L
Matanzas 0-17 y L Vichuqueacuten 1-66) El δ15N fluctuacutea hacia valores maacutes
negativos Ademaacutes se observa un incremento del NT en los sedimentos lacustres
cuanto mayor es la superficie forestal
Entre 1800-1940 las cuencas eran fundamentalmente agriacutecolas y
ganaderas (Cap1 Fig 7 y Cap 2 Fig 12) el δ15N de los sedimentos lacustres
oscila con valores maacutes positivos (L Matanzas +40 plusmn 05permil y L Vichuqueacuten 56 plusmn
033permil Fig 1) hay una baja bioproductividad en el lago (BrTi ~002 TOC~17)
lo que coincide con un bajo contenido de NT (~478 μg) Este es un periacuteodo de altas
precipitaciones (Christie et al 2009) que tienden a favorecer el lavado de la cuenca
131
y con ello el aporte de MO sedimentos y nutrientes desde la cuenca tiende a verse
favorecido Aunque las principales actividades humanas en estas cuencas son
diferentes (ganaderiacutea en Laguna Matanzas Contreras-Lopez et al 2014
agricultura en Lago Vichuqueacuten Odone 1997) en ambos casos hubo un reemplazo
de la cobertura natural de bosques nativo que favorecioacute mayores aportes de N y
sedimentos desde la cuenca en un efecto sumado con el aumento de las
precipitaciones
A partir de 1980 CE en ambos lagos se observa una disminucioacuten en los
valores de δ15N y un incremento del NT que no tiene precedentes en todo el registro
y pese a que ambos lagos son limnologicamente muy diferentes En Lago
Vichuqueacuten el δ15N tiende a oscilar en antifase con el NT (Fig 1) En el caso de
Laguna Matanzas se observa una tendencia hacia valores maacutes negativos y a partir
de 1990s a valores maacutes positivos mientras que el NT tiende a incrementarse de
manera sostenida Ello podriacutea estar relacionado con el crecimiento de la actividad
forestal habriacutea una mayor retencioacuten de N y sedimentos en la cuenca ligado al
incremento de la superficie forestal Ademaacutes en el caso de L Matanzas el
incremento de la actividad agriacutecola (sumado al de las plantaciones forestales)
podriacutea expicar el cambio en la tendencia en la deacutecada de los 1990s
En el contexto de Antropoceno esta tesis nos permite identificar un gran
impacto de las actividades humanas en Chile central a partir de la deacutecada de 1940
y una Gran Aceleracoacuten a partir de la deacutecada de 1970s En el registro sedimentario
de los lagos en estudio se observa una gran acumulacioacuten de MO el δ15N oscila
hacia valores maacutes negativos y un gran incremento del NT y el crecimiento de la
actividad forestal parece ser la principal responsable Ademaacutes la sobreexplotacioacuten
132
del agua junto al cambio climaacutetico global ha generado una combinacioacuten letal para
los lagos costeros de Chile central
Figura 1 Comparacioacuten de las fluctuaciones de δ15N durante los uacuteltimos 300
antildeos en Laguna Matanzas y Lago Vichuqueacuten
133
Referencias
Battye W Aneja VP Schlesinger WH 2017 Earthrsquos Future Is nitrogen the next carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia
UC de V 2014 Elementos de la historia natural del An Mus Hist Natulas Vaplaraiso 27 51ndash67
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Evans RD Evans RD 2001 Physiological mechanisms influencing
plant nitrogen isotope composition 6 121ndash126 Fogel ML Cifuentes LA 1993 Isotope fractionation during primary
production Org Geochem 73ndash98 httpsdoiorg101007978-1-4615-2890-6_3 Galloway JN Schlesinger WH Ii HL Schnoor L Tg N 1995
Nitrogen fixation Anthropogenic enhancement-environmental response reactive N Global Biogeochem Cycles 9 235ndash252
Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J
Christie D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Koba K Koyama LA Kohzu A Nadelhoffer KJ 2003 Natural 15 N
Abundance of Plants Coniferous Forest httpsdoiorg101007s10021-002-0132-6
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos
Transactions American Geophysical Union httpsdoiorg1010292007EO070007 McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE
2007 Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
134
Phytologist N Oct N 2006 Tansley Review No 95 15N Natural Abundance in Soil-Plant Systems Peter Hogberg 137 179ndash203
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Turner BL Kasperson RE Matson PA McCarthy JJ Corell RW
Christensen L Eckley N Kasperson JX Luers A Martello ML Polsky C Pulsipher A Schiller A 2003 A framework for vulnerability analysis in sustainability science Proc Natl Acad Sci U S A httpsdoiorg101073pnas1231335100
Vitousek PM Aber JD Howarth RW Likens GE Matson PA
Schindler DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
Vitousek PM Howarth RW 1991 Nitrogen limitation on land and in the
sea How can it occur Biogeochemistry httpsdoiorg101007BF00002772 von Gunten L Grosjean M Rein B Urrutia R Appleby P 2009 A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 Holocene 19 873ndash881 httpsdoiorg1011770959683609336573
Xu R Tian H Pan S Dangal SRS Chen J Chang J Lu Y Maria
Skiba U Tubiello FN Zhang B 2019 Increased nitrogen enrichment and shifted patterns in the worldrsquos grassland 1860-2016 Earth Syst Sci Data httpsdoiorg105194essd-11-175-2019
Zaehle S 2013 Terrestrial nitrogen-carbon cycle interactions at the global
scale Philos Trans R Soc B Biol Sci 368 httpsdoiorg101098rstb20130125
5
A mis padres Arturo y Malena
A mis hijos Xavi y Panchito
6
AGRADECIMIENTOS
Quiero agradecer a mi tutor y mentor Dr Claudio Latorre por brindarme su
apoyo sin el cual no habriacutea logrado concluir esta tesis de doctorado Claudio tu
apoyo constante incentivo y el fijarme metas que a veces me pareciacutean imposibles
de alcanzar no solo han dado forma a esta tesis sino tambieacuten me ha hecho maacutes
exigente como cientiacutefica Claudio destacas no solo por ser un gran cientiacutefico si no
tambieacuten por tu gran calidad humana eres un gran ejemplo
Quiero agradecer Dr Blas Valero-Garceacutes por nuestras numerosas
conversaciones viacutea Skype que incluiacutean vacaciones y fines de semana para discutir
los resultados de la tesis y que han dado forma a esta investigacioacuten principalmente
al primer capiacutetulo Ademaacutes por haberme acogido como un miembro maacutes en el
laboratorio de Paleoambientes Cuaternarios durante las estancias que he realizado
en el transcurso de estos antildeos Blas eres un ejemplo para miacute conjugas ciencia de
calidad calidez y dedicacioacuten por tus estudiantes
A quienes han financiado mi doctorado la Comisioacuten Nacional de
Investigacioacuten Cientiacutefica y Tecnoloacutegica (CONICYT) con sus becas de manutencioacuten
doctoral (2013) gastos operacionales pasantiacutea (2016) postnatal (2017) y termino
de tesis doctoralrdquo (2013) A FONDECYT a traveacutes del proyecto 1160744 de C
Santoro Al Departamento de InvestigacioacutenAl Instituto de Ecologiacutea y Biodiversidad
(IEB) a traveacutes de del PIA financiamiento basal 170008 la Pontificia Universidad
Catoacutelica de Chile por la beca incentivo para tesis interdisciplinaria para doctorandos
(2015)
Agradezco a mis compantildeeros del laboratorio de Paleoecologiacutea y
Paleoclimatologiacutea Karla Matias Dani Carolina Mauricio y Pancho que han hecho
grato mi tiempo en el laboratorio Agradecimientos especiales a Carolina Matiacuteas y
Leo por acompantildearme a terreno
7
Agradezco a los miembros del laboratorio de Paleoambientes Cuaternarios
(IPE) en especial a Raquel Elena Fernando Ana Penelope Graciela Miguel
Sevilla Mariacutea y Miguel Bartolomeacute
Agradezco a los miembros de la comisioacuten Dr Joseacute Miguel Farintildea Dr Juan
Armesto y Dr Ricardo De Pol-Holz por la gran ayuda durante el desarrollo de mi
doctorado en especial por las correcciones finales de la tesis
Al departamento de Ecologiacutea en especial a Rosario Galaz Edita Luengo
Daniela Mora y Valeria Cavallero por su apoyo
A mis compantildeeros de doctorado Beleacuten Gallardo Pancho Diacuteaz Bryan Bularz
Bian Latorre Renato Borras Juan Carlos Hernaacutendez y Paulina Troncoso con
quienes estudieacute aprendiacute y compartimos muy buenos momentos durante los
primeros antildeos del doctorado
A Pablo Sarricolea mi compantildeero de vida Gracias por tu constante e
incondicional apoyo Sin ti durante este proceso todo habriacutea sido cuesta arriba
A mis padres Arturo y Malena gracias por su carintildeo apoyo y compromiso
Papaacute aunque no esteacutes a mi lado siempre te recuerdo con mucho carintildeo A mi
madre que ha sido un constante apoyo sobre todo en el cuidado de mis hijos xavi
y panchito
A mis hermanos Rodrigo y David por estar presentes durante toda esta
etapa Siempre con carintildeo y hermandad
A Rodrigo David Matiacuteas Jany Paola y Julio por cuidar a mis hijos con carintildeo
siempre que estuve ausente por el doctorado
8
ABREVIATURAS
N Nitroacutegeno (Nitrogen)
DIN Nitroacutegeno Inorgaacutenico Disuelto (Dissolved Inorganic Nitrogen)
C Carbono (Carbon)
TOC Carbono Inorgaacutenico Total (Total Organic Carbon)
TIC Carbono Inorgaacutenico Total (Total inorganic Carbon)
TC Carbono Total (Total Carbon)
TS Azufre Total (Total Sulfur)
LUCC Cambios de UsoCobertura del Suelo (Land Use and Cover Change)
OM Materia Orgaacutenica (Organic Matter)
POM Particulate Organic Matter (materia orgaacutenica particulada)
CE Common Era
BCE Before Common Era
Cal BP Calibrado en antildeos radiocarbono antes de 1950
ie id est (esto es)
e g Exempli gratia (por ejemplo)
9
RESUMEN
El ldquoAntropocenordquo se caracteriza por pertubaciones de origen humano que
conllevan impactos globales Por ejemplo los cambios de usocobertura del suelo
(LUCC) son conocidos por perturbar el ciclo del N a partir de la Revolucioacuten Industrial
pero especialmente desde la Gran Aceleracioacuten (1950 CE) en adelante Sin
embargo existen incertezas asociadas a la magnitud del impacto y su efecto
acoplado con los cambios climaacuteticos actuales (ie megasequiacutea disminucioacuten de las
precipitaciones) y acoplamientos con otros ciclos biogeoquiacutemicos (eg ciclo del
Carbono) Este impacto ha cambiado la disponibilidad de N en los ecosistemas
terrestres y acuaacuteticos y sus enlaces Los sedimentos lacustres contienen
informacioacuten de las condiciones paleoambientales del lago y su cuenca en el
momento en que se depositaron y el anaacutelisis de los isoacutetopos estables de N (δ15N)
en los sedimentos permiten reconstruir los cambios en la disponibilidad de N a
traveacutes del tiempo En este trabajo usamos una aproximacioacuten multiproxy que incluye
10
anaacutelisis sedimentoloacutegicos geoquiacutemicos e isotoacutepicos (δ15N δ13C) de los sedimentos
lacustres columna de agua y suelo-vegetacioacuten de la cuenca y la reconstruccioacuten de
los LUCC entre 1975 y 2016 a partir de imaacutegenes satelitales El objetivo de esta
tesis fue evaluar el rol de LUCC como principal control del ciclo del N en el sistema
cuenca-lago durante las uacuteltimas centurias en Chile central Nuestros principales
resultados muestran que valores maacutes positivos de δ15N en los sedimentos lacustres
estaacuten relacionados con grandes aportes de N de la cuenca que a su vez son
mayores cuanto mayor la superficie agriacutecola yo los pastizales mientras que tanto
las plantaciones forestales como los bosques favorecen la retencioacuten de nutrientes
en la cuenca (lo que se ve reflejado en un δ15N maacutes negativo) Esta tesis plantea
un cambio de estado en la dinaacutemica del N asociado a la introduccioacuten y expansioacuten
de las plantaciones forestales o bosques los que retienen el Nitroacutegeno en las
cuencas mientras que el clima juega un rol secundario
11
ABSTRACT
The Anthropocene is characterized by human disturbances at the global
scale For example changes in land use are known to disturb the N cycle since the
industrial revolution but especially since the Great Acceleration (1950 CE) onwards
This impact has changed N availability in both terrestrial and aquatic ecosystems
However there are some important uncertainties associated with the extent of this
impact and how it is coupled to ongoing climate change (ie megadroughts rainfall
variability and shifting stationarity) or to other nutrient cycles (e g carbon cycle)
Lake sediments contain paleoenvironmental information regarding the conditions of
the watershed and associated lakes and which the respective sediments are
deposited Nitrogen stable isotope analysis (δ15N) on lake sediments allows us to
reconstruct the changes in N availability through time Here we used a multiproxy
approach that uses sedimentological geochemical and isotopic analyses on
lacustrine sediments water column and soilvegetation from the watershed as well
12
as land use and cover change (LUCC) from 1975 to 2016 obtained from satellite
images The goal of this thesis is to evaluate the role of LUCC as the main driver for
N cycling in a coastal watershed system of central Chile over the last centuries Our
main results show that more positive δ15N values in lake sediments are related to
higher N contributions from the watershed which in turn increase with increased
agricultural andor pasture cover whereas either forest plantations or native forests
can favor nutrient retention in the watershed (δ15N more negative) This thesis
proposes that N dynamics are mainly driven by the introduction and expansion of
forest or tree plantations that retain nitrogen in the watershed whereas climate plays
a secondary role
13
INTRODUCCIOacuteN
El N es un elemento esencial para la vida y limita la productividad en
ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al 1997) Las actividades
humanas han tenido un profundo impacto sobre el ciclo del N global principalmente
a partir la revolucioacuten industrial (IPCC 2013 2007) Las concentraciones de N se
han incrementado por la fijacioacuten industrial de N atmosfeacuterico (viacutea el proceso Haber-
Bosch Erisman et al 2008) por el uso de fertilizantes sobre la base de N para
mejorar el rendimiento de los cultivos la produccioacuten ganadera y ademaacutes de los
cambios en el uso del suelo (abreviado en ingleacutes- LUCC) (Robertson y Vitousek
2009 Galloway et al 2008 Battye et al 2017 Xu et al 2019) Estas actividades
contribuyen significativamente al incremento de la disponibilidad bioloacutegica de N
cuyas consecuencias para los ecosistemas incluye la perdida de diversidad
modificacioacuten de la disponibilidad de nutrientes y acidificacioacuten de los suelos entre
otras Sin embargo se desconoce la cantidad destino y el alcance que ha tenido
14
el N movilizado entre los ecosistemas generado por la influencia de las actividades
humanas (Vitousek et al 1997)
La entrada natural de N en los ecosistemas terrestres y acuaacuteticos es viacutea
fijacioacuten atmosfeacuterica la que es llevada a cabo por microorganismos muchos de ellos
en relaciones simbioacuteticas (eg Rhizobium en leguminosas) y las algas (Vitousek et
al 1997 Battye et al 2017) Las principales salidas estaacuten relacionadas con la
desnitrificacioacuten volatilizacioacuten del amoniaco y por escurrimiento superficial y
subterraacuteneo (Galloway et al 2008 Diacuteaz et al 2016) En el caso de los sistemas
lacustres ademaacutes estaacuten la resuspencioacuten de N del fondo del lago los aportes desde
la cuenca (entradas de N) La depositacioacuten de MO en el fondo del lago que marca
la salida de N de la columna de agua Estas relaciones de intercambio de N tienen
un control geograacutefico (eg latitud exposicioacuten relieve etc) climaacutetico y biotico
(McLauchlan et al 2013) La alteracioacuten provocada por los LUCC no solo genera
las peacuterdidas de nutrientes del suelo por erosioacuten y escurrimiento superficial sino que
tambieacuten modifica la disponibilidad y transferencia de N entre los ecosistemas
terrestres y acuaacuteticos (Elbert et al 2012 McLauchlan et al 2013) Por ejemplo el
reemplazo de la vegetacioacuten nativa por la agricultura yo plantaciones forestales
altera el ciclo del N debido al despeje de los terrenos viacutea quema de la vegetacioacuten
pero tambieacuten debido al reemplazo de la vegetacioacuten preexistente por un
monocultivo (eg trigo pino) y el uso ineficiente de fertilizantes para mejorar el
rendimiento agriacutecola Estas actividades favorecen un incremento de los aportes de
N (y otros nutrientes) viacutea sistemas de drenajes a lagos los que actuacutean como
sumideros El incremento del N derivado de las actividades humanas tanto en los
ecosistemas terrestres como acuaacuteticos genera incertezas relacionados con la
trayectoria y la magnitud de la disponibilidad de N en ambos ecosistemas (Batye et
15
al 2017 Mclauchlan et al 2017) Por lo tanto se requiere de una liacutenea base de
largo plazo para saber coacutemo estas perturbaciones impactan la disponibilidad de N
en las uacuteltimas deacutecadassiglos en los ecosistemas lacustres y por lo tanto el alcance
real que los LUCC han tenido en el ciclo del N
Los ecosistemas mediterraacuteneos y el ciclo del N
Los ecosistemas mediterraacuteneos son especialmente sensibles a los LUCC
pues estos se encuentran sometidos a un estreacutes hiacutedrico durante largas temporadas
estivales y las precipitaciones se concentran en eventos puntuales y a veces con
altos montos pluviomeacutetricos Las precipitaciones tienen un alto potencial de arrastre
de sedimentos y nutrientes los que actuacutean como fertilizantes al llegar a los
ecosistemas lacustres A su vez un incremento en la disponibilidad de N puede
generar un desbalance con otros nutrientes (eg Fosforo) con efectos sobre la
productividad y diversidad de los lagos (Pentildeuelas et al 2012 Sardans et al 2012
McLauchlan 2013) En este sentido en Chile central se observa una disminucioacuten
de las precipitaciones especialmente fuerte en las uacuteltimas deacutecadas por lo que se ha
denoacuteminado como Megasequiacutea (Garreaud et al 2017) y el efecto de las
precipitaciones esporaacutedicas pueden generar un arrastre de nutrientes que no ha
sido evaluado
Los ecosistemas mediterraacuteneos concentran el 20 de la biodiversidad global
(Myers et al 2000) pero existe una escasez de conocimiento respecto a los
efectos del incremento de N en los cuerpos de agua como consecuencia de las
actividades humanas en el pasado histoacuterico (Ochoa-Hueso et al 2011) y coacutemo la
disponibilidad de N ha variado en el tiempo Los LUCC han aumentado el flujo de
N en las cuencas hidrograacuteficas viacutea escurrimiento superficial y lixiviacioacuten
16
favoreciendo la peacuterdida de sedimentos y nutrientes del suelo en el mundo entero
(Elser et al 2011 McLauchlan et al 2013 Xu et al 2017 y 2019) Ello ha
contribuido a la acidificacioacuten y eutrofizacioacuten generalizada de riacuteos y lagos
(McLauchlan et al 2013 Schindler et al 2008)
El ecosistema mediterraacuteneo de Chile central es una regioacuten ampliamente
intervenida desde al menos la colonizacioacuten espantildeola (1600-1800 CE) Los LUCC
han tenido efectos negativos en la disponibilidad de agua especialmente
observados con el desarrollo de las actividades minera (Prieto et al 2019) Aunque
se sabe de los impactos en los ecosistemas de bosque en teacuterminos de cobertura
debidos al desarrollo silvo-agropecuarias (Armesto et al 2010) se desconoce el
impacto de estas actividades sobre el ciclo del N en los cuerpos de agua o en queacute
momento cobran fuerza suficiente para alterar el flujo de N hacia los lagos de Chile
Central Por lo tanto en esta tesis nos centramos en entender coacutemo los LUCC han
afectado la disponibilidad de N en los lagos costeros de Laguna Matanzas y Lago
Vichuqueacuten desde la llegada de los espantildeoles a Chile hasta el presente
Los lagos como sensores ambientales
Los sedimentos lacustres son buenos sensores de cambios en los aportes
de nutrientes contaminacioacuten y eutroficacioacuten de los cuerpos de agua ya que son
capaces de preservar sentildeales quiacutemicas y fiacutesicas al momento de su formacioacuten y
ademaacutes con alta resolucioacuten y a largo plazo (Leng et al 2006) Por lo tanto
constituyen una herramienta uacutetil para reconstruir el flujo de N entre ecosistemas
terrestres y acuaacuteticos en el pasado sus consecuencias (eg incremento de la
productividad lacustre) y el rol del clima y los LUCC en el pasado (McLauchlan et
al 2013 von Gunten et al 2009) Por ejemplo el cultivo de maiacutez por parte de los
17
nativos iroqueses en Canadaacute provocoacute la eutrofizacioacuten del Lago Crawford (43degN)
durante los siglos XII y XV Entre las consecuencias registradas se documentoacute un
claro incremento de la productividad primaria y cambios en la estructura comunitaria
de diatomeas foacutesiles (incremento de Stephanodiscus complex y disminucioacuten de
Cyclotella bodanica en Ekdahl et al 2007) Otro ejemplo demuestra como las
actividades agriacutecolas en la cuenca del riacuteo Mississippi modificaron los patrones de
sedimentacioacuten y aporte de nutrientes al lago Pepin EE UU (44degN) a partir del
asentamiento europeo en 1840 CE y principalmente desde 1940 CE (Engstrom et
al 2009) Para Chile von Gunten et al (2009) a partir de indicadores
limnogeoloacutegicos evidencioacute la contaminacioacuten y eutroficacioacuten de cinco lagos cercanos
a la ciudad de Santiago (33ordmS) como consecuencia de la deposicioacuten atmosfeacuterica
de metales pesados (especialmente Cu) quema de combustible foacutesil y flujo de
nutrientes durante los uacuteltimos 200 antildeos
Caracteriacutesticas limnoloacutegicas de los lagos
Los procesos limnoloacutegicos afectan la distribucioacuten y abundancia de los
organismos en los lagos Estaacuten influenciados por forzamientos externos por
ejemplo la precipitacioacuten o la penetracioacuten de la luz y la morfometriacutea del lago En este
sentido el nivel del lago dependeraacute del balance entre las entradas y salidas de agua
(precipitaciones escurrimiento superficial y subterraacuteneo y la evaporacioacuten) y la forma
de la cuenca (profundidad pendiente aacuterea del espejo de agua)
En funcioacuten de la profundidad de penetracioacuten de la luz es posible diferenciar
dos aacutereas que determinan la distribucioacuten de los organismos la zona foacutetica (donde
penetra la luz y permite la fotosiacutentesis) y la zona afoacutetica (subyacente a la zona
foacutetica) Al depender de la radiacioacuten la zona foacutetica variacutea estacionalmente Ademaacutes
18
puede variar por el grado de turbidez la morfometriacutea del lago y la produccioacuten de
materia orgaacutenica en la columna de agua
Otro factor que influye en la productividad es el reacutegimen de mezcla de la
columna de agua pues favorece la oxigenacioacuten y el reciclado de nutrientes La
mezcla ocurre en condiciones de gran inestabilidad usualmente relacionado con el
reacutegimen de viento Por el contrario un lago estratificado resulta de grandes
diferencias en temperatura o salinidad entre la superficie (epilimnion) y el fondo del
lago (hipolimnion) que separa las masas de agua superficial y de fondo por una
termoclina (si la estratificacioacuten estaacute relacionada con diferencias de temperatura de
las masas de agua) o haloclina (diferencias de salinidad) En funcioacuten del reacutegimen
de mezcla los lagos se pueden clasificar en (Lewis 1983)
1 Amiacutecticos no hay mezcla vertical de la columna de agua
2 Monomiacutecticos friacuteos y caacutelidos se mezcla una vez al antildeo
3 Dimiacutecticos la columna de agua se mezcla dos veces al antildeo
4 Oligomiacutecticos Lagos estratificados la mayor parte del tiempo pero se mezclan a
intervalos irregulares mayores a 1 antildeo
5 Polimiacutecticos friacuteos y caacutelidos la columna de agua se mezcla varias veces al antildeo
El ciclo del N en lagos
Al estar presente en aacutecidos nucleicos aminoaacutecidos y proteiacutenas el N es un
nutriente esencial de los organismos Por lo tanto su disponibilidad en la columna
de agua es clave para la produccioacuten composicioacuten y acumulacioacuten de la MO a traveacutes
del tiempo (Olson 1998 Talbot 2002) Aunque el principal reservorio de N estaacute en
19
la atmosfera (N2) solo unos pocos organismos (eg cianobacterias) pueden fijarlo
directamente Asiacute el Nitroacutegeno Inorgaacutenico Disuelto (DIN) representa la principal
fuente de N disponible para la produccioacuten primaria en los ecosistemas acuaacuteticos
(Talbot et al 2002) El DIN estaacute compuesto por nitrato (NO3-) nitrito (N02
-) y amonio
(NH4+) y los cambios en su disponibilidad condicionan el tipo de produccioacuten primaria
(eg fijadores vs asimiladores de N ver Scott y Grant 2013 Scott 2008)
La Figura 1 resume los principales componentes en lagos del ciclo del N y
sus interacciones La principal entrada natural de N es viacutea fijacioacuten de N atmosfeacuterico
y es llevado a cabo principalmente por bacterias y algas verde-azules capaces de
romper el triple enlace entre los dos aacutetomos de N a un alto costo energeacutetico (Torres
et al 2012) El N fijado es reducido a NH3 Tras la muerte de los organismos el N
es liberado de sus tejidos por hongos y bacterias implicados en la descomposicioacuten
de la MO Una parte se deposita en el fondo del lago y la otra queda disponible para
ser asimilada por el fitoplancton como amonio mediante el proceso de
amonificacioacuten (Talbot 2002) La amonificacioacuten resulta de la reduccioacuten bacteriana
del amoniaco y se da preferentemente en condiciones anoacutexicas La volatizacioacuten del
amoniaco es un proceso por medio este se incorpora a la atmoacutesfera Este proceso
se da preferentemente en lagos aacutecidos (pH gt 85) El nitrato es otra forma de N
bioloacutegicamente disponible para el fitoplancton En los lagos el NO3 del DIN estaacute
compuesto por los aportes desde la cuenca hidrograacutefica en la que se circunscriben
por la resuspensioacuten desde el fondo del lago favorecido por procesos de mezcla
(descrito en seccioacuten anterior) y por los nitratos de la columna de agua Estos uacuteltimos
son el resultado de la nitrificacioacuten proceso realizado por bacterias aeroacutebicas
mediante el cual el amonio se oxida para formar nitrito y nitrato El ciclo se completa
20
con la desnitrificacioacuten que es la reduccioacuten bacteriana de nitrato al N atmosfeacuterico
Este proceso se da preferentemente en condiciones anoacutexicas
Figura 1 Materia Orgaacutenica sedimentaria y su relacioacuten con el ciclo del N y las
variaciones de δ15N (elaborado a partir de Talbot et al 2002) En la figura se
representan los factores clave en la acumulacioacuten de la MO sedimentaria y su
relacioacuten con el ciclo del N El δ15N de la MO es resultado de los aportes de MO
desde la cuenca MO de macroacutefitas y de la productividad bioloacutegica La productividad
en ecosistemas lacustres estaacute condicionada por DIN y la fijacioacuten de N atmosfeacuterico
El δ15N del pool del DIN disponible se va enriqueciendo por la asimilacioacuten
preferencial de 14N por parte del fitoplancton Ademaacutes el DIN remanente se va
enriqueciendo por accioacuten microbiana (amonificacioacuten nitrificacioacuten desnitrificacioacuten)
Reconstruyendo el ciclo del N a partir de variaciones en δ15N
La sentildeal isotoacutepica de N (δ15N) en los sedimentos lacustres puede ser usada
para reconstruir los cambios pasados del ciclo N la transferencia de N entre
ecosistemas terrestres y acuaacuteticos y el estado troacutefico del lago (Bruland y Mackenzie
2015 McLauchlan et al 2013 Woodward et al 2012 von Gunten et al 2009
Leng et al 2006 Meyers y Teranes 2001) La Figura 1 resume los principales
procesos y fuentes que afectan la sentildeal isotoacutepica δ15N (total o ldquobulkrdquo) en la MO de
21
los sedimentos lacustres Entre los que destacan las fuentes de MO (aloacutectona vs
autoacutectona) la bioproductividad lacustre y las entradas de N (ie fijacioacuten bioloacutegica
de N2) (Torres et al 2012) Estos procesos conducen a un fraccionamiento
isotoacutepico cineacutetico que favorece la disociacioacuten entre sus isotopos ldquopesadosrdquo (15N) y
ldquoligerosrdquo (14N) Asiacute por ejemplo los valores de δ15N de la MO se enriquecen en 15N
en la medida que los sedimentos lacustres estaacuten sometidos a perdidas de 14N viacutea
desnitrificacioacuten (Bruland y Mackenzie 2015 Diaz et al 2016 Talbot et al 2002)
Por otra parte el fitoplancton en la medida que aumenta su biomasa (eg
durante la estacioacuten caacutelida) puede llegar a asimilar el DIN hasta agotarse En este
caso el N remanente se va empobreciendo en 14N si no hay reposicioacuten de N (eg
aporte de NO3 de la cuenca) y la MO queda enriquecida en 15N Esta disminucioacuten
induce a una limitacioacuten de fitoplancton por N y los organismos diztroacuteficos pueden
verse estimulados por lo que se incrementa la entrada de N viacutea fijacioacuten de N2 (Scott
y Grantsz 2013) como resultado la MO oscila en valores maacutes bajos (en torno a 0permil)
La cantidad de MO que se deposita en el fondo del lago depende del
predominio de contribuciones provenientes desde la cuenca (aloacutectona) versus las
producidas dentro del mismo lago (autoacutectona) (Torres et al 2012) Aunque en
general los lagos reciben permanentemente aportes de sedimentos y MO desde
su cuenca los lagos pequentildeos tienden a reflejar maacutes los procesos que ocurren
solamente en su cuenca directa y reflejan los valores isotoacutepicos de la misma (Gu et
al 2006) Woodward et al (2012) realizaron un estudio en 50 lagos en Irlanda que
les permitioacute correlacionar diferentes patrones de LUCC (bosques de coniacuteferas
agricultura ganaderiacutea entre otros) con los valores de δ15N (y δ13C) en los
sedimentos superficiales de los lagos Encontraron que los valores de δ15N son maacutes
negativos en cuencas poco impactadas y maacutes positivos en aquellas con alto
22
impacto humano (eg usos de suelo agriacutecola y ganadero) McLauchlan et al (2007)
encontraron valores maacutes positivos de δ15N en los sedimentos del Mirror Lake New
Hampshire (29ordm N) a partir de la desforestacioacuten de su cuenca en 1790 CE (inicio
del asentamiento europeo en el lago) para el desarrollo de la actividad agriacutecola
Estos valores se volvieron maacutes negativos hacia valores similares al pre-
asentamiento europeo despueacutes del cese de la agricultura en la cuenca y la
recuperacioacuten del bosque a partir de 1929 CE
El anaacutelisis de δ15N en los sedimentos lacustres es una herramienta casi sin
explorar en Chile central Proponemos reconstruir la dinaacutemica de transferencia de
N cuenca-lago como consecuencia de los LUCC desde la colonizacioacuten espantildeola en
los lagos costeros de Laguna Matanzas (34ordmS) y Lago Vichuqueacuten (36ordmS) Como son
muacuteltiples los procesos que pueden afectar la sentildeal isotoacutepica de N (medido como
δ15N) en los sedimentos lacustres existen muchos problemas para su
interpretacioacuten paleoambiental (Leng et al 2006) Por ello en esta tesis utilizamos
un enfoque multiproxy que consiste en integrar el anaacutelisis limnoloacutegico y geoquiacutemico
de los sedimentos lacustres con los valores isotoacutepicos modernos de la columna de
agua (mediante monitoreo del POM) la relacioacuten vegetacioacuten-suelo de la cuenca y la
reconstruccioacuten de los principales usos de suelo de las cuencas desde 1975 CE
mediante el uso de imaacutegenes satelitales Considerando estas muacuteltiples liacuteneas de
evidencia nos planteamos utilizar las variaciones del δ15N para reconstruir los
cambios en la disponibilidad de N en los lagos costeros de Chile central y establecer
coacutemo y desde cuaacutendo las actividades humanas en la cuenca son lo suficientemente
importantes como para modificar el ciclo del N en los lagos desde la colonizacioacuten
espantildeola (siglo XVII)
23
Como hipoacutetesis planteamos que la dinaacutemica de transferencia de sedimentos
y nutrientes entre la cuenca y el lago tiene controles geoloacutegicos climaacuteticos y
bioloacutegicos que interactuacutean en el tiempo registraacutendose en la acumulacioacuten de MO de
los sedimentos lacustres (Fig1) Bajo este marco los LUCC modifican esta
dinaacutemica alterando la transferencia de N a los lagos ya que las actividades agriacutecolas
y ganaderas tienden a aportar maacutes N (aumentando δ15N) que la actividad forestal
(la que disminuye δ15N)
En el primer capiacutetulo titulado ldquoA combined approach to establishing the timing
and magnitude of anthropogenic nutrient alteration in a mediterranean coastal lake-
watershed systemrdquo se evaluoacute el rol de los LUCC en el control de la descarga de N
y sedimentos a Laguna Matanzas en un sistema cuenca-lago desde el siglo XVII
Para ello se realizoacute un anaacutelisis de la dinaacutemica sedimentaria lacustre mediante el
anaacutelisis de muacuteltiples proxys sedimentoloacutegicos (ie minerales y textura)
geoquiacutemicos (eg geoquiacutemica orgaacutenica e isotoacutepica) y bioloacutegicos (ie diatomeas) de
Laguna Matanzas que cubren los uacuteltimos 200 antildeos Ademaacutes se realizoacute una
reconstruccioacuten de los usos de suelo entre 1975 y 2016 a partir de imaacutegenes de
sateacutelites y se colectaron muestras de suelo de las principales coberturas de la
cuenca a los cuales se midioacute el δ15N
Entre los principales resultados obtenidos se destaca la influencia de la
ganaderiacutea y la denitrificacioacuten lo que explica los altos valores de δ15N evidenciados
por el registro lacustre durante la colonizacioacuten espantildeola e inicios de la repuacuteblica A
partir de la Gran Aceleracioacuten (1950 CE) la desforestacioacuten y el reemplazo de la
ganaderiacutea por plantaciones forestales tienen un correlato en el registro
sedimentario con valores de δ15N maacutes bajos Estos resultados demuestran que los
LUCC son el factor de primer orden para explicar los cambios observados en
24
nuestro registro de δ15N y a su vez implica un rol secundario del clima como posible
control de la disponibilidad de N en Laguna Matanzas Este capiacutetulo fue sometido
a la revista Scientific Reports y estaacute en revisioacuten desde el 26 de marzo de 2019 En
la elaboracioacuten del mauscrito han colaborado Claudio Latorre Blas Valero-Garceacutes
Matiacuteas Frugone Pablo Sarricolea Santiago Giralt Manuel Contreras-Lopez
Ricardo Prego y Patricia Bernardez
El segundo capiacutetulo se titula ldquoStable isotopes track land use and cover
changes in a mediterranean lake in central Chile over the last 600 yearsrdquo Aquiacute
evaluamos la dinaacutemica del N actual en el sistema cuenca-lago considerando los
valores isotoacutepicos modernos tanto de la vegetacioacuten como del suelo ademaacutes de los
cambios estacionales del POM de la columna de agua Esta informacioacuten se utiliza
como un anaacutelogo moderno para conocer el rol de los LUCC en la disponibilidad de
N en los uacuteltimos 600 antildeos Para ello se colectaron muestras filtradas de la columna
de agua a 2 5 y 20 m dos veces por antildeo entre noviembre de 2015 y julio de 2018
y se obtuvo el δsup1⁵N del POM Ademaacutes se colectoacute muestras de vegetacioacuten y suelo
de la cuenca diferenciando entre especies nativas plantaciones forestales y
vegetacioacuten herbaacutecea Esta informacioacuten se analizoacute en conjunto con la reconstruccioacuten
de los LUCC entre 1975 y 2014 y el anaacutelisis limnoloacutegico del lago Ello nos permitioacute
evaluar coacutemo el δ15N de los sedimentos lacustres refleja los valores δ15N de la
cuenca en el Lago Vichuqueacuten y el control que ejerce el clima en la sentildeal isotoacutepica
de POM Este capiacutetulo seraacute sometido a la revista Science of total environmet
proximamente En la elaboracioacuten del mauscrito han colaborado Claudio Latorre
Blas Valero-Garceacutes Matiacuteas Frugone Pablo Sarricolea Leonardo Villacis y M Laura
Carrevedo
25
Entre los principales resultados encontramos que el δ15N en los sedimentos
lacustres es maacutes positivo con presencia de agricultura en la cuenca y maacutes negativo
cuando esta es reemplazada por cobertura arboacuterea bien sea plantaciones
forestales bien sea bosque nativo A su vez durante la estacioacuten seca y caacutelida la
mayor entrada al lago es viacutea fijacioacuten de N atmosfeacuterico (bajos valores de δ15NPOM)
Durante la estacioacuten huacutemeda en cambio prevalecen los aportes de la cuenca con
altos valores de δ15NPOM Ello estariacutea evidenciando un cambio estacional en la
composicioacuten de especies en verano ligado a especies fijadoras de N y en invierno
las algas y microorganismos que consumen el DIN de la columna de agua
Referencias
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea AM Marquet PA 2010 From the Holocene to the Anthropocene A historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the
next carbon Earth rsquo s Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409 httpsdoiorg102134jeq20090005
Diacuteaz FP Frugone M Gutieacuterrez RA Latorre C 2016 Nitrogen cycling in an
extreme hyperarid environment inferred from δ15 N analyses of plants soils and herbivore diet Sci Rep 6 1ndash11 httpsdoiorg101038srep22226
Ekdahl EJ Teranes JL Wittkop CA Stoermer EF Reavie ED Smol JP
2007 Diatom assemblage response to Iroquoian and Euro-Canadian eutrophication of Crawford Lake Ontario Canada J Paleolimnol 37 233ndash246 httpsdoiorg101007s10933-006-9016-7
Elbert W Weber B Burrows S Steinkamp J Buumldel B Andreae MO
Poumlschl U 2012 Contribution of cryptogamic covers to the global cycles of carbon and nitrogen Nat Geosci 5 459ndash462
26
httpsdoiorg101038ngeo1486 Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506
httpsdoiorg101126science1215567 ARTICLE Engstrom DR Almendinger JE Wolin JA 2009 Historical changes in
sediment and phosphorus loading to the upper Mississippi River Mass-balance reconstructions from the sediments of Lake Pepin J Paleolimnol 41 563ndash588 httpsdoiorg101007s10933-008-9292-5
Erisman J W Sutton M A Galloway J Klimont Z amp Winiwarter W (2008)
How a century of ammonia synthesis changed the world Nature Geoscience 1(10) 636
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892
httpsdoiorg101126science1136674 Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J Christie
D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592 Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
IPCC 2013 Climate Change 2013 The Physical Science BasisContribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change Cambridge University Press Cambridge United Kingdom and New York NY USA
httpsdoiorg101017CBO9781107415324 Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007 Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32iumliquestfrac12S 3050iumliquestfrac12miumliquestfrac12asl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470
27
httpsdoiorg101073pnas0701779104 McLauchlan KK Gerhart LM Battles JJ Craine JM Elmore AJ Higuera
PE Mack MC McNeil BE Nelson DM Pederson N Perakis SS 2017 Centennial-scale reductions in nitrogen availability in temperate forests of the United States Sci Rep 7 1ndash7 httpsdoiorg101038s41598-017-08170-z
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Myers N Mittermeler RA Mittermeler CG Da Fonseca GAB Kent J
2000 Biodiversity hotspots for conservation priorities Nature httpsdoiorg10103835002501
Pentildeuelas J Poulter B Sardans J Ciais P Van Der Velde M Bopp L
Boucher O Godderis Y Hinsinger P Llusia J Nardin E Vicca S Obersteiner M Janssens IA 2013 Human-induced nitrogen-phosphorus imbalances alter natural and managed ecosystems across the globe Nat Commun 4 httpsdoiorg101038ncomms3934
Prieto M Salazar D Valenzuela MJ 2019 The dispossession of the San
Pedro de Inacaliri river Political Ecology extractivism and archaeology Extr Ind Soc httpsdoiorg101016jexis201902004
Robertson GP Vitousek P 2010 Nitrogen in Agriculture Balancing the Cost of
an Essential Resource Ssrn 34 97ndash125 httpsdoiorg101146annurevenviron032108105046
Sardans J Rivas-Ubach A Pentildeuelas J 2012 The CNP stoichiometry of
organisms and ecosystems in a changing world A review and perspectives Perspect Plant Ecol Evol Syst 14 33ndash47 httpsdoiorg101016jppees201108002
Schindler DW Howarth RW Vitousek PM Tilman DG Schlesinger WH
Matson PA Likens GE Aber JD 2007 Human Alteration of the Global Nitrogen Cycle Sources and Consequences Ecol Appl 7 737ndash750 httpsdoiorg1018901051-0761(1997)007[0737haotgn]20co2
Scott E and EM Grantzs 2013 N2 fixation exceeds internal nitrogen loading as
a phytoplankton nutrient source in perpetually nitrogen-limited reservoirs Freshwater Science 2013 32(3)849ndash861 10189912-1901
28
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Talbot M R (2002) Nitrogen isotopes in palaeolimnology In Tracking
environmental change using lake sediments (pp 401-439) Springer Dordrecht
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable
isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K
Yeager KM 2016 Different responses of sedimentary δ15N to climatic changes and anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
29
CAPIacuteTULO 1 A COMBINED APPROACH TO ESTABLISHING THE TIMING
AND MAGNITUDE OF ANTHROPOGENIC NUTRIENT ALTERATION IN A
MEDITERRANEAN COASTAL LAKE- WATERSHED SYSTEM
30
A combined approach to establishing the timing and magnitude of anthropogenic
nutrient alteration in a mediterranean coastal lake- watershed system
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugoneabc Pablo
Sarricolead Santiago Giralte Manuel Contreras-Lopezf Ricardo Prego g Patricia
Bernaacuterdez g Blas Valero-Garceacutesch
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Institute of Earth Sciences Jaume Almera (ICTJA-CSIC) CLluis Solegrave Sabaris sn Barcelona E-
08028 Spain
f Facultad de Ingenieriacutea y Centro de Estudios Avanzados Universidad de Playa Ancha Traslavintildea
450 Vintildea del Mar Chile
g Instituto de Investigaciones Marinas (CSIC) CEduardo Cabello 6 36208 Vigo Spain
h Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding author
E-mail address
clatorrebiopuccl magdalenafuentealbagmailcom
Abstract
Since the industrial revolution and especially during the Great Acceleration (1950
CE) human activities have profoundly altered the global nutrient cycle through land
use and cover changes (LUCC) However the timing and intensity of recent N
variability together with the extent of its impact in terrestrial and aquatic ecosystems
and coupled effects of regional LUCC and climate are not well understood Here
we used a multiproxy approach (sedimentological geochemical and isotopic
31
analyses historical records climate data and satellite images) to evaluate the role
of LUCC as the main control for N cycling in a coastal watershed system of central
Chile during the last few centuries The largest changes in N dynamics occurred in
the mid-1970s associated with the replacement of native forests and grasslands for
livestock grazing by plantations of introduced Monterey pine (Pinus radiata) and
eucalyptus (Eucalyptus globulus) Lake productivity increased and is evidenced by
an increase trend in δ15N values Our study shows that anthropogenic land
usecover changes are key in controlling nutrient supply and N availability in
Mediterranean watershed ndash lake systems and that large-scale forestry
developments during the mid-1970s likely caused the largest changes in central
Chile
Keywords Anthropocene Organic geochemistry watershedndashlake system Stable
Isotope Analyses Land usecover change Nitrogen cycle Mediterranean
ecosystems central Chile
1 INTRODUCTION
Human activities have become the most important driver of the nutrient cycles in
terrestrial and aquatic ecosystems since the industrial revolution (Gruber and
Galloway 2008 Galloway et al 2008 Holtgrieve et al 2011 Fowler et al 2013
Goyette et al 2016) Among these N is a common nutrient that limits productivity
in terrestrial and aquatic ecosystems (Vitousek and Howarth 1991 McLauchlan et
al 2013) With the advent of the Haber-Bosch industrial N fixation process in the
early 20th century total N fluxes have surpassed previous planetary boundaries
32
(Galloway et al 2008 Elser 2011) reaching unprecedented values (ie tipping
points) in the Earth system especially during what is now termed the Great
Acceleration (which began in the 1950s) which intensified in the 1970s (Howarth
2004 Steffen et al 2015) Yet dramatic changes in LUCC have occurred in the last
few centuries mainly linked to forestry and agro-pastoral activities (Poraj-Goacuterska et
al 2017 Ge et al 2019 Cheng et al 2019) These have had large impacts on N
(and Carbon -C-) availability in terrestrial and aquatic ecosystems but the synergic
effect with climate change and global N dynamics has not been established
(Vitousek et al 1997 Mclauchlan et al 2007 2013 Pongratz et al 2010
Woodward et al 2012 Mclauchlan et al 2017)
The onset of the Anthropocene poses significant challenges in mediterranean
regions that have a strong seasonality of hydrological regimes and an annual water
deficit (Stocker et al 2013) Mediterranean climates occur in all continents
(California central Chile Australia South Africa circum-Mediterranean regions)
providing a unique opportunity to investigate global change processes during the
Anthropocene in similar climate settings but with variable geographic and cultural
contexts The effects of global change in mediterranean watersheds have been
analyzed from different perspectives hydrology (Giorgi 2006 Arnell and Gosling
2013 Prudhomme et al 2014) vegetation dynamics (Lenihan et al 2003 Vicente-
Serrano et al 2013 Matesanz and Valladares 2014) sediment dynamics (Garciacutea-
Ruiz et al 2013 Garciacutea-Ruiz 2010 Syvitski et al 2005) changes in
biogeochemical cycles (Thomas et al 2013 Fowler et al 2013 Frank et al 2015)
carbon storage (Muntildeoz-Rojas et al 2015) and biodiversity (Hooper et al 2012) A
recent review showed an extraordinarily high variability of erosion rates in
mediterranean watersheds positive relationships with slope and annual
33
precipitation and the paramount effect of land use (Garciacutea-Ruiz et al 2015)
However the temporal context and effect of LUCC on nutrient supply to
mediterranean lakes has not been analyzed in much detail
Major LUCC in central Chile occurred during the Spanish Colonial period
(17th-18th centuries) (Armesto 1994 Gurnell et al 2001 Figueroa et al 2004
Martiacuten-Foreacutes et al 2012 Contreras-Loacutepez et al 2014) with the onset of
industrialization and mostly during the mid to late 20th century (von Gunten et al
2009 Gayo et al 2012) Recent copper pollution caused by 20th century mining
and industrial smelters has been documented in cores throughout the Andes
(Laguna el Ocho and Laguna Ensuentildeo von Gunten et al 2009) and also from our
surveys in the central valley (Batuco wetland) coastal range (Cordillera de Name)
and along the coast itself (Bucalemito and Colejuda) (Valero-Garceacutes et al 2010
unpublished data)
Paleolimnological studies have shown how these systems respond to
climate LUCC and anthropogenic impacts during the last millennia (Jenny et al
2002 2003 Villa Martiacutenez et al 2002 Frugone-Aacutelvarez et al 2017 Contreras et
al 2018) Furthermore changes in sediment and nutrient cycles have also been
identified in associated terrestrial ecosystems dating as far back as the Spanish
Conquest and related to fire clearance and wood extraction practices of the native
forests (Armesto 1994 Contreras-Loacutepez et al 2014) Nevertheless pollen and
limnological evidence argue for a more recent timing of the largest anthropogenic
impacts in central Chile For example paleo records show that during the mid-20th
century increased soil erosion followed replacement of native forest by Pinus
radiata and Eucalyptus globulus plantations at Laguna Matanzas Aculeo and
34
Vichuqueacuten lakes (Jenny et al 2002 Villa-Martiacutenez et al 2002 2003 Frugone-
Aacutelvarez et al 2017)
Lakes are a central component of the global carbon cycle Lakes act as a
sink of the carbon cycle both by mineralizing terrestrially derived organic matter and
by storing substantial amounts of organic carbon (OC) in their sediments (Anderson
et al 2009) Paleolimnological studies have shown a large increase in OC burial
rates during the last century (Heathcote et al 2015) however the rates and
controls on OC burial by lakes remain uncertain as do the possible effects of future
global change and the coupled effect with the N cycle LUCC intensification of
agriculture and associated nutrient loading together with atmospheric N-deposition
are expected to enhance OC sequestration by lakes Climate change has been
mainly responsible for the increased algal productivity since the end of the 19th
century and during the late 20th century in lakes from both the northern (Ruumlhland et
al2015) and southern hemispheres (Michelutti et al 2015 Carrevedo et al 2015)
but many studies suggest a complex interaction of global warming and
anthropogenic influences and it remains to be proven if climate is indeed the only
factor controlling these transitions (Catalaacuten et al 2013) Alternative causes for
recent N (Galloway et al 2008) increases in high altitude lakes such as catchment
mediated processes cannot be ruled out (Camarero and Catalan 2012 Fritz and
Anderson 2013) Few lake-watershed systems have robust enough chronologies of
recent changes to compare variations in C and N with regional and local processes
and even fewer of these are from the southern hemisphere (McLauchlan et al
2007 Holtgrieve et al 2011)
In this paper we present a multiproxy lake-watershed study including N and
C stable isotope analyses on a series of short cores from Laguna Matanzas in
35
central Chile focused in the last 200 years We complemented our record with land
use surveys satellite and aerial photograph studies Our major objectives are 1) to
reconstruct the dynamics among climate human activities and changes in the N
cycle over the last two centuries 2) to evaluate how human activities have altered
the N cycle during the Great Acceleration (since the mid-20th century)
2 STUDY SITE
Laguna Matanzas (33ordm45rsquoS 71ordm 40acuteW 7 m asl Fig 1) is a coastal lake located
in central Chile near to a large populated area (Santiago gt6106 inhabitants) The
lake has a surface area of 15 km2 with a max depth of 3 m and a watershed of 30
km2 The lake basin is emplaced over Pleistocene-Holocene aeolian and alluvial fan
deposits (SERNAGEOMIN 2003) A recent phase of dune activity occurred from the
mid to late Holocene which mostly sealed off the basin from the ocean (Villa-
Martinez 2002) Climate in Laguna Matanzas is characterized by cool-wet winters
and hot-dry summers with annual precipitation of ~510 mm and a mean annual
temperature of 12ordmC Central Chile is in the transition zone between the southern
hemisphere mid-latitude westerlies belt and the South Pacific Anticyclone (or SPA)
(Garreaud et al 2009) In winter precipitation is modulated by the north-west
displacement of the SPA the northward shift of the westerlies wind belt and an
increased frequency of storm fronts stemming off the Southern Hemisphere
Westerly Winds (SWW) (Frugone-Aacutelvarez et al 2017) Austral summers are
typically dry and warm as a strong SPA blocks the northward migration of storm
tracks stemming off the SWW
36
Historic land cover changes started after the Spanish conquest with a Jesuit
settlement in 1627 CE near El Convento village and the development of a livestock
ranch that included the Matanzas watershed After the Jesuits were expelled from
South America in 1778 CE the farm was bought by Pedro Balmaceda and had more
than 40000 head of cattle around 1800 CE (Contreras-Lopez et al 2014) The first
Pinus radiata and Eucalyptus globulus trees were planted during the second half of
the 19th century and were mostly used for dune stabilization (Albert 1900 Gibson
1972) However the main plantation phase occurred 60 years ago (Villa-Martinez
2002) as a response to the application of Chilean Forestry Laws promulgated in
1931 and 1974 and associated state subsidies
Major land cover changes occurred recently from 1975 to 2008 as shrublands
were replaced by more intensive land uses practices such as farmland and tree
plantations (Schulz et al 2010) Laguna Matanzas is part of the Reserva Nacional
Humedal El Yali protected area (Ramsar Ndeg 878) Despite this protected status the
lake and its watershed have been heavily affected by intense agricultural and
farming activities during the last decades The main inlet ldquoLas Rosasrdquo has been
diverted for crop irrigation causing a significant loss of water input to the lake
Consequently the flooded area of the lake has greatly decreased in the last couple
of decades (Fig 1b) Exotic tree species cover a large surface area of the
watershed Recently other activities such as farms for intensive chicken production
have been emplaced in the watershed
37
Figure 1 Laguna Matanzas a) Digital Elevation Model showing the watershed and
the surface hydrologic connection with Colejuda and Cabildo lakes b) Climograph
depicting the warm dry season in austral summer c) Annual precipitation from
1965-2015 (arrow shows start of the most recent ldquomega-droughtrdquo- see Garreaud et
al 2017) d) A decline of lake area occurs between 2007 and 2018 Lake surface
area decreased first along the western sector (in 2007) followed by more inland
areas (in 2018)
38
3 RESULTS
31 Age Model
The age model for the Matanzas sequence was developed using Bacon software
to establish the deposition rates and age uncertainty (Blaauw and Christen 2011)
It is based on 210Pb and 137Cs dates and two 14C dates (Table 2) According to this
age model the lake sequence spans the last 1000 years (Fig 2) A major breccia
layer (unit 3b) was deposited during the early 18th century which agrees with
historic documents indicating that a tsunami impacted Laguna Matanzas and its
watershed in 1730 CE (Contreras-Loacutepez et al 2014) Here we focus in the last 200
years were the most important changes occurred in terms of LUCC (after the
sedimentary hiatus caused by the tsunami) The Spanish colonial period (17th -18th
century) brought new forms of territorial management along with an intensification
of watershed use which remained relatively unchanged until the 1900s
39
Figure 2 a) Age-depth model obtained for the Laguna de Matanzas sedimentary
sequence A hiatus occurs at 80 cm depth (unit 3b) The section of core used for our
analysis is highlighted in a red rectangle b) Close up of the age model used for
analysis of recent anthropogenic influences on the N cycle c) Information regarding
the 14C dates used to construct age model
Lab code Sample ID
Depth (cm) Material Fraction of modern C
Radiocarbon age
Pmc Error BP Error
D-AMS 021579
MAT11-6A 104-105 Bulk Sediment
8843 041 988 37
D-AMS 001132
MAT11-6A 1345-1355
Bulk Sediment
8482 024 1268 21
POZ-57285
MAT13-12 DIC Water column 10454 035 Modern
Table 2 Laguna Matanzas radiocarbon dates
32 The sediment sequence
Laguna Matanzas sediments consist of massive to banded mud with some silt
intercalations They are composed of silicate minerals (plagioclase quartz and clay
minerals) with relatively high TOC content (Fig 3) Pyrite is a common mineral
indicating dominant anoxic conditions in the lake sediments whereas aragonite
occurs only in the uppermost section Mineralogical analyses visual descriptions
texture and geochemical composition were used to characterize five main facies
(Fig 3) F1 (organic-rich mud) represents baseline sedimentation in a shallow well-
mixed brackish highly productive lake and F1acute (dark orange) is a less organic facies
than F1 (more details see table in the supplementary material) F2 (massive to
banded silty mud) indicates periods of higher clastic input into the lake but finer
(mostly clay minerals) likely from suspension deposition associated with flooding
40
events Aragonite (up to 15 ) occurs in both facies but only in samples from the
uppermost 30 cm and it is interpreted as endogenic linked to higher MgFe waters
and elevated biologic productivity
Figure 3 Sedimentary facies and units mineralogy grain size elemental geochemical
and isotopic composition of Laguna Matanzas core Values highlighted in gray indicate
that these are above average
The banded to laminated fining upward silty clay layers (F3) reflect
deposition by high energy turbidity currents The presence of aragonite suggests
that littoral sediments were incorporated by these currents Non-graded laminated
coarse silt layers (F4) do not have aragonite indicating a dominant watershed
41
sediment source Both facies are interpreted as more energetic flood deposits but
with different sediment sources A unique breccia layer with coarse silt matrix and
cm-long soft clasts (F5) is interpreted as a high-energy event (ie a tsunami)
capable of eroding the littoral zone and depositing coarse clastic material in the
distal zone of the lake Similar coarse breccia layers have been found at several
coastal sites in Chile and interpreted as tsunami-related deposits (Vargas et al
2005 Le Roux et al 2008)
33 Sedimentary units
Three main units and six subunits have been defined (Fig 3) based on
sedimentary facies and sediment composition We use ZrTi as an indicator of the
mineral fraction transported from the watershed (Marzecoacuteva et al 2011) as higher
ZrTi (F3 and F4) is commonly associated with coarser sediments (Cuven et al
2010) A high correlation among Br BrTi and TOC (r = 046-087 p value=0)
supports the use of BrTi as an indicator of lake productivity (Marzecoacuteva et al 2011
Frugone-Aacutelvarez et al 2017) The MnFe ratio is indicative of lake bottom
oxygenation (Naecher et al 2013) as under reducing conditions Mn mobilizes more
than Fe leading to a decreased MnFe ratio (Marzecoacuteva et al 2011) SrTi indicates
periods of increased aragonite formation as Sr is preferentially included in the
aragonite mineral structure (Veizer et al 1971) (See supplementary material)
The basal unit (3c 129 to 99 cm) is relatively organic-rich (TOC mean=26
BioSi mean=5) and composed by F1 (without aragonite) with some coarser F4
flood layers (ZrTi mean=025) Unit 3b (98 to 80 cm) is interpreted as a tsunami or
storm surge deposit (breccia F5 grading into massive to banded silt F2) Unit 3a
(79 to 73 cm) is characterized by relatively low productivity (TOC=2 BrTi=002
42
BioSi=4) under anoxic conditions (MnFe=001) Unit 2a (72 to 31cm) has
relatively less organic content and more intercalated clastic facies F3 and F4 The
top of this unit (43-30 cm) has elevated TS values The Subunit 1b (30-20 cm)
shows increasing TOC BioSi and BrTi values (TOC mean = 29 BioSi mean =
54 BrTi mean=004) and the upper subunit 1a (19-0 cm) has the highest TOC
(mean = 64) and BioSi (mean = 56) high BrTi mean = 010 and the presence
of aragonite More frequent anoxic conditions (MnFe lower than 001) during units
3 and 2 shifted towards more oxic episodes (MnFe = 003) in unit 1 (Fig 3)
34 Isotopic signatures
Figure 4 shows the isotopic signature from soil samples of the major land
usescover present in the Laguna Matanzas used as an end member in comparison
with the lacustrine sedimentary units δ15N from cropland samples exhibit the
highest values whereas grassland and soil samples from lake shore areas have
intermediate values (Fig 4) Tree plantations and native forests have similarly low
δ15N values (+11 permil SD=24) All samples (except those from the lake shore)
exhibit low δ13C values (from ndash285permil to ndash298permil) CNmolar from agriculture land
lakeshore area and non-vegetation areas samples display the lowest values (about
18) CNmolar from tree plantations and native forest have the highest values (383
and 267 respectively)
43
Figure 4 C-N stable isotope plot showing a comparison of lake sediments grouped
by sedimentary units (MAT11-6A) with the soil end members of present-day (lake
shore and land usecover) from Laguna Matanzas
The δ15N values from sediment samples (MAT11-6A) range from ndash15 and
+53permil (mean= +35 SD=05) δ13C values range from ndash266permil to ndash202permil (mean=
ndash240 SD=14) In unit 3c δ15N and δ13C show relatively high values (mean=
+41permil and ndash233permil respectively) δ15N and δ13C from unit 3b and 3a fluctuate at
slightly lower values than in 3c (mean δ15N from 3a= +38permil and 3b mean= +39permil
mean δ13C from 3a= ndash242permil and 3b mean= ndash244permil) In unit 2a δ15N values are
relatively high (mean= +38permil) but show a slightly decreasing trend (from +52 to
+24permil at 66 cm and 36 cm respectively) δ13C also decreases (mean= ndash242permil)
reaching minimum values at 45 cm (ndash266permil) with a sharp increase towards the top
of this unit with maximum values ca ndash210permil Unit 1 exhibits the lowest δ15N values
(1a mean= +11permil 1b mean= +28permil) with negative values in the uppermost
44
sediments (ndash04permil at 14 cm) δ13C values show a decreasing trend over most of
subunit 1b and increase only near the very top of this unit
35 Recent land use changes in the Laguna Matanzas watershed
Major LUCC from 1975 (unit 1b) to 2016 CE in the Laguna Matanzasacutes
watershed is summarized in Figure 5 The watershed has a surface area of 30 km2
of which native forest (36) and grassland areas (44) represented 80 of the
total surface in 1975 The area occupied by agriculture was only 02 and tree
plantations were absent Isolated burned areas (33) were located mostly in the
northern part of the watershed By 1989 tree plantations surface area had increased
to 5 burned areas to 17 and agricultural fields to 9 (a 45-fold increase) and
native forest and grassland sectors decreased to 23 and 27 respectively By
2016 agricultural land and tree plantations have increased to 17 of the total area
whereas native forests decreased to 21
45
Figure 5 Land uses and cover changes from 1975 to 2016 in Laguna Matanzas
watershed from natural cover and areas for livestock grazing (grassland) to the
expansion of agriculture and forest plantation
4 DISCUSSION
41 N and C dynamics in Laguna Matanzas
Small lakes with relatively large watersheds such as Laguna Matanzas would
be expected to have relatively high contributions of allochthonous C to the sediment
OM (Gu et al 2006) Terrestrial C3 plants (δ13C =ndash26 to ndash28permil Meyers and Terranes
2001 Ku et al 2007) are dominant in the Laguna Matanzas watershed Likewise
our soil samples ranged across similar although slightly more negative values
46
(δ13C = ndash30 to ndash28permil Fig 4) to those proposed by Meyers and Terranes (2001)
and are used here as terrestrial end members oil samples were taken from the lake
shore (δ13C = ndash22permil) and POM from surface water (δ13C = ndash24permil) were more
positive than the terrestrial end member and are used as lacustrine end members
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from terrestrial vegetation and more positive δ13C values have increased
aquatic OM (Gu et al 2006 Torres et al 2012) Phytoplankton preferentially uptake
12C leaving the DIC pool enriched in 13C (Gu et al 2006) especially when there are
no important external sources of C (eg decreased C input from the watershed)
Therefore during events of elevated primary productivity the phytoplankton uptakes
12C until its depletion and are then obligated to use the heavier isotope resulting in
an increase in δ13C Changes in lake productivity thus greatly affect the C isotope
signal (Torres et al 2012) with high productivity leading to elevated δ13C values
(Torres et al 2012 Gu et al 2006)
In a similar fashion the N isotope signatures in Laguna Matanzas reflect a
combination of factors including different N sources (autochthonousallochthonous)
and lake processes such as productivity isotope fractionation in the water column
and sediment denitrification Elevated δ15N values from a POM sample (+22permil) and
average values from the lake shore (mean=+34permil SD=028) are used as aquatic
end members whereas terrestrial samples have values from +10 +24 (tree species)
to +77permil +35 (agriculture) which we use as terrestrial end members (Fig 4)
Autochthonous OM in aquatic ecosystems typically displays low δ15N values
when the OM comes from N-fixing species Atmospheric fixation of N2 by
cyanobacteria ends in OM with δ15N values close to 0permil (Leng et al 2006)
Phytoplankton preferentially uptake 14N from Dissolved Inorganic Nitrogen (DIN) in
47
the water column and derived OM typically have δ15N values lower than DIN values
When productivity increases the remaining DIN becomes depleted in 14N which in
turn increases the δ15N values of phytoplankton over time especially if the N not
replenished (Torres et al 2012) Thus high POM δ15N values from Laguna
Matanzas reflect elevated phytoplankton productivity with a 14N- depleted DIN In
addition N-watershed inputs also contribute to high δ15N values Heavily impacted
watersheds by human activities are often reflected in isotope values due to land use
changes and associated modified N fluxes For example the input of N runoff
derived from the use of inorganic fertilizers leads to the presence of elevated δ15N
(between ndash4 to +4permil) values in the water bodies (Elliott and Brush 2006 Diebel and
Vander Zanden 2009) Widory et al (2004) reported a direct relationship between
elevated δ15N values and increased nitrate concentration from manure in the
groundwater of Brittany (France) Terranes and Bernasconi (2000) found a good
correlation between augmented nutrient loading and a progressive increase in δ15N
values of sedimentary OM related to agricultural land use
Post-depositional diagenetic processes can further affect C and N isotope
signatures Several studies have shown a decrease in δ13C values of OM in anoxic
environments particularly during the first years of burial related to the selective
preservation of OM depleted in 13C (Hollander and Smith 2001 Lehmann et al
2002 Gӓlman et al 2009 Torres et al 2012) Diagenetic processes can also lead
to post-burial N isotope enrichment of the sediments Indeed 14N is consumed more
rapidly than 15N by denitrifying bacteria which intensifies under anoxic conditions
(Diebel and Vander Zanden 2009) Thus the remaining OM pool will be enriched
in 15N leaving to elevated δ15N values (Granger et al 2008 Menzel et al 2013)
48
In summary the relatively high δ15N values in sediments of Laguna Matanzas
reflect N input from an agriculturegrassland watershed with positive synergetic
effects from increased lake productivity enrichment of DIN in the water column and
most likely denitrification The increase of algal productivity associated with
increased N terrestrial input andor recycling of lake nutrients (and lesser extent
fixing atmospheric N) and denitrification under anoxic conditions can all increase
δ15N values (Fig 3) In addition elevated lake productivity without C replenishing
(eg by terrestrial C input) produces shifts towards positive δ13C values whereas C
input from the watershed generates more negative δ13C values
42 Recent evolution of the Laguna Matanzas watershed
Sedimentological compositional and geochemical indicators show three
depositional phases in the lake evolution under the human influence in the Laguna
Matanzas over the last two hundred years Although the record is longer (around
1000 years) we analyzed the last two centuries (unit 2 and 1) to provide a recent
historical context for the large changes detected during the 20th century
The first phase lasted from the beginning of the 19th century until ca 1940
(unit 2a Fig 3 4) and was characterized by moderate productivity with elevated
sediment input from the watershed as indicated by our geochemical proxies (BrTi
= 004 AlTi gt 01 Fig 6) Higher lake levels and dominant anoxic bottom conditions
(MnFe = 001 Fig3) can be related to increased precipitation (mean= 557 mmyr)
and lower temperatures (summer annual temperature lt19ordmC) During the Spanish
colonial period the Laguna Matanzas watershed was used as a livestock farm
(Jesuits 1627-1780 unit 3 followed by the Pedro Balmaceda era 1780-1810- unit
2a) (Contreras-Loacutepez et al 2014) The main buildings and farm facilities at the El
49
Convento village During this period livestock grazing and lumber extraction for
mining would have involved extensive deforestation and loss of native vegetation
(eg Armesto et al 1994 2010) However the Matanzas pollen record does not
show any significant regional deforestation during this period (Villa Martiacutenez 2002)
suggesting that the impact may have been highly localized
Lake productivity sediment input and elevated precipitation (Fig 6) all
suggest that N availability was related to this increased input from the watershed
The N from cow manure and soil particles would have led to higher δ15N values
(Elliot and Brush 2016) In addition anoxic lake bottom conditions would lead to
even further enrichment of buried sediment N The δ13C values lend further support
to our interpretation of increased sediment input -and N- from the watershed
Decreased δ13C values reach their lowest values in the entire record (ndash270permil) at
ca 1910 CE (Fig 4 6)
During most of the 19th century human activities in Laguna Matanzas were
similar to those during the Spanish Colonial period However the appearance of
Pinus radiata and Eucalyptus globulus pollen ca 1890 CE (Gibson 1972) the dune
stabilization-afforestation program which began ca 1900 CE (Albert 1900) and the
application of the Chilean forestry law of 1931 (DFL nordm265) contributed to an
increased capacity of the surrounding vegetation to retain nutrients and sediments
The law subsidized forest plantations in areas devoid of vegetation and prohibited
the cutting of forest on slopes greater than 45ordm These land use changes were coeval
with decreased sediment inputs (AlTi trend) from the watershed slightly increased
lake productivity (BrTi from 001 to 003) and a decrease in annual precipitation
(Fig 6) N isotope values become more negative during this period although they
remained high (from +49permil to +37permil) whereas the δ13C trend towards more
50
positive values reflects changes in the N source from watershed to in-lake dynamics
(e g increased endogenic productivity)
The second phase started after 1940 and is clearly marked by an abrupt
change in the general trend of δ15N that oscillates around +41permil (SD = 04permil) during
the first phase to +24permil (SD = 14permil) during the second These δ15N trends reflect
the lowest watershed nutrient and sediment inputs (based on the AlTi record)
decreased precipitation (mean = 318 mm year) and a slight increase in lake
productivity (increased BrTI) Depositional dynamics in the lake likely crossed a
threshold as human activity intensified throughout the watershed and lake levels
decreased
During the Great Acceleration δ15N values shifted towards higher values to
ca 3permil with an increase in δ13C values that are not reflected either in lake
productivity or lake level As the sediment input from the watershed increased and
precipitation remained as low as the previous decade δ15N values during this period
are likely related to watershed clearance which would have increased both nutrient
and sediment input into the lake
The δ13C trend to more positive values reaching the peaks in the 1960s (ndash
212permil) at the same time as the peaks in temperature (196 ordmC mean annual) with a
downward trend in precipitation A shift in OM origin from macrophytes and
watershed input influences to increased lake productivity could explain this trend
(Fig 4 1b)
In the 1970s the Laguna Matanzasacute watershed was mostly covered by native
forest (36) and grassland areas were intended for livestock grazing (44 Fig 5)
Both soils have δ15NSoil lt 5permil (Fig 4) Farming fields occupied a very small area and
tree plantations were almost nonexistent The decreasing trend in δ15N values seen
51
in our record is interrupted by several large peaks that occurred between ca 1975
and ca 1989 when the native forest and grassland areas fell by 23 and 27
respectively largely due to fires affecting 17 of the forests Agriculture fields
increased by 4 and forest plantations increased by 9 (unit 1a) Concomitantly
sediment ndash and likely N - inputs from the watershed decreased (as indicated by the
trend in AlTi) the precipitation was still relatively low (Fig 6) These changes are
likely related to the increase of vegetation cover especially of tree plantations (which
have more negative δ15N values) The small increase in productivity in the lake could
have been favored by increased temperature (von Gunten et al 2009) After 1989
the increase in agriculture land (17 in 2016) correlates with increasing δ15N δ13C
and TOC trends in spite of declining rainfall The increase of forest plantations was
mostly in response to the implementation of the Law Decree of Forestry
Development (DL 701 of 1974) that subsidized forest plantation After 1989 the
increase in agricultural land (17 in 2016) is synchronous with increasing δ15N
δ13C TOC and MnFe trends despite the decline in rainfall and overall lower lake
levels as more water is used for irrigation
The third phase started c 1990 CE (unit 1a) when OM accumulation rates
increase and δ13C δ15N decreased reaching their lowest values in the sequence
around 2000 CE Afterward during the 21st century δ13C and δ15N values again
began to increase The onset of unit 1 is marked by increased lake productivity and
decreased sediment input (AlTi lt 02) synchronous with intensive farming replacing
forestry and extensive agriculture (Fig 5 6)
A change in the general trend of δ15N values which decreased until 1990
(unit 1a) as the native forest and grassland area fell to 23 and 27 respectively
is most likely due to deforestation and fires Agriculture surface increased to 4 and
52
forest plantations increased by 9 (Fig 5) Concomitantly sediment ndash and likely N
ndash inputs from the watershed decreased probably related to the low precipitation (Fig
1b) and the increase of vegetation cover in the watershed in particularly by tree
plantations (with more negative δ15N Fig 4)
At present agriculture and tree plantations occupy around 34 of the
watershed surface whereas native forests and grassland cover 21 and 25
respectively Lake productivity as indicated by BrTi (Fig6) is higher and generates
OM with lower δ15N and higher δ13C values (about +2permil and ndash20permil ca 2000 CE
respectively) due to in-lake processes (ie biological N fixation and nutrient
recycling) and driven by changes in the arboreal cover which diminishes nutrient
flux into the lake (Fig6)
53
Figure 6 Anthropogenic and climatic forcing and lake dynamics response
(productivity sediment input N and C cycles) at Matanzas Lake over the last two
54
centuries Mean annual precipitation reconstructed and temperatures (von Gunten
et al 2009) Vertical gray bars indicate mega-droughts
5 CONCLUSIONS
Human activities have been the main factor controlling the N and C cycle in
the Laguna Matanzas during the last two centuries The N isotope signature in the
lake sediments reflects changes in the watershed fluxes to the lake but also in-lake
processes such as productivity and post-depositional changes Denitrification could
have been a dominant process during periods of increased anoxic conditions which
were more frequent prior to Great Acceleration (1950 CE) Higher δ15N and lower
δ13C values are associated with increased nutrient input from the watershed due to
increased livestock grazing and agriculture pressure (phase 1 Fig 7a) whereas
lower isotope values occurred during periods of increased forest plantations (phase
3 Fig 7c) During periods of increased lake productivity - such as in the last few
decades - δ15N values increased significantly
The most important change in C and N dynamics in the lake occurred after the
1950s (Phase 2 and 3) as tree plantations dominated the watershed Recent
changes in N dynamics can be explained by the higher nutrient contribution
associated with intensive agriculture (i e fertilizers) since the 1990s Although the
replacement of livestock activities with forestry and farming seems to have reduced
nutrient and soil export from the watershed to the lake the inefficient use of fertilizer
(by agriculture) can be the ultimate responsible for lake productivity increase during
the last decades
55
Figure 7 Schematic diagrams illustrating the main factors controlling the
isotope N signal in sediment OM of Laguna Matanzas N input from watershed
depends on human activities and land cover type Agriculture practices and cattle
(grassland development) contribute more N to the lake than native forest and
plantations Periods of higher productivity tend to deplete the dissolved inorganic N
in 14N resulting in higher δ15N (OM) The denitrification processes are more effective
in anoxic conditions associated with higher lake levels
6 METHODS
Short sediment cores were recovered from Laguna Matanzas using an Uwitec
gravity piston corer in 2011 and 2013 (Fig 1a) Sediment cores (MAT11-6A 129 cm
MAT13-2A 149 cm MAT13-3A 33 cm MAT13-4A 99 cm) were split
photographed sub-sampled and stored at the Pyrenean Institute of Ecology (IPE-
CSIC Spain) Core MAT11-6A was obtained from the central sector of the lake and
56
was selected for detailed multiproxy analyses (including elemental geochemistry C
and N isotope analyses XRF and 14C dating)
The isotope analyses (δ13C and δ15N) were performed at the Laboratory of
Biogeochemistry and Applied Stable Isotopes (LABASI-PUC Chile) using a Delta
V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via a
Conflo IV interface Isotope results are expressed in standard delta notation (δ) in
per mil (permil) relative to the standards Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Sediment samples
for δ13Corg were pre-treated with 50 ml of HCl and 50 ml of deionized water and
dried at 60ordmC for 4 hr to remove carbonates (Harris et al 2011)
Total Carbon (TC) Total Organic Carbon (TOC) Total Inorganic Carbon (TIC)
and Total Sulphur (TS) were measured every cm with a LECO SC 144 DR at IPE-
CSIC XRF measurements were carried out every 4 mm in MAT11-1A-1G core using
an AVAATECH X-ray Fluorescence II core scanner at the University of Barcelona
(Spain) Results are expressed as element intensities in counts per second (cps)
Tube voltage was operated at 30 kV and 10 kV to obtain the abundances of 15
elements (Al Si S Cl K Ca Ti V Mn Fe Br Rb Sr Y Zr) with an average at
least of 1600 cps (less for Br=1000)
Biogenic silica content mineralogy and grain size were measured every 4
cm Biogenic silica was measured following Mortlock and Froelich (1989) and
Bernaacuterdez et al (2005) using an Auto Analyzer Technicon AAII for dissolved silicate
analysis Mineralogy was analyzed with a Siemens D-500 X-ray diffractometer (Cu
kα 40 kV 30 mA graphite monochromator) at the ICTJA-CSIC (Spain) Grain size
analyses were performed in a Beckmann Coulter LS 13 320 Particle Size Analyzer
57
at the IPE-CSIC The samples were classified according to textural classes as
follows clay (lt2 μm) silt (20-2 microm) and sand (gt2 microm) fractions
The age-depth model for the Laguna Matanzas sedimentary sequence was
constructed using 210Pb and 137Cs dating techniques (MAT13-4A) as well as two 14C
AMS dates on bulk sediment samples (MAT11-6A fig 2) We dated the dissolved
inorganic carbon (DIC) in the water column and no significant reservoir effect is
present in the modern-day water column (10454 + 035 pcmc Table 2) An age-
depth model was obtained with the Bacon R package to estimate the deposition
rates and associated age uncertainties along the core (Blaauw and Christen 2011)
To estimate LUCC of Laguna Matanzasacutes watershed we use satellite images
Landsat MSS for 1975 Landsat TM for 1989 and Landsat OLI for 2016 all taken in
summer or autumn (Table 1) We performed supervised classification of land uses
(maximum likelihood algorithm) for each year (1975 1989 and 2016) and the results
were mapped using software ArcGIS 102 in 2017
Satellite Images Acquisition Date
Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat OLI 20160404 30 m
Table 1 Landsat imagery
Surface water samples were filtered for obtained particulate organic matter In
addition soil samples from the main land usecover present in the Laguna Matanzas
watershed were collected Elemental C N and their corresponding isotopes from
POM and soil were obtained at the LABASI and used here as end members
Daily precipitation at Santo Domingo (33ordm39`S 71ordm3 6`W)- the nearest weather
station to Laguna Matanzasndash was compiled using the redPrec R package (Fig1 d
Serrano-Notivoli et al 2017 Sarricolea et al 2019) To extend our precipitation
58
reconstruction back to 1824 we correlated this dataset with that available for
Santiago The Santiago data was compiled from data published in the Anales of
Universidad of Chile (1850) for the years 1824 to 1849 Almeyda (1940) for the years
1849 to 1864 and the Quinta Normal series from 1866 to the present (Direccioacuten
Meteoroloacutegica de Chile) We generated a linear regression model between the
presentday Santo Domingo station and the compiled Santiago data with a Pearson
coefficient of 087 and p-valuelt 001
Acknowledgments This research was funded by grants CONICYT AFB170008
to the Institute of Ecology and Biodiversity (IEB) and FONDECYT 1150763 (to CL)
Doctoral grant Becas Chile 21150224 MEDLANT (Spanish Ministry of Economy
and Competitiveness grant CGL2016-76215-R) Additional funding was provided
by the Laboratorio Internacional de Cambio Global (LINCGlobal PUC-CSIC) We
thank R Lopez E Royo and M Gallegos for help with sample analyses We thank
the Laboratory of Biogeochemistry and Applied Stable Isotopes (LABASI) of the
Department of Ecology (PUC) for sample analyses
References
Albert F 1900 Las dunas o sean arenas volantes voladeros arenas muertas invasion de las arenas playas I meacutedanos del centro de Chile comprendiendo el litoral desde el liacutemite norte de la provincia de Aconcagua hasta el liacutemite sur de la de Arauco Memorias cientiacuteficas I Lit
Anderson NJ W D Fritz SC 2009 Holocene carbon burial by lakes in SW
Greenland Glob Chang Biol httpsdoiorg101111j1365-2486200901942x
Armesto J Villagraacuten C Donoso C 1994 La historia del bosque templado
chileno Ambient y Desarro 66 66ndash72 httpsdoiorg101007BF00385244
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A
59
historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Arnell NW Gosling SN 2013 The impacts of climate change on river flow
regimes at the global scale J Hydrol 486 351ndash364 httpsdoiorg101016JJHYDROL201302010
Bernaacuterdez P Prego R Franceacutes G Gonzaacutelez-Aacutelvarez R 2005 Opal content in
the Riacutea de Vigo and Galician continental shelf Biogenic silica in the muddy fraction as an accurate paleoproductivity proxy Cont Shelf Res httpsdoiorg101016jcsr200412009
Blaauw M Christen JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474 httpsdoiorg10121411-BA618
Brush GS 2009 Historical land use nitrogen and coastal eutrophication A
paleoecological perspective Estuaries and Coasts 32 18ndash28 httpsdoiorg101007s12237-008-9106-z
Camarero L Catalan J 2012 Atmospheric phosphorus deposition may cause
lakes to revert from phosphorus limitation back to nitrogen limitation Nat Commun httpsdoiorg101038ncomms2125
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego
R Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and environmental change from a high Andean lake Laguna del Maule central Chile (36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Catalan J Pla-Rabeacutes S Wolfe AP Smol JP Ruumlhland KM Anderson NJ
Kopaacuteček J Stuchliacutek E Schmidt R Koinig KA Camarero L Flower RJ Heiri O Kamenik C Korhola A Leavitt PR Psenner R Renberg I 2013 Global change revealed by palaeolimnological records from remote lakes A review J Paleolimnol httpsdoiorg101007s10933-013-9681-2
Chen Y Y Huang W Wang W H Juang J Y Hong J S Kato T amp
Luyssaert S (2019) Reconstructing Taiwanrsquos land cover changes between 1904 and 2015 from historical maps and satellite images Scientific Reports 9(1) 3643 httpsdoiorg101038s41598-019-40063-1
Contreras S Werne J P Araneda A Urrutia R amp Conejero C A (2018)
Organic matter geochemical signatures (TOC TN CN ratio δ 13 C and δ 15 N) of surface sediment from lakes distributed along a climatological gradient on the western side of the southern Andes Science of The Total Environment 630 878-888 httpsdoiorg101016jscitotenv201802225
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia UC
60
de V 2014 ELEMENTOS DE LA HISTORIA NATURAL DEL An Mus Hist Natulas Vaplaraiso 27 51ndash67
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-010-9453-1
Diebel M Vander Zanden MJ 201209 Nitrogen stable isotopes in streams
stable isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19 1127ndash1134 httpsdoiorg10189008-03271
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506 Espizua LE Pitte P 2009 The Little Ice Age glacier advance in the Central
Andes (35degS) Argentina Palaeogeogr Palaeoclimatol Palaeoecol 281 345ndash350 httpsdoiorg101016JPALAEO200810032
Figueroa JA Castro S A Marquet P A Jaksic FM 2004 Exotic plant
invasions to the mediterranean region of Chile causes history and impacts Invasioacuten de plantas exoacuteticas en la regioacuten mediterraacutenea de Chile causas historia e impactos Rev Chil Hist Nat 465ndash483 httpsdoiorg104067S0716-078X2004000300006
Fowler D Coyle M Skiba U Sutton MA Cape JN Reis S Sheppard
LJ Jenkins A Grizzetti B Galloway N Vitousek P Leach A Bouwman AF Butterbach-bahl K Dentener F Stevenson D Amann M Voss M 2013 The global nitrogen cycle in the twenty- first century Philisophical Trans R Soc B httpsdoiorghttpdxdoiorg101098rstb20130164
Frank D Reichstein M Bahn M Thonicke K Frank D Mahecha MD
Smith P van der Velde M Vicca S Babst F Beer C Buchmann N Canadell JG Ciais P Cramer W Ibrom A Miglietta F Poulter B Rammig A Seneviratne SI Walz A Wattenbach M Zavala MA Zscheischler J 2015 Effects of climate extremes on the terrestrial carbon cycle Concepts processes and potential future impacts Glob Chang Biol httpsdoiorg101111gcb12916
Fritz SC Anderson NJ 2013 The relative influences of climate and catchment
processes on Holocene lake development in glaciated regions J Paleolimnol httpsdoiorg101007s10933-013-9684-z
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
61
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS) implications for past sea level and environmental variability J Quat Sci 32 830ndash844 httpsdoiorg101002jqs2936
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892 httpsdoiorg101126science1136674
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924 httpsdoiorg104319lo20095430917
Garciacutea-Ruiz JM 2010 The effects of land uses on soil erosion in Spain A
review Catena httpsdoiorg101016jcatena201001001
Garciacutea-Ruiz JM Loacutepez-Moreno JI Lasanta T Vicente-Serrano SM
Gonzaacutelez-Sampeacuteriz P Valero-Garceacutes BL Sanjuaacuten Y Begueriacutea S Nadal-Romero E Lana-Renault N Goacutemez-Villar A 2015 Los efectos geoecoloacutegicos del cambio global en el Pirineo Central espantildeol una revisioacuten a distintas escalas espaciales y temporales Pirineos httpsdoiorg103989Pirineos2015170005
Garciacutea-Ruiz JM Nadal-Romero E Lana-Renault N Begueriacutea S 2013
Erosion in Mediterranean landscapes Changes and future challenges Geomorphology 198 20ndash36 httpsdoiorg101016jgeomorph201305023
Garreaud RD Vuille M Compagnucci R Marengo J 2009 Present-day
South American climate Palaeogeogr Palaeoclimatol Palaeoecol httpsdoiorg101016jpalaeo200710032
Gayo EM Latorre C Jordan TE Nester PL Estay SA Ojeda KF
Santoro CM 2012 Late Quaternary hydrological and ecological changes in the hyperarid core of the northern Atacama Desert (~21degS) Earth-Science Rev httpsdoiorg101016jearscirev201204003
Ge Y Zhang K amp Yang X (2019) A 110-year pollen record of land use and land
cover changes in an anthropogenic watershed landscape eastern China Understanding past human-environment interactions Science of The Total Environment 650 2906-2918 httpsdoiorg101016jscitotenv201810058
Goyette J Bennett EM Howarth RW Maranger R 2016 Global
Biogeochemical Cycles 1000ndash1014 httpsdoiorg1010022016GB005384Received
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and
oxygen isotope fractionation during dissimilatory nitrate reduction by
62
denitrifying bacteria 53 2533ndash2545 httpsdoiorg10230740058342
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
Gurnell AM Petts GE Hannah DM Smith BPG Edwards PJ Kollmann
J Ward J V Tockner K 2001 Effects of historical land use on sediment yield from a lacustrine watershed in central Chile Earth Surf Process Landforms 26 63ndash76 httpsdoiorg1010021096-9837(200101)261lt63AID-ESP157gt30CO2-J
Heathcote A J et al Large increases in carbon burial in northern lakes during the
Anthropocene Nat Commun 610016 doi 101038ncomms10016 (2015) Hollander DJ Smith MA 2001 Microbially mediated carbon cycling as a
control on the δ13C of sedimentary carbon in eutrophic Lake Mendota (USA) New models for interpreting isotopic excursions in the sedimentary record Geochim Cosmochim Acta httpsdoiorg101016S0016-7037(00)00506-8
Holtgrieve GW Schindler DE Hobbs WO Leavitt PR Ward EJ Bunting
L Chen G Finney BP Gregory-Eaves I Holmgren S Lisac MJ Lisi PJ Nydick K Rogers LA Saros JE Selbie DT Shapley MD Walsh PB Wolfe AP 2011 A coherent signature of anthropogenic nitrogen deposition to remote watersheds of the Northern Hemisphere Science (80) 334 1545ndash1548 httpsdoiorg101126science1212267
Hooper DU Adair EC Cardinale BJ Byrnes JEK Hungate BA Matulich
KL Gonzalez A Duffy JE Gamfeldt L Connor MI 2012 A global synthesis reveals biodiversity loss as a major driver of ecosystem change Nature httpsdoiorg101038nature11118
Horvatinčić N Sironić A Barešić J Sondi I Krajcar Bronić I Borković D
2016 Mineralogical organic and isotopic composition as palaeoenvironmental records in the lake sediments of two lakes the Plitvice Lakes Croatia Quat Int httpsdoiorg101016jquaint201701022
Howarth RW 2004 Human acceleration of the nitrogen cycle Drivers
consequences and steps toward solutions Water Sci Technol httpsdoiorg1010382Fscientificamerican0490-56
Jenny B Valero-Garceacutes BL Urrutia R Kelts K Veit H Appleby PG Geyh
M 2002 Moisture changes and fluctuations of the Westerlies in Mediterranean Central Chile during the last 2000 years The Laguna Aculeo record (33deg50primeS) Quat Int 87 3ndash18 httpsdoiorg101016S1040-
63
6182(01)00058-1
Jenny B Wilhelm D Valero-Garceacutes BL 2003 The Southern Westerlies in
Central Chile Holocene precipitation estimates based on a water balance model for Laguna Aculeo (33deg50primeS) Clim Dyn 20 269ndash280 httpsdoiorg101007s00382-002-0267-3
Leng M J Lamb A L Heaton T H Marshall J D Wolfe B B Jones M D
amp Arrowsmith C 2006 Isotopes in lake sediments In Isotopes in palaeoenvironmental research (pp 147-184) Springer Dordrecht
Le Roux JP Nielsen SN Kemnitz H Henriquez Aacute 2008 A Pliocene mega-
tsunami deposit and associated features in the Ranquil Formation southern Chile Sediment Geol 203 164ndash180 httpsdoiorg101016jsedgeo200712002
Lehmann M Bernasconi S Barbieri a McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66 3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Lenihan JM Drapek R Bachelet D Neilson RP 2003 Climate change
effects on vegetation distribution carbon and fire in California Ecol Appl httpsdoiorg101890025295
McLauchlan K K Gerhart L M Battles J J Craine J M Elmore A J
Higuera P E amp Perakis S S (2017) Centennial-scale reductions in nitrogen availability in temperate forests of the United States Scientific reports 7(1) 7856 httpsdoi101038s41598-017-08170-z
Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32ordmS 3050 masl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
Martiacuten-Foreacutes I Casado MA Castro I Ovalle C Pozo A Del Acosta-Gallo
B Saacutenchez-Jardoacuten L Miguel JM De 2012 Flora of the mediterranean basin in the chilean ESPINALES Evidence of colonisation Pastos 42 137ndash160
Marzecovaacute A Mikomaumlgi a Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54 httpsdoiorg103176eco2011105
Matesanz S Valladares F 2014 Ecological and evolutionary responses of
Mediterranean plants to global change Environ Exp Bot httpsdoiorg101016jenvexpbot201309004
64
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Menzel P Gaye B Wiesner MG Prasad S Stebich M Das BK Anoop A
Riedel N Basavaiah N 2013 Influence of bottom water anoxia on nitrogen isotopic ratios and amino acid contributions of recent sediments from small eutrophic Lonar Lake central India Limnol Oceanogr 58 1061ndash1074 httpsdoiorg104319lo20135831061
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Michelutti N Wolfe AP Cooke CA Hobbs WO Vuille M Smol JP 2015
Climate change forces new ecological states in tropical Andean lakes PLoS One httpsdoiorg101371journalpone0115338
Muntildeoz-Rojas M De la Rosa D Zavala LM Jordaacuten A Anaya-Romero M
2011 Changes in land cover and vegetation carbon stocks in Andalusia Southern Spain (1956-2007) Sci Total Environ httpsdoiorg101016jscitotenv201104009
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Poraj-Goacuterska A I Żarczyński M J Ahrens A Enters D Weisbrodt D amp
Tylmann W (2017) Impact of historical land use changes on lacustrine sedimentation recorded in varved sediments of Lake Jaczno northeastern Poland Catena 153 182-193 httpdxdoiorg101016jcatena201702007
Pongratz J Reick CH Raddatz T Claussen M 2010 Biogeophysical versus
biogeochemical climate response to historical anthropogenic land cover change Geophys Res Lett 37 1ndash5 httpsdoiorg1010292010GL043010
Prudhomme C Giuntoli I Robinson EL Clark DB Arnell NW Dankers R
Fekete BM Franssen W Gerten D Gosling SN Hagemann S Hannah DM Kim H Masaki Y Satoh Y Stacke T Wada Y Wisser D 2014 Hydrological droughts in the 21st century hotspots and uncertainties from a global multimodel ensemble experiment Proc Natl Acad Sci httpsdoiorg101073pnas1222473110
65
Ruumlhland KM Paterson AM Smol JP 2015 Lake diatom responses to
warming reviewing the evidence J Paleolimnol httpsdoiorg101007s10933-
015-9837-3
Sarricolea P Meseguer-Ruiz Oacute Serrano-Notivoli R Soto M V amp Martin-Vide
J (2019) Trends of daily precipitation concentration in Central-Southern Chile Atmospheric research 215 85-98 httpsdoiorg101016jatmosres201809005
Sernageomin 2003 Mapa geologico de chile version digital Publ Geol Digit Serrano-Notivoli R de Luis M Begueriumlia S 2017 An R package for daily
precipitation climate series reconstruction Environ Model Softw httpsdoiorg101016jenvsoft201611005
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Stine S 1994 Extreme and persistent drought in California and Patagonia during
mediaeval time Nature 369 546ndash549 httpsdoiorg101038369546a0
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL
Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Syvitski JPM Voumlroumlsmarty CJ Kettner AJ Green P 2005 Impact of humans
on the flux of terrestrial sediment to the global coastal ocean Science 308(5720) 376-380 httpsdoiorg101126science1109454
Thomas RQ Zaehle S Templer PH Goodale CL 2013 Global patterns of
nitrogen limitation Confronting two global biogeochemical models with observations Glob Chang Biol httpsdoiorg101111gcb12281
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Urbina Carrasco MX Gorigoitiacutea Abbott N Cisternas Vega M 2016 Aportes a
la historia siacutesmica de Chile el caso del gran terremoto de 1730 Anuario de Estudios Americanos httpsdoiorg103989aeamer2016211
Vargas G Ortlieb L Chapron E Valdes J Marquardt C 2005 Paleoseismic
inferences from a high-resolution marine sedimentary record in northern Chile
66
(23degS) Tectonophysics httpsdoiorg101016jtecto200412031 399(1-4) 381-398httpsdoiorg101016jtecto200412031
Valero-Garceacutes BL Latorre C Morelliacuten M Corella P Maldonado A Frugone M Moreno A 2010 Recent Climate variability and human impact from lacustrine cores in central Chile 2nd International LOTRED-South America Symposium Reconstructing Climate Variations in South America and the Antarctic Peninsula over the last 2000 years Valdivia p 76 (httpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-yearshttpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-years
Vicente-Serrano SM Gouveia C Camarero JJ Begueria S Trigo R
Lopez-Moreno JI Azorin-Molina C Pasho E Lorenzo-Lacruz J Revuelto J Moran-Tejeda E Sanchez-Lorenzo A 2013 Response of vegetation to drought time-scales across global land biomes Proc Natl Acad Sci httpsdoiorg101073pnas1207068110
Villa Martiacutenez R P 2002 Historia del clima y la vegetacioacuten de Chile Central
durante el Holoceno Una reconstruccioacuten basada en el anaacutelisis de polen sedimentos microalgas y carboacuten PhD
Villa-Martiacutenez R Villagraacuten C Jenny B 2003 Pollen evidence for late -
Holocene climatic variability at laguna de Aculeo Central Chile (lat 34o S) The Holocene 3 361ndash367
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Human alteration of the global nitrogen cycle sources and consequences Ecological applications 7(3) 737-750 httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Rein B Urrutia R amp Appleby P (2009) A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 The Holocene 19(6) 873-881 httpsdoiorg1011770959683609336573
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Widory D Kloppmann W Chery L Bonnin J Rochdi H Guinamant JL
2004 Nitrate in groundwater An isotopic multi-tracer approach J Contam Hydrol 72 165ndash188 httpsdoiorg101016jjconhyd200310010
67
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
68
Supplementary material
Facie Name Description Depositional Environment
F1 Organic-rich
mud
Massive to banded black
organic - rich (TOC up to 14 )
mud with aragonite in dm - thick
layers Slightly banded intervals
contain less OM (TOClt4) and
aragonite than massive
intervals High MnFe (oxic
bottom conditions) High CaTi
BrTi and BioSi (up to 5)
Distal low energy environment
high productivity well oxygenated
and brackish waters and relative
low lake level
F2 Massive to
banded silty clay
to fine silt
cm-thick layers mostly
composed by silicates
(plagioclase quartz cristobalite
up to 65 TOC mean=23)
Some layers have relatively high
pyrite content (up to 25) No
carbonates CaTi BrTi and
BioSi (mean=48) are lower
than F1 higher ZrTi (coarser
grain size)
Deposition during periods of
higher sediment input from the
watershed
69
F3 Banded to
laminated light
brown silty clay
cm-thick layers mostly
composed of clay minerals
quartz and plagioclase (up to
42) low organic matter
(TOC mean=13) low pyrite
and BioSi content
(mean=46) and some
aragonite
Flooding events reworking
coastal deposits
F4 Laminated
coarse silts
Thin massive layers (lt2mm)
dominated by silicates Low
TOC (mean=214 ) BrTi
(mean=002) MnFe (lt02)
TIC (lt034) BioSi
(mean=46) and TS values
(lt064) and high ZrTi
Rapid flooding events
transporting material mostly
from within the watershed
F5 Breccia with
coarse silt
matrix
A 17 cm thick (80-97 cm
depth) layer composed by
irregular mm to cm-long ldquosoft-
clastsrdquo of silty sediment
fragments in a coarse silt
matrix Low CaTi BrTi and
MnFe ratios and BioSi
Rapid high energy flood
events
70
(mean=43) and high ZrTi
(gt018)
Table Sedimentological and compositional characteristics of Laguna Matanzas
facies
71
CAPIacuteTULO 2 STABLE ISOTOPES TRACK LAND USE AND COVER
CHANGES IN A MEDITERRANEAN LAKE IN CENTRAL CHILE OVER THE
LAST 600 YEARS
72
Stable isotopes track land use and cover changes in a mediterranean lake in
central Chile over the last 600 years
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugone-Aacutelvarezabc Pablo
Sarricolead Leonardo Villaciacutese M Laura Carrevedob Blas Valero-Garceacutescg
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Departamento de Ecologiacutea Universidad de ChileLas Palmeras 3425 Nuntildeoa Santiago Chile
f Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding authors E-mail addresses clatorrebiopuccl magdalenafuentealbagmailcom
Key words Anthropocene lake sediment δsup1⁵N δsup1sup3C Great Acceleration organic
geochemistry watershedndashlake system Stable Isotope Analyses land usecover
change Nitrogen cycle mediterranean ecosystems central Chile
73
Abstract
Nutrient fluxes in many aquatic ecosystems are currently being overridden by
anthropic controls especially since the industrial revolution (mid-1800s) and the
Great Acceleration (mid-1900s) Land use and cover changes (LUCC) alter the
availability and fluxes of nutrients such as nitrogen that are transferred via runoff
and groundwater into lakes By altering lake productivity and trophic status these
changes are often preserved in the sedimentary record Here we use
geolimnological proxies and stable isotope analyses (δsup1⁵N δsup13C) on lake sediments
to reconstruct the lake-watershed nutrient transfer in a central Chilean lake (Lago
Vichuqueacuten) over the past 600 years We compare our multiproxy record from recent
lake sediments to the soilvegetation relationship across the watershed as well as
land usecover changes from 1975 to 2014 derived from satellite images Our results
show that lake sediment δsup1⁵N values increased with meadow cover but decreased
with tree plantations suggesting increased nitrogen retention when trees dominate
the watershed Our δsup1⁵N record from central Chile can thus be interpreted as a proxy
for nutrient availability over the last 600 years mainly derived from land use changes
coupled with climate drivers Although variable sources of organic matter and in situ
fractionation often hinder straightforward environmental interpretations of stable N
isotope signatures our study shows that lake sediment δsup1⁵N can be a useful tool for
assessing the contribution of past human activities in nutrient and nitrogen cycling
1 Introduction
Nitrogen (N) is a limiting nutrient in terrestrial and aquatic ecosystems (Vitousek
et al 1997) Changes in its availability can drive eutrophication and increase
pollution in these ecosystems (McLauchlan et al 2013) Although recent human
74
impacts on the global N cycle have been significant the consequences of increased
anthropogenic N on these ecosystems are unclear (Gerhart and Mclauchlan 2014
Battye et al 2017) Isotopic signatures are key to understand N dynamics in lakes
nevertheless in situ andor diagenetic fractionation along with multiple sources of
organic matter (OM) often hinder straightforward environmental interpretations from
isotopes Monitoring δ15N and δ13C values as components of the N cycle
specifically those related to the link between terrestrial and aquatic ecosystems can
help differentiate between effects from processes versus sources in stable isotope
values (eg from Particulate Organic Matter -POM- soil and vegetation) and
improve how we interpret variations in δ15N (and δ13C) values at longer temporal
scales
The main processes controlling stable N isotopes in bulk lake OM are source
lake productivity diagenesis and denitrification (Talbot 2001 Leng et al 2006
Fariacuteas et al 2009 Torres et al 2012 Brahney et al 2014) N sources depend on
contributions from the watershed (ie soil and biomass) the transfer of atmospheric
N to the lake (ie atmospheric N2 fixation) and nutrient recycling (Leng et al 2006)
Atmospheric N2 (isotopically defined as δ15N =0) is fixed by cyanobacteria with
minimal fractionation resulting in δ15NOM values that fluctuate from -2 to +2permil (Fogel
and Cifuentes 1993 Talbot 2001 Elliot and Brush 2006) Besides N2 fixation by
cyanobacteria phytoplankton species assimilate dissolved inorganic nitrogen (DIN)
and values typically range from +7 to +10permil (Xu et al 2016 Meyers et al 2003) In
addition seasonal changes in POM occur in the lake water column Gu et al (2006)
sampled the water column of Lake Wauberg (Florida USA) during a hydrologic year
and found a higher development of N fixing species during the summer A major
factor behind this increase are human activities in the watershed which control the
75
inputs of the sediment and nutrients to the lake (Woodward et al 2011) Some
studies have shown higher δ15N values in lake sediments from watersheds that are
highly affected by human activities (Bruland and Mackenzie 2010 Woodward et al
2011) Syntenic fertilizers displayed δ15N values between -4 to +4permil cattle manure
around +8 to +22permil and surface and groundwater OM between 0 to +10permil (Elliott
and Brush 2006 Leng et al 2006) Although relatively low δ15N values from
fertilizers constitute major N input to human-altered watersheds the elevated loss
of 14N via volatilization of ammonia and denitrification leaves the remaining total N
input enriched in 15N (Bruland and Mackenzie 2010)
In addition to the different sources and variations in lake productivity early
diagenesis at the sedimentndashwater interface in the sediment can further alter
sediment OM isotope values (Brahney et al 2014 Galman et al 2009) During
diagenetic loss of N in lacustrine systems kinetic fractionation occurs and the
remaining N is enriched in 15N leading to higher δ15N values (Galman et al 2006
Talbot 2001) Denitrification tends to enrich lake sediments in 15N due to the
assimilation of nitrates by denitrifying bacteria (Granger et al 2008) and is more
prevalent in anoxic environments (Brahney et al 2016 Lehman et al 2002)
Carbon isotopes in lake sediments can also provide useful information about
paleoenvironmental changes OM origin and depositional processes (Meyers et al
2003) Allochthonous organic sources (high CN ratios) produce isotope values
similar to values from catchment vegetation Autochthonous organic matter (low CN
ratio) is influenced by fractionation both in the lake and the watershed leading up to
carbon assimilation by lacustrine biota (Woodward et al 2011) Changes in
productivity greatly affect δ13COM values (Torres et al 2012) as during C uptake
plankton preferentially assimilate 12C leaving the dissolved inorganic carbon (DIC)
76
pool enriched in 13C (Gu et al 2006) with phytoplankton averaging ca 20 permil lower
than the respective δ13CDIC values (Leng et al 2006) Under conditions of low to
moderate primary productivity plankton preferentially uptake the lighter 12C
resulting in low δ13C values displayed by the OM (Torres et al 2012) Conversely
during high primary productivity phytoplankton will uptake 12C until its depletion and
is then forced to assimilate the heavier isotope resulting in an increase in δ13C
values Higher productivity in C-limited lakes due to slow water-atmosphere
exchange of CO2 also results in high δ13C values (Galman et al 2009) In these
cases algae are forced to uptake dissolved bicarbonate with δ13C values between
7 to 10permil higher than those of the atmospheric CO2 dissolved in water (Sun et al
2016 Torres et al 2012 Galman et al 2009)
Stable isotope analyses from lake sediments are thus useful tools to
reconstruct shifts in lake-watershed dynamics caused by changes in limnological
parameters and LUCC Our knowledge of the current processes that can affect
stable isotope signals in a watershed-lake system is limited however as monitoring
studies are scarce Besides in order to use stable isotope signatures to reconstruct
past environmental changes we require a multiproxy approach to understand the
role of the different variables in controlling these values Hence in this study we
carried out a detailed survey of current N dynamics in a coastal central Chilean lake
(Lago Vichuqueacuten) and a multiproxy study of a sediment record spanning the last
600 years The characterization of the recent changes in the watershed since 1970s
is based on satellite images to compare recent changes in the lake and assess how
these are related with climate variability and an ever increasing human footprint
(Lara et al 2012 Gallardo et al 2015 Steffen et al 2015) Our main goal is to
investigate how stable isotope values from lake sediment reflect changes in the lake
77
ndash watershed system during periods of high watershed disruption (eg Spanish
Conquest late XIX century Great Acceleration) and recent climate change (eg
Little Ice Age and current global warming)
2 Study Site
Lago Vichuqueacuten (34ordm49`S 72ordm04W 4 m asl 30 m of depth Fig 1) is a
mesotrophic to eutrophic lake with dominant anoxic bottom conditions that is
stratified year around (Frugone-Aacutelvarez et al 2017) The main inlets are the
Vichuqueacuten and Huintildee rivers and the only outlet is the Llico estuary that drains into
the Pacific Ocean High tides can sporadically shift the flow direction of the Llico
estuary which increases the marine influence in the lake Dune accretion gradually
limited ocean-lake connectivity until the estuary was almost completely closed off
by ca 10 ky BP and the current lake regime formed (Frugone-Aacutelvarez et al 2017)
The area is characterized by a mediterranean climate with cold-wet winters and
hot-dry summers and an annual precipitation of ~650 mm and a mean annual
temperature of 15ordmC During the austral winter months (June - August) precipitation
is modulated by north-west shifts of the South Pacific Anticyclone (SPA) followed by
an increased frequency of storm fronts stemming off the South Westerly Winds
(SWW) A strengthened SPA during austral summers (December - March) which
are typically dry and warm blocks the northward migration of storm tracks stemming
off the SWW (Fig1 Frugone-Aacutelvarez et al 2017)
78
Figure 1 a) The location of the Lago Vichuqueacuten in central Chile b) Map of land
uses and cover in 2016 c) Ombrothermic diagram mediterranean climates are
characterized by cold-wet winters with surplus moisture from June to August and
hot-dry summers d) Lake bathymetry showing location of cores and water sampling
sites used in this study
Although major land cover changes in the area have occurred since 1975 to the
present as the native forests were replaced by tree (Monterey pine and eucalyptus)
plantations the region was settled before the Spanish conquest (Frugone-Alvarez
et al 2018) Historical documents indicate that Vichuqueacuten was occupied by a
Mitimaes colony as part of the Inka empire (Odone 1998) Unlike other Andean
areas during the Inka period the introduction of corn and quinoa in the Vichuqueacuten
watershed do not seem to have intensified land use The Spanish colonial period in
Chile lasted from 1542 CE to the independence in 1810 CE The first historical
document (1550 CE) shows that the areas around Vichuqueacuten were settled by the
Spanish early on as the Vichuqueacuten watershed was given as an ldquoencomiendardquo
system to Juan Cuevas The ldquoencomiendardquo system provided the owners with land
79
and indigenous people to work but also the introduction of wheat wine cattle
grazing and logging of native forests for lumber extraction and increasing land for
agricultural activities (Odone et al 1998 Armesto et al 2010) During the 19th
century (the Republic) the export of wheat to Australia and Canada generated
intensive changes in land cover use The town of Vichuqueacuten became the regional
capital and plans to build a maritime port in the bay of Vichuqueacuten were drawn
However the fall of international markets in 1880 paralyzed these plans During the
20th century the plantation of trees (Pinus radiata and Eucalyptus globulus) in areas
cleared of native forests was subsidized by two national laws DFL nordm265 (1931) and
DFL nordm 701 (1974) both of which provided funds for such plantations During the last
decades the urbanization with summer vacation homes along the shorelines of
Lago Vichuqueacuten has greatly increased Extensive lake eutrophication is currently a
large environmental problem (EULA 2008)
3 Methods
Coring campaigns were organized from 2011 to 2015 (Fig 1d) We recovered
12 short cores and 4 long cores (up to 14 m long) at several sites using a hammer-
modified UWITEC gravity corer Sediment cores (VIC11-2A 170 cm VIC13-2B 170
cm VIC13-2D 1200 cm and VIC15-1A 119 cm) were split photographed sub-
sampled and stored in the Pyrenean Institute of Ecology (IPE-CSIC Spain) Core
VIC13-2B was selected for detailed multiproxy analyses (including elemental
geochemistry C and N isotopes XRF and 14C dating) Stable isotope analyses
(δ13C δ15N) were performed at the Laboratory of Biogeochemistry and Applied
Stable Isotopes (LABASI-PUC Chile) Sediment samples for δ13Corg were pre-
treated with 50 ml of HCl and 50 ml of deionized water and dried at 60ordmC for 4 hr to
remove carbonates (Harris et al 2011) Isotope analyses were conducted using a
80
Delta V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via
a Conflo IV interface Isotope results are expressed in standard delta notation (δ)
and expressed in per mil (permil) relative to the Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Total carbon (TC)
Total Organic Carbon (TOC) Total Inorganic Carbon (TIC) and Total Sulfur (TS)
were measured every 2 cm with a LECO SC 144 DR at the IPE-CSIC
An AVAATECH X-Ray Fluorescence II core scanner with a Rh X-ray tube from
the University of Barcelona was used to obtain XRF logs every 4 mm of resolution
Results are expressed as element intensities in counts per second (cps) Tube
voltage was operated at 30 kV and 10 kV to obtain the abundances of 18 elements
(Pb Al Si S Cl K Ca Ti V Mn Fe Co Zn Br Rb Sr Y Zr) with an average of
at least of 1800 cps (less for Pb=437 and Zn=934 cps) Pb was included due to
similar behavior with Co and Fe Element ratios were calculated to describe changes
in redox conditions (MnFe) endogenic productivity (BrTi) carbonate formation
(SrCa and CaTi) and sediment input (ZrTi) (Barreiro-Lostres et al 2014
Carrevedo et al 2015 Frugone-Aacutelvarez et al 2017 Kylander et al 2011 Moreno
et al 2007a)
Several campaigns were carried out to sample the POM from the water column
two per hydrologic year from November 2015 to August 2018 A liter of water was
recovered in three sites through to the lake two are from the shallower areas (with
samples taken at 2 and 5 m depth at each site) and one in the deeper central portion
(at 2 5 and 20 m depth) of the lake (Fig 1d) The samples were filtered using glass
fiber filters (25 m) and then frozen to obtain the stable carbon and nitrogen isotope
signal of lacustrine POM Additionally soil and vegetation samples from the
following communities native species meadow hydrophytic vegetation and
81
Monterrey pine (Pinus radiata) were taken for stable isotope analyses (see in
supplementary material)
The age model for the complete Lago Vichuqueacuten sedimentary sequence is
based on Frugone-Aacutelvarez et al (2017) with the last 1000 years based on
210Pb137Cs dating techniques in VIC15-1A and 14C AMS dates on bulk sediment
samples (Supplementary Table S1) The 14C measurements of lake water DIC show
a moderate reservoir effect of 180 plusmn 25 14C yr) The revised age-depth model used
here includes three more 14C AMS dates performed with the program Bacon to
establish the deposition rates along the core (Blaauw and Christen 2011 Fig 2)
The age-depth model indicates that average resolution between 0 to 87 cm is lt2
cm per year and from 88 to 170 cm it is lt47 cm per year
82
Figure 2 Chronological age-depth model for the Lago Vichuqueacuten sedimentary
sequence spanning the last 1000 yr (see also Frugone-Alvarez et al 2017)
To estimate land use changes in the watershed we use Landsat MSS images
for 1975 and Landsat TM for 1989 and 2014 all taken in either summer or autumn
(Table 1) We performed supervised classification of land uses (maximum likelihood
83
algorithm) for each year (1975 1989 and 2014) and results were mapped using
ArcGIS 102
Table 1 Images using for LUCC reconstruction
Source of LUCC
Acquisition
Date Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat TM 19991226 30 m
CONAF 2009 30 m
Land cover Chile 2014 30 m
CONAF 2016 30 m
Previous Work on Lago Vichuqueacuten sedimentary sequence
The sediments are organic-poor dark brown to brown laminated silt with some
intercalated thin coarser clastic layers Lacustrine facies have been classified
according to elemental composition (TOC TS TIC and TN) grain size and
sedimentary textures (Frugone-Aacutelvarez et al 2017) Four main types of lacustrine
facies were identified in this short core Facies L1 is a laminated (1cm) black to dark
brown diatom-poor silt with relatively higher TOC percentages (TOC about 185)
TS (about of 18) and TOCTN (about 92) Facies L2 is characterized by a
homogeneous black organic-poor silty clay with low TOC (mean= 13) TS (mean=
13) TOCTN (mean= 88) values Facies L3 is a laminated (1cm) black organic-
poor (TOC mean 14) silts with higher TS (mean= 21) and TOCTN ratios
(mean= 88) Facies L3 is interpreted as ldquobaselinerdquo deposition on the distal areas
84
of Vichuqueacuten lake Mineralogy is dominated by mica (muscovite) clay minerals
(chlorite) and plagioclase (albite) Quartz and calcite occur at the top and pyrite
occurs in the lower part of the sequence Facies T is composed by massive banded
sediments 4 cm thick and coarse grain size It is interpreted as an instantaneous
depositional event (for more detail see Frugone-Aacutelvarez et al 2017) In this work
we identified four subunits based on geochemical and stable isotope signals
4 Results
41 Geochemistry and PCA analysis
High positive correlations exist between Al Si K and Ti (r = 078 ndash 096
supplementary Fig S1) and between Zn Zr Rb and Y (r = 072 - 096) which reflect
the silicate content of the watershed-derived sediments (Marzecoacuteva et al 2011) Zr
is commonly associated with minerals more abundant in coarser deposits Thus the
ZrTi profile could reflect grain size (Cuven et al 2010) showing higher variability
in the upper part of the Lake Vichuqueacuten sequence and in the alternation between
facies L1 and L2 Another group of elements made up of Co Fe and Pb displayed
positive correlations (r = 067ndash 097) and represents the input of heavy metals Br
Cl and Ca show a positive correlation (r = 049 ndash 06 p-value = 00001) BrTi ratio
is interpreted as a productivity indicator due to Br having a strong affinity with humic
and organic compounds (Marzecoacuteva et al 2011 Frugone-Aacutelvarez et al 2017) In
our Lago Vichuqueacuten sequence BrTi values are high from 170 to 132 cm and from
36 cm to the top of core where it shows several sharps peaks (Fig 3) The MnFe
ratio is indicative of lake bottom oxygenation (Naecher et al 2013) as under
reducing conditions Mn tends to become more mobile than Fe leading to a
decreased MnFe (Marzecoacuteva et al 2011) Several sharp peaks in MnFe occurred
85
from 127 to 120 cm and from 57 to 30 cm (MgFegt03) The silicate group and the
Br Cl Ca Mn group are negatively correlated (r= -012 and -066)
Principal Component Analysis (PCA) was undertaken on the XRF
geochemical data to investigate the main factors controlling sediment deposition in
Lago Vichuqueacuten (Fig 3) The four main eigenvectors explain 801 of the variance
(supplementary material Table S2) The principal component (PC1) explain 437
of the total of variance and grouped elements are associated with terrigenous input
to the lake Positive values of the biplot have been attributed to higher heavy metals
deposition (Pb Co and Fe) and negative values to higher silicate input (Al K Ti and
Si) in a high energy catchment environment (Zr) The PC2 explains 198 of the
total of variance and highlights the endogenic productivity in the lake The positive
loading is compound by Br Cl Ca and Sr and to a lesser extent by Y Zn Rb and
Mn The PC2 may explain both the precipitation of carbonates (Ca Sr) and biological
production (Br)
86
Figure 3 Principal Component Analysis of XRF geochemical measurements in
VIC13-2B Lago Vichuqueacuten lake sediments
42 Sedimentary units
Based on geochemical and stable isotope analysis we identified four
lithological subunits in the short core sedimentary sequence Our PCA analyses and
Pearson correlations pointed out which variables were better for characterizing the
subunits (Fig 4) in terms of endogenic productivity heavy metals and terrestrial
input The unit 2c (170-131 cm) is characterized by the occurrence of some clastic
layers (L2 from 160 and 150 cm) high δsup1⁵N values (mean= +59 plusmn 04permil) with
Magnetic Susceptibility (MS) BrTi ZrTi and SrCa values increase towards the top
Also it has relatively low δsup1sup3C values (mean= -265 plusmn 11permil) and CNmolar ratios
(mean=78 plusmn 04) The ZrTi show upwards increasing values reaching the highest
values of the sequence at the top of this unit suggesting a coarsening upward trend
and relatively higher depositional energy The MS trend also indicates higher
erosion in the watershed and enhanced delivery of ferromagnetic minerals likely
from soil particles particularly from 150 to 160 cm (Frugone-Aacutelvarez at al 2017)
The subunit 2b (130-118 cm) is also composed of black silts but it has the
lowest MS values of the whole sequence and its onset is marked by a sharp
decrease in ZrTi This subunit is marked by sharp peaks of MnFe (at 126 and 120
cmgt027) SrCa (at 125 and 120 cmgt033) CaTi (at 124 and 120 cmgt199) TOC
(about 26 at 130 124 122 and 118 cm) and δsup1⁵N (gt+67permil at 126 and 120 cm)
BrTi oscillates around low values (about 01) whereas the δsup1sup3C values range
between -262 and -282permil
87
The unit 2a (58-117 cm) shows increasing and then decreasing MS values
and displayed sharps peaks of δsup1⁵N (gt64 at 102 to 99 87 and 75 cm) and CN
(about 10 at 86 74 66 and 60 cm) SrCa (mean = 04 plusmn 013) BrTi (mean = 008
plusmn 002) MnFe (mean = 002 plusmn 001) and δsup1sup3C (mean = -260 plusmn 07permil) oscillate in
low values The upper part of this unit (72 ndash 57 cm) shows an increase of SrCa
(from 03 to 05)
The upper unit 1a (0 - 57 cm) starts with a fine clastic layer (T at 57 to 54
cm) interpreted as deposition during a high-energy event It is characterized by
lower BrTi (mean = 003 plusmn 002) TOC (mean = 12 plusmn 03) and δsup1sup3C (mean= -
266 plusmn 06permil) higher δsup1⁵N (mean = +55 plusmn 08 permil) This unit is made of alternating
fine (lt 1cm) black and dark brown laminae with carbonate (L1 facies) Recently
deposited sediments show decreasing MS values increasing TOC (mean= 15 plusmn
04 ) sharp peaks of BrTi (gt015 at 33 21 5 and 1 cm) and the lowest δsup1⁵N values
of the entire record (mean= 45 plusmn 06 permil) and δsup1sup3C values (mean= -277 plusmn 09 permil)
Heavy metal content increases abruptly in the upper section of the core (2 - 20 cm)
(peaks of FeTi CoTi and PbTi)
88
Figure 4 Sedimentary units of Lago Vichuqueacuten sedimentary sequence Selected
variables illustrate watershed input (Magnetic Susceptibility ZrTi CoTi and MnFe)
endogenic productivity (SrCa CaTi TS) organic matter source (BrTi TOC
CNmolar and stable isotope records (δ13Corg and δ15Nbulk)
43 Recent seasonal changes of particulate organic matter on water column
The average of δ15NPOM from the three sites across to the lake was +86 plusmn 58
permil oscillating widely from -14 to +193permil (Fig 5) Important seasonal differences
occur at 2 and 5 m depth with more negative values during summer (~53 plusmn 52 permil)
than winter (~122 plusmn 40permil) At 20 m depth δ15NPOM values exhibit small seasonal
ranges (mean = +10permil for winter and +87permil for summer) The δ13CPOM average was
-296 plusmn 33permil with slightly seasonal and water column depth differences However
more positive δ13CPOM values occurred during winter (mean=-290 plusmn 41permil) than in
summer (mean= -301 plusmn 19 permil) Like the δ15NPOM values CNPOM (mean= 83 plusmn 17)
displayed important seasonal and water depth differences Lower CNPOM ratios
89
occurred during summer at 2 and 5 m depth (mean= 95 plusmn 17) and higher and more
constant values during winter (mean = 73 plusmn 09) at the same depth At 20 m CNPOM
shows similar values in both winter (70) and summer (74)
Figure 5 Seasonal POM averages from samples taken of the Lago Vichuqueacuten
water column Samples were taken from 2015 to 2018 at 2 (n=16) 5 (n=14) and 20
(n=8) meters depth
44 Stable isotope values across the Lake Vichuqueacuten watershed
Figure 6 shows modern vegetation soil and sediment isotope values found for
the watershed Huintildee tributary and Vichuqueacuten lake The δsup1⁵NFoliar values from
meadow plantations and macrophytes have similar range values with a mean of
+89 plusmn50 permil (n=18) +74 plusmn 75 permil (n=5) +82 plusmn 36 permil (n=6) respectively Native
vegetation has overall lower δsup1⁵NFoliar values with a mean of 14 plusmn 21 permil (n=28 see
Table 2 in supplementary material) The δsup1sup3CFoliar from terrestrial samples exhibit
similar values across the different plant communities (tree plantation mean=-274 plusmn
13permil meadow mean = -275 plusmn 23 permil native species mean=-272 plusmn 14permil) whereas
macrophytes display slightly more negative values with a mean of -287 plusmn 23permil
Macrophytes substrate displayed the highest δsup1⁵N values with a mean of +60 plusmn
14permil (n=3) Similar and relatively high δsup1⁵N values for soil of plantation (mean= +54
plusmn 13permil n=4) meadow (mean= +49 plusmn 04permil n=8) and surface river sediment
90
(mean= +52 plusmn 17permil n=10) Native forest soils oscillate around slightly more
negative values with a mean of +45 plusmn 12permil (n=4) Slightly more positive soil δsup1sup3C
values occur both underneath native forests and in tree plantations with means of -
284 plusmn 23permil and -298 plusmn 13permil respectively compared to either meadow soils
(mean= -302 plusmn 11permil) aquatic sediments underneath macrophytes (-311 plusmn 10permil)
or from surface river sediments (mean= -312 plusmn 10permil)
Figure 6 Stable isotope content of modern soil river sediment and foliar vegetation
used as end members in the sedimentary sequence of Lago Vichuqueacuten a)
Scatterplot of foliar δsup1⁵N and δsup13C values for vegetation from Lago Vichuqueacuten
watershed (plantation meadow and native species) and macrophytes on Lake
Vichuqueacuten See supplementary material for more detail of vegetation types b)
Scatterplot of soil δsup1⁵N and δsup13C values across the watershed the sediments of the
Huintildee and Vichuqueacuten rivers and the fluvial sediment corresponding to the
macrophyte vegetation
45 Land use and cover change from 1975 to 2014
Major land use changes between 1975 CE and 2016 CE in the Lago
Vichuqueacuten watershed are summarized in Figure 7 The watershed has a surface
area of 535329 km2 of which native vegetation (26) and shrublands (53)
represent 79 of the total surface in 1975 Meadows are confined to the valley and
91
represent 17 of watershed surface Tree plantations initially occupied 1 of the
watershed and were first located along the lake periphery By 1989 the areas of
native forests shrublands and meadows had decreased to 22 31 and 14
respectively whereas tree plantations had expanded to 30 These trends
continued almost invariably until 2016 when shrublands and meadows reached 17
and 5 of the total areas while tree plantations increased to 66 Native forests
had practically disappeared by 1989 and then increased up to 7 of the total area
in 2016 (Fig 6) preferentially occupying the southeastern sector of the watershed
Figure 7 Changes in land use and cover between 1975 and 2016 in the Lago
Vichuqueacuten watershed as measured from satellite images The major change is
represented by the replacement of native forest shrubland and meadows by
plantations of Monterrey pine (Pinus radiata)
Figure 8 shows correlations between lake sediment stable isotope values and
changes in the soil cover from 1975 to 2013 Positive relationships occurred
between sediment δsup1⁵N and δsup13C values from core VIC13-2B (unit 1) and the
92
percentage of native forest (r=089 for δsup1⁵N 082 for δsup13C) shrublands (r = 071 for
δsup1⁵N 057 for δsup13C) and meadow cover (r = 088 for δsup1⁵N 081 for δsup13C) All of these
correlations are significant (p value lt 0001) In contrast significant negative
correlations (p lt0001) occurred between tree plantation cover and lake sediment
stable isotope values (δsup1⁵N r = -082 and δsup13C r = -067) native forest (r = -097)
meadows (r = -086) and shrubland (r =-093)
Figure 8 Correlation plots of land use and cover change versus lake sediment
stable isotope values The δsup1⁵N values are positively correlated with native forests
agricultural fields and meadow cover across the watershed Total Plantation area
increases are negatively correlated with native forest meadow and shrubland total
area Significance levels are indicated by the symbols p-values (0 0001 001
005 01 1) lt=gt symbols ( )
93
5 Discussion
51 Seasonal variability of POM in the water column
The stable isotope values of POM can vary during the annual cycle due to
climate and biologic controls namely temperature and length of the photoperiod
which affect phytoplankton growth rates and isotope fractionation in the water
column (Gu et al 2006 Leng et al 2006) POM from Lago Vichuqueacuten surface
samples (2 and 5 m depth) displayed lower δ13CPOM values during the summer than
in winter During C uptake phytoplankton preferentially utilize 12C leaving the
DICpool enriched in 13C Therefore as temperature increases during the summer
phytoplankton growth generates OM enriched in 12C until this becomes depleted
and then the biomas come to enriched u At the onset of winter the DICpool is now
enriched in 13C and despite an overall decrease in phytoplankton production the
OM produced will also be enriched in 13C In contrast δ13CPOM values at 20 m depth
did not reflect these seasonal differences probably due to water-column
stratification that maintains similar temperatures and biological activity throughout
the year
Lake N availability depends on N sources including inputs from the
watershed and the atmosphere (ie deposition of N compounds and fixation of
atmospheric N2) which varies during the hydrologic year The fixation of atmospheric
N2 is an important natural source of N to the lake occurring mainly during the
summer season associated with higher temperature and light (Gu et al 2006)
Because the δ15NPOM of N2 fixers is close to 0permil with almost no overall isotope
fractionation (Leng et al 2006) when N2 is the major N source δ15NPOM values are
typically low However when DIN concentrations are high or alternatively when little
94
N2 fixation occurs δ15NPOM values will be higher (Gu et al 2006) δ15NPOM values
from summer Lago Vichuqueacuten samples were lower than those from winter with large
differences between the seasons (~5permil Fig 5 and 9) Furthermore δ15NPOM values
were high when monthly average temperature was low and monthly precipitation
was high (Fig 9) This pattern is consistent with the dominance of high N2 fixation
by cyanobacteria associated with increased summer temperatures This correlation
of δ15NPOM values with temperature further suggests a functional group shift i e
from N fixers to phytoplankton that uptake DIN The correlation between wetter
months and higher δ15NPOM values could be caused by increased N input from the
watershed due to increased runoff during the winter season The lack of data of the
δ15NPOM between the 30 mm and the 160 mm of precipitation is due to the
mediterranean-type climate that concentrates precipitations in the winter months
Finally Lago Vichuqueacuten is characterized by pronounced seasonality leading to
higher phytoplankton biomass in summer characterized by low δ15NPOM In winter
low biomass production and increased input from watershed is associated to high
δ15NPOM
95
Figure 9 Seasonal changes can drive δ15NPOM values in Lago Vichuqueacuten Data
correspond to average monthly temperature and total monthly precipitation for the
months when the water samples were taken (years 2015 - 2018) P-valuelt005
52 Stable isotope signatures in the Lake Vichuqueacuten watershed
The natural abundance of 15N14N isotopes of soil and vegetation samples
from the Lago Vichuqueacuten watershed appear to result from a combination of factors
isotope fractionation different N sources for plants and soil microorganisms (eg N2
fixation Nitrates) depth in the soil where N uptake by vegetation occurs and N loss
mechanisms (ie denitrification leaching and ammonia volatilization Hogberg
1997) The lowest δsup1⁵Nfoliar values are associated with native species and are
probably due to the contribution of N-fixing species (eg Sophora) (Fig6 and for
more detail see Table S3 in supplementary material) The number native N-fixers
species present in the Chilean mediterranean vegetation are not well known
however so there could be other N-fixing species in this group Alternatively δsup1⁵Nfoliar
values reflect soil N uptake (Kahmen et al 2008) In environments limited by N
plants have δsup1⁵N values similar to the soil δsup1⁵N values however the denitrification
and volatilization of ammonia can lead to the remain N of soil to come enriched in
15N (Cifuentes et al 1989 Craine et al 2009 Bruland and Mckenzie 2010) Soil N
isotope samples from native species communities tends to display relatively high
δsup1⁵N values respect to foliar samples due to loss of N-soil
The higher foliar and soil δsup1⁵N values obtained from samples of meadows
aquatic macrophytes and tree plantations can be attributed to the presence of
greater amounts of nitrate available for plant uptake (Fig 6) Houmlgberg (1997)
suggests that the availability of different N sources in soils (ie nitrates versus
96
ammonia) with different residence times can also explain these δsup1⁵NFoliar values
Indeed Feigin et al (1974) described differences of up to 20permil between ammonia
and nitrates sources Denitrification and nitrification discriminate much more against
15N than N2 fixation does (Craine et al 2009) leading to soils (and plants after
uptake) enriched in 14N
In general multiple processes that affect the isotopic signal result in similar
δsup1⁵N values between the soil of the watershed and the sediments of the river
However POM isotope fluctuations allow to say that more negative δsup1⁵N values are
associated to lake productivity while more positive δsup1⁵N values are associated with
N input from the watershed
δ13C values in Lago Vichuqueacuten watershed reflect CO2 gas exchange between
C3 plants and algae with the atmosphere During photosynthesis plants
discriminate against the heavier stable isotope of carbon (13C) in favor of the lighter
isotope 12C which results in δ13CFoliar values between -220permil and -320permil (Browman
and Cook 2002 Liu et al 2007) Likewise the δ13CFoliar values from Lago Vichuqueacuten
oscillate around -27permil (Fig 6) Plants transform atmospheric CO2 into soil organic
carbon (C) which in turn reflects this initial discrimination against 13C during C
uptake and post photosynthetic fractionation (Farquhar et al 1982 1989 Badeck
et al 2005 Pausch and Kuzyakov 2017) Slightly more positive δ13CFoliar values
(about 15permil) were measured in comparison with their δ13CSoil values This may be
reflecting the C transference from plants to the soil but also a soil-atmosphere
interchange The preferential assimilation of the light isotopes (12C) during soil
respiration carried by the roots and the microbial biomass that is associated with the
decomposition of litter roots and soil organic matter explain this differential
(Berhardt et al 2006 Blagodatskaya et al 2010 Midwood and Millard 2011)
97
In general the δ13CSoil values from the Lago Vichuqueacuten watershed oscillated
around -290permil and did not vary with our plant classification types Here we use
these values as terrestrial-end members to track changes in source OM (Fig 6)
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from the terrestrial watershed By the other hand more positive δ13C
values most likely reflect an increased aquatic OM component as indicated by POM
isotope fluctuations (Fig 9)
53 Recently land use and cover change and its influences on N inputs to the lake
Tree plantations in the Lago Vichuqueacuten watershed have increased greatly in
the last few decades (from 1 in 1975 to 66 in 2016 Fig 7) replacing previous
native forests (from 26 to 7) meadows (17 to 5) and shrublands (53 to
17) In 1975 tree plantations were confined to the lake perimeter with discrete
patches distributed throughout the watershed The ldquoForest Lawrdquo (DFL 701- passed
in 1974) allocated state funding to afforestation efforts and management of tree
plantations which greatly favored the replacement native forests by introduced trees
This increase is marked by a sharp and steady decrease in lake sediment δ15N and
δ13C values because tree plantations function as a nutrient sink whereas other land
uses such as meadows favor nutrient loss from the watershed (Fig8) Bruland and
Mackenzie (2014) noted a decrease in wetland δ15N values when watershed
forested cover increased and concluded that N inputs to the wetlands are lower from
the forested areas as they generally do not export as much N as agricultural lands
A positive correlation between native vegetation and δ15Ncore values can be
explained by the relatively scarce arboreal cover in the watershed in 1975 when
native forest occupied just 26 of the watershed surface whereas shrublands and
98
meadows occupied more than the 70 of the surface of the watershed with the
concomitant elevated loss of N (Fig 7 and Fig 8)
54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years
Sedimentological compositional and geochemical indicators all show changes in
the N cycle associated to LUCC lake dynamics and climate changes (Fig 10) From
the pre-Columbian indigenous settlement including the Spanish colonial period up
to the start of the Republic (1300 - 1800 CE) the introduction of crops such as
quinoa and wheat but also the clearing of land for extensive agriculture would have
favored the entry of N into the lake Conversely major changes observed during the
last century were characterized by a sharp decrease of N input that were coeval
with the increase of tree plantations
From 1365 until 1550 CE (unit 2c 2b and half of 2a Fig 3 and Fig 10) pre-
Columbian agricultural societies planted mostly crops of corn and quinoa (Ramirez
and Vidal 1985) This period is characterized by large peaks seen in the δsup1⁵N record
(eg 1403 1452 1476 and 1505) with overall high δsup1⁵N values (up to 5permil) indicating
that N input from watershed was elevated and oscillating to the beat of the NT These
positive δsup1⁵N peaks could be due to several causes including a) the clearing of land
for farming b) N loss via denitrification which would be generally augmented in
anoxic environments (Diebel et al 2009) such as Lago Vichuqueacuten (low MnFe
values about 0001 Fig 4) or c) climate especially cold-wet winters or hot-dry
summers can also exert control on the δsup1⁵N record Indeed tree-ring records and
summer temperature reconstructions show overall wetcold conditions during this
period (Christie et al 2009 Von Gunten et al 2009 Fig 10) Increased
precipitation would bring more sediment (and nutrients) from the watershed into the
99
lake and increase lake productivity which is also detected by the geochemical
proxies for productivity (peaks of BrTi SrCa and TOC Fig 4 and Fig 10) (see also
Frugone-Alvarez et al 2017)
Figure 10 Changes in the N availability during the last six centuries in Lago
Vichuqueacuten a) lake sediment δsup1⁵N values displayed more positive values during the
prehistoric period Spanish Colony and the starting 19th century which is associated
with enhanced N input from the watershed by extensive clearing and crop
plantations The inset shows this relationship between sediment δsup1⁵N and
100
percentage of meadow cover over the last 30 years b) Summer temperature
reconstruction from central Chile (von Gunten et al 2009) showing a
correspondence with cold phases and high δsup1⁵N values at Vichuqueacuten except for the
last 100 years c) Palmer Drought Severity Index (PDSI) is an interannual moisture
variability reconstruction for late springndashearly summer during the last six centuries
(Christie et al 2009) Grey shadow indicating higher precipitation periods
From 1550 and 1810 CE (unit 2a) and during the Colonial period several peaks
of δ15N could be further related not only to high precipitation (Fig 10 grey shadow)
but also pulses of enhanced N input from the watershed linked to human land use
In 1550 CE Juan Cuevas was granted lands and indigenous workers under the
encomienda system for agricultural and mining development of the Vichuqueacuten
village (Ramiacuterez and Vidal 1985) Historical documents show that around 1579 CE
the Vichuqueacuten watershed was occupied by indigenous communities dedicated to
wheat plantations and vineyards wood extraction and gold mining (Odone 1998)
The introduction of the Spanish agricultural system implied not just a change in the
types of crops used (from quinoa to vineyards and wheat) but also a clearing of
native species for the continuous increase of agricultural surface and wood
extraction During the Colonial period wheat from Vichuqueacuten was exported to Peru
(Ramiacuterez and Vidal 1985) Armesto et al (2009) indicate that during the XVIII and
XIX centuries the extraction of wood for mining operations was important enough to
cause extensive loss of native forests The independence and instauration of the
Chilean Republic did not change this prevailing system Increases in the
contributions of N to the lake during the second half of the XIX century (peaks in
δsup1⁵N ca 1870 CE) could be due to expanding farm land dedicated for wheat
101
production and increased commercial trade with California and Canada (Ramiacuterez
and Vidal 1985)
In contrast LUCC in the last century are clearly related to the development of
large-scale tree plantations (Fig 7) The δsup1⁵N record reaches the lowest values of
the entire sequence in the last few decades (Fig 10) A marked increase in lake
productivity NT concentration and decreasing sediment input is synchronous (unit
1 Fig 4) with trees replacing meadows shrublands and areas with native forests
(Fig 7 and 8) The implementation of the lsquoForest Lawrsquo in 1974 had a stronger impact
on the landscape and lake ecosystem dynamics than the impacts of ongoing climate
change in the region which is much more recent (Garreaud et al 2018) although
the prevalence of hot dry summers seen over the last decade would also be
associated with low δsup1⁵N values and increase of NT (Fig 10) Heavy metals ratios
(FeTi CoTi and PbTi) show notable increases in lake sediments from 1992 to 2011
CE (Fig 4) Although this could be related to mining in the El Maule region the
closest mines are 60 Km away (Pencahue and Romeral) so local factors related to
shoreline urbanization for the summer homes and an increase in tourist activity
could also be a major factor
6 Conclusions
The N isotope signal in the watershed depends on the rates of exchange
between vegetation and soil cover (Fig 11) Plants preferentially uptake 14N and the
underlying soils become enriched in 15N especially when the terrestrial ecosystem
is N-limited andor significant N loss occurs (ie denitrification andor ammonia
volatilization) LUCC in the Lago Vichuqueacuten watershed have heavily modified the
links between terrestrial and aquatic ecosystems with agriculture practices
102
contributing more N to the lake than tree plantations or native forests In situ lake
processes can also fractionate N isotopes An increase of N-fixing species results
in OM depleted in 15N which results in POM with lower δsup1⁵N values during these
periods During winter phytoplankton is typically enriched in 15N due to the
decreased abundance of N-fixing species and increased N input from the
watershed This in turn generates the highest δsup1⁵NPOM values in Lago Vichuqueacuten
Periods of elevated productivity tend to deplete the dissolved inorganic N of 14N
resulting in even higher δ15N values
Our study illustrates the usefulness of δsup1⁵N as a tool for reconstructing the past
influence of LUCC on N availability in lake ecosystems To constrain the relative
roles of the diverse forcing mechanisms that can alter N cycling in mediterranean
ecosystems all main components of the N cycle should be monitored seasonally
(or monthly) including the measurements of δ15N values in land samples
(vegetation-soil) as well as POM
103
Figure 11 Summary of human and environmental factors controlling the δ15N
values of lake sediments Particulate organic matter(POM) δ15N values in
mediterranean lakes are driven by N input from the watershed that in turn depend
on land use and cover changes (ie forest plantation agriculture) andor seasonal
changes in N sources andor lake ecosystem processes (ie bioproductivity redox
condition denitrification) Arrows indicate the direction of N fluxes (inputoutput from
the N cycle) N cycle processes that deplete lake sediments of 15N are shown in
blue whereas those that enrich sediments in 15N are shown in red
104
Supplementary material
Figure S1 Pearson correlate coefficient between geochemical variables in core
VIC13-2B Positive and large correlations are in blue whereas negative and small
correlations are in red (p valuelt0001)
Figure S2 Principal Component Analysis of geochemical elements from core
VIC13-2B
105
Table S1 Lago Vichuqueacuten radiocarbon samples
RADIOCARBON
LAB CODE
SAMPLE
CODE
DEPTH
(m)
MATERIAL
DATED
14C AGE ERROR
D-AMS 029287
VIC13-2B-
1 043 Bulk 1520 24
D-AMS 029285
VIC13-2B-
2 085 Bulk 1700 22
D-AMS 029286
VIC13-2B-
2 124 Bulk 1100 29
Poz-63883 Chill-2D-1 191 Bulk 945 30
D-AMS 001133
VIC11-2A-
2 201 Bulk 1150 44
Poz-63884
Chill-2D-
1U 299 Bulk 1935 30
Poz-64089
VIC13-2D-
2U 463 Bulk 1845 30
Poz-64090
VIC13-2A-
3U 469 Bulk 1830 35
D-AMS 010068
VIC13-2D-
4U 667 Bulk 2831 25
Poz-63886
VIC13-2D-
4U 719 Bulk 3375 35
106
D-AMS 010069
VIC13-2D-
5U 775 Bulk 3143 27
Poz-64088
VIC13-2D-
5U 807 Bulk 3835 35
D-AMS-010066
VIC13-2D-
7U 1075 Bulk 6174 31
Poz-63885
VIC13-2D-
7U 1197 Bulk 6440 40
Poz-5782 VIC13-15 DIC 180 25
Table S2 Loadings of the trace chemical elements used in the PCA
Elementos PC1 PC2 PC3 PC4
Zr 0922 0025 -0108 -0007
Zn 0913 -0124 -0212 0001
Rb 0898 -0057 -0228 0016
K 0843 0459 0108 0113
Ti 0827 0497 0060 -0029
Al 0806 0467 0080 0107
Si 0803 0474 0133 0136
Y 0784 -0293 -0174 0262
V 0766 0455 0090 -0057
Br 0422 -0716 -0045 0226
Ca 0316 -0429 0577 0489
Sr 0164 -0420 0342 -0182
Cl 0151 -0781 -0397 0162
107
Mn -0121 -0091 0859 0095
S -0174 -0179 -0051 0714
Pb -0349 0414 -0282 0500
Fe -0700 0584 -0023 0280
Co -0704 0564 -0107 0250
Table S3 Stable Isotope values from vegetation of Lago Vichuqueacutenacutes watershed
Taxa Classification δsup1⁵N δsup1sup3C CN
molar
Poaceae Meadow 1216 -2589 3602
Juncacea Meadow 1404 -2450 3855
Cyperaceae Meadow 1031 -2596 1711
Taraxacum
officinale Meadow 836 -2400 2035
Poaceae Meadow 660 -2779 1583
Poaceae Meadow 453 -2813 1401
Poaceae Meadow 966 -2908 4010
Juncus Meadow 1247 -2418 3892
Poaceae Meadow 747 -3177 6992
Poaceae Meadow 942 -2764 3147
Poaceae Meadow 1479 -2634 2895
Poaceae Meadow 1113 -2776 1795
Poaceae Meadow 2215 -2737 7971
Poaceae Meadow 1121 -2944 2934
Poaceae Meadow 638 -3206 1529
108
Macrophytes Macrophytes 886 -3044 2286
Macrophytes Macrophytes 1056 -2720 2673
Macrophytes Macrophytes 769 -3297 1249
Macrophytes Macrophytes 967 -2763 1442
Macrophytes Macrophytes 959 -2670 2105
Macrophytes Macrophytes 334 -2728 1038
Acacia dealbata
Introduced
species 656 -2696 1296
Acacia dealbata
Introduced
species 487 -2941 1782
Acacia dealbata
Introduced
species 220 -2611 3888
Luma apiculata Native species 433 -2542 4135
Luma apiculata Native species 171 -2664 7634
Luma apiculata Native species -001 -2736 6283
Luma apiculata Native species 029 -2764 6425
Azara sp Native species 159 -2868 8408
Azara sp Native species 101 -2606 2885
Baccharis concava Native species 104 -2699 5779
Baccharis concava Native species 265 -2488 4325
Baccharis concava Native species 287 -2562 7802
Baccharis concava Native species 427 -2781 5204
Baccharis linearis Native species 190 -2610 4414
Baccharis linearis Native species 023 -2825 5647
109
Peumus boldus Native species 042 -2969 6327
Peumus boldus Native species 205 -2746 4110
Peumus boldus Native species 183 -2743 6293
Chusquea quila Native species 482 -2801 4275
Poaceae meadow 217 -2629 7214
Lobelia sp Native species 224 -2645 3963
Lobelia sp Native species -091 -2565 4538
Aristotelia chilensis Native species -035 -2785 5247
Aristotelia chilensis Native species -305 -2889 2305
Aristotelia chilensis Native species 093 -2836 5457
Chusquea quila Native species 173 -2754 3534
Chusquea quila Native species 045 -2950 6739
Quillaja saponaria Native species 223 -2838 9385
Scirpus meadow 018 -2820 7115
Sophora sp Native species -184 -2481 2094
Sophora sp Native species -181 -2717 1721
Pinus radiata
Introduced
trees 1581 -2602 3679
Pinus radiata
Introduced
trees 1431 -2784 4852
Pinus radiata
Introduced
trees -091 -2708 9760
Pinus radiata
Introduced
trees 153 -2568 3470
110
Salix sp
Introduced
trees 632 -2878 1921
LITERATURE CITED
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A historical
framework for land cover change in southwestern South America in the past 15000
years Land use policy 27 148ndash160
httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the next
carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014
httpsdoiorg101002eft2235
Blaauw M Christeny JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474
httpsdoiorg10121411-BA618
Bowman DMJS Cook GD 2002 Can stable carbon isotopes (δ13C) in soil
carbon be used to describe the dynamics of Eucalyptus savanna-rainforest
boundaries in the Australian monsoon tropics Austral Ecol
httpsdoiorg101046j1442-9993200201158x
Brahney J Ballantyne AP Turner BL Spaulding SA Otu M Neff JC 2014
Separating the influences of diagenesis productivity and anthropogenic nitrogen
deposition on sedimentary δ15N variations Org Geochem 75 140ndash150
httpsdoiorg101016jorggeochem201407003
111
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409
httpsdoiorg102134jeq20090005
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego R
Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and
environmental change from a high Andean lake Laguna del Maule central Chile
(36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the
Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from
tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Craine JM Elmore AJ Aidar MPM Dawson TE Hobbie EA Kahmen A
Mack MC Kendra K Michelsen A Nardoto GB Pardo LH Pentildeuelas J
Reich B Schuur EAG Stock WD Templer PH Virginia RA Welker JM
Wright IJ 2009 Global patterns of foliar nitrogen isotopes and their relationships
with cliamte mycorrhizal fungi foliar nutrient concentrations and nitrogen
availability New Phytol 183 980ndash992 httpsdoiorg101111j1469-
8137200902917x
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty
Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-
010-9453-1
Diebel M Vander Zanden MJ 2012 Nitrogen stable isotopes in streams stable
isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19
1127ndash1134 httpsdoiorg10189008-03271
112
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in freshwater
wetlands record long-term changes in watershed nitrogen source and land use SO
- Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash
2916
Fariacuteas L Castro-Gonzales M Cornejo M Charpentier J Faundez J
Boontanon N Yoshida N 2009 Denitrification and nitrous oxide cycling within the
upper oxycline of the oxygen minimum zone off the eastern tropical South Pacific
Limnol Oceanogr 54 132ndash144
Farquhar GD Ehleringer JR Hubick KT 1989 Carbon Isotope Discrimination
and Photosynthesis Annu Rev Plant Physiol Plant Mol Biol
httpsdoiorg101146annurevpp40060189002443
Farquhar GD OrsquoLeary MH Berry JA 1982 On the relationship between
carbon isotope discrimination and the intercellular carbon dioxide concentration in
leaves Aust J Plant Physiol httpsdoiorg101071PP9820121
Fogel ML Cifuentes LA 1993 Isotope fractionation during primary production
Org Geochem httpsdoiorg101007978-1-4615-2890-6_3
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A
Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-
resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS)
implications for past sea level and environmental variability J Quat Sci 32 830ndash
844 httpsdoiorg101002jqs2936
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924
httpsdoiorg104319lo20095430917
113
Gerhart LM McLauchlan KK 2014 Reconstructing terrestrial nutrient cycling
using stable nitrogen isotopes in wood Biogeochemistry 120 1ndash21
httpsdoiorg101007s10533-014-9988-8
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and oxygen
isotope fractionation during dissimilatory nitrate reduction by denitrifying bacteria 53
2533ndash2545 httpsdoiorg10230740058342
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater
eutrophic lake Limnol Oceanogr 51 2837ndash2848
httpsdoiorg104319lo20065162837
Harris D Horwaacuteth WR van Kessel C 2001 Acid fumigation of soils to remove
carbonates prior to total organic carbon or CARBON-13 isotopic analysis Soil Sci
Soc Am J 65 1853 httpsdoiorg102136sssaj20011853
Houmlgberg P 1997 Tansley review no 95 natural abundance in soil-plant systems
New Phytol httpsdoiorg101046j1469-8137199700808x
Kylander ME Ampel L Wohlfarth B Veres D 2011 High-resolution X-ray
fluorescence core scanning analysis of Les Echets (France) sedimentary sequence
New insights from chemical proxies J Quat Sci 26 109ndash117
httpsdoiorg101002jqs1438
Lara A Solari ME Prieto MDR Pentildea MP 2012 Reconstruccioacuten de la
cobertura de la vegetacioacuten y uso del suelo hacia 1550 y sus cambios a 2007 en la
ecorregioacuten de los bosques valdivianos lluviosos de Chile (35o - 43o 30acute S) Bosque
(Valdivia) 33 03ndash04 httpsdoiorg104067s0717-92002012000100002
Lehmann M Bernasconi S Barbieri A McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during
114
simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66
3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007
Liu G Li M An L 2007 The Environmental Significance Of Stable Carbon
Isotope Composition Of Modern Plant Leaves In The Northern Tibetan Plateau
China Arctic Antarct Alp Res httpsdoiorg1016571523-0430(07-505)[liu-
g]20co2
Marzecovaacute A Mikomaumlgi A Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54
httpsdoiorg103176eco2011105
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash
1643 httpsdoiorg1011770959683613496289
Meyers PA 2003 Application of organic geochemistry to paleolimnological
reconstruction a summary of examples from the Laurention Great Lakes Org
Geochem 34 261ndash289 httpsdoiorg101016S0146-6380(02)00168-7
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland
Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Odone C 1998 El Pueblo de Indios de Vichuqueacuten siglos XVI y XVII Rev Hist
Indiacutegena 3 19ndash67
Pausch J Kuzyakov Y 2018 Carbon input by roots into the soil Quantification of
rhizodeposition from root to ecosystem scale Glob Chang Biol
httpsdoiorg101111gcb13850
115
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98
httpsdoiorg1011772053019614564785
Sun W Zhang E Jones RT Liu E Shen J 2016 Biogeochemical processes
and response to climate change recorded in the isotopes of lacustrine organic
matter southeastern Qinghai-Tibetan Plateau China Palaeogeogr Palaeoclimatol
Palaeoecol httpsdoiorg101016jpalaeo201604013
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of
different trophic status J Paleolimnol 47 693ndash706
httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl
httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M 2009 High-resolution quantitative climate
reconstruction over the past 1000 years and pollution history derived from lake
sediments in Central Chile Philos Fak PhD 246
Woodward C Chang J Zawadzki A Shulmeister J Haworth R Collecutt S
Jacobsen G 2011 Evidence against early nineteenth century major European
induced environmental impacts by illegal settlers in the New England Tablelands
south eastern Australia Quat Sci Rev 30 3743ndash3747
httpsdoiorg101016jquascirev201110014
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K Yeager
KM 2016 Different responses of sedimentary δ15N to climatic changes and
116
anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau
J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
117
DISCUSION GENERAL
El Nitroacutegeno (N) es un elemento clave que controla la dinaacutemica y
funcionamiento de ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al
1997) Pese a que el N (N2) es muy abundante en la atmoacutesfera (78) es escaso
en los ecosistemas pues la mayoriacutea de la biota no puede acceder a su forma
molecular (Galloway et al 1995 Zaehle et al 2013) En este sentido la entrada
natural de N en los ecosistemas es viacutea fijacioacuten bioloacutegica del N2 por bacterias que lo
convierten en compuestos bioloacutegicamente disponibles (Battye et al 2017) Debido
a que la fijacioacuten es un proceso energeacuteticamente costoso los ecosistemas
comuacutenmente estaacuten limitados por N (Vitousek and Howarth 1991) Los LUCC
contribuyen al incremento del N disponible y son una de las principales causas de
eutroficacioacuten y contaminacioacuten de los cuerpos de agua (Woodward et al 2012)
En Chile central los LUCC principalmente relacionados con las actividades
agriacutecolas y forestales han causado grandes cambios en el ciclo del N debido al
118
reemplazo de la cobertura natural por un monocultivo y al uso de fertilizantes que
modifican los aportes de MO y N a los cuerpos de agua El programa de
estabilizacioacuten de laderas impulsada por el gobierno (Albert 1900) y la ley forestal
de 1974 han fomentado la forestacioacuten con plantaciones de Pinus radiata y
Eucalyptus globulus y en el caso de Lago Vichuqueacuten estas actualmente cubren maacutes
del 60 de su cuenca (Cap2 Fig 5) Ademaacutes en Laguna Matanzas la
sobreexplotacioacuten del agua en conjunto con el cambio climaacutetico actual el que ha
conducido a una de las megasequiacuteas maacutes severas en las uacuteltimas deacutecadas
(Garreaud et al 2017) generoacute una combinacioacuten letal que secoacute el lago en tan solo
10 antildeos (Cap1 Fig1) Las evidencias obtenidas a partir de estos dos lagos
permiten identificar las huellas del Antropoceno en Chile central basadas en el
registro sedimentario lacustre
La transferencia de N desde los ecosistemas terrestres a acuaacuteticos es un
proceso natural con controles bioacuteticos climaacuteticos y geoloacutegicos El enlace
hidroloacutegico entre los ecosistemas terrestres y acuaacuteticos hace que el estado troacutefico
de los lagos esteacute vinculado al de su cuenca (McLauchlan et al 2013) En Chile
central se sabe poco del impacto de los LUCC sobre el ciclo de nutrientes en los
ecosistemas lacustres aunque al menos desde la colonizacioacuten espantildeola se tienen
registros de influencia humana en las cuencas Durante la colonia espantildeola
Laguna Matanzas fue una hacienda ganadera que exportaba productos caacuternicos al
Peruacute (Contreras-Lopez et al 2014) A su vez en el Lago Vichuqueacuten se realizaban
extensas actividades agriacutecolas hasta comienzos del s XX como por ejemplo
cultivos de vid y trigo ademaacutes de extraccioacuten de madera y mineriacutea de oro (Odone
1997) En las uacuteltimas deacutecadas los LUCC han estado vinculados principalmente con
el desarrollo forestal al compaacutes de una estrategia gubernamental de fomentar con
119
incentivos financieros (Ley de Bosques DFL nordm265 de 1931 y DFL 701 de 1974)
esta actividad El incremento de la superficie forestal es especialmente fuerte en
ambos lagos a partir de la deacutecada de 1980 y hasta 2016 (Laguna de Matanzas 0-
17 Lago Vichuqueacuten 1-66) Este aumento habriacutea sido en detrimento del bosque
nativo y los pastizales (Cap 1 Fig 5 y Cap 2 Fig 8) El incremento de la superficie
forestal generariacutea una disminucioacuten de los aportes de sedimentos y nutrientes al lago
y en este sentido un cambio de estado en los flujos de N (e g tipping points) que
a su vez ha quedado registrado en la sentildeal isotoacutepica (δ15N) y la acumulacioacuten de
MO en los sedimentos lacustres
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central
Los sedimentos lacustres son verdaderos ldquosumiderosrdquo que pueden llegar a
registrar lo que ocurre en cuanto a la historia ambiental de su cuenca En esta tesis
se utilizan los isoacutetopos estables de N (15N y 14N) en sedimentos lacustres para
reconstruir los cambios en la disponibilidad de N en el tiempo y establecer la
magnitud de impacto generado por actividades humanas El fraccionamiento
cineacutetico en los lagos es mediado por procesos bioloacutegicos que mediante la
asimilacioacuten de N genera un producto (eg fitoplancton) maacutes ldquoligerordquo (valores maacutes
bajos de 15N) que su remanente (eg DIN) (Fogel y Cifuentes 1993) Sin embargo
en periacuteodos en que los lagos estaacuten limitados por N el fitoplancton discrimina menos
y la MO resultante queda enriquecida en 15N (Fig 1 a y b) Ademaacutes la
desnitrificacioacuten acentuada en lagos de fondo anoacutexicos (eg Laguna Matanzas
entre 1800 y 1940 Cap1 Fig 6) conduce a un enriquecimiento en 15N de los
sedimentos y de la columna de agua Esta informacioacuten queda reflejada en la MO
120
de los sedimentos lacustres lo que permite conocer la disponibilidad del N en el
tiempo a partir de las variaciones de 15N
En esta tesis analizamos la MO de los sedimentos lacustres para reconstruir
la influencia de los LUCC en los aportes de sedimentos y N al lago En teacuterminos de
asimilacioacuten de N se puede distinguir entre dos grupos principales de productores
primarios que componen el POM (Fig1)
1 Los que asimilan el DIN compuesto por amonio nitratos y nitritos Aquiacute el
δ15N resultante fluctuaraacute en torno a valores maacutes positivos en la medida que
la productividad se incrementa o no hay reposicioacuten del DIN (Fig 1)
2 Los fijadores de N Dado que es un proceso energeacuteticamente costoso en
ambientes que no estaacuten limitados por N muchas veces son excluiacutedas
competitivamente por el resto del fitoplancton Si el DIN queda agotado por
el incremento de la productividad (o bien si el ambiente estaacute limitado ya sea
por N o por otros nutrientes) los fijadores de N pueden florecer y la MO que
se genera a partir de ello tendraacute un δ15N con valores en torno a 0permil
De este modo la MO en los sedimentos lacustres dependeraacute de la
composicioacuten de productores primarios (ie fijadores vs asimiladores del DIN)
ademaacutes de los aportes desde la cuenca tanto de MO como del N de la cuenca que
pasa a formar parte del DIN (eg fertilizantes estieacutercol) (Fig 1)
La MO de los lagos estudiados en esta tesis ha sido analizada a partir de
variables geoquiacutemicas isotoacutepicas frotis y estaacute compuesta principalmente por
diatomeas y MO amorfa A partir de los datos obtenidos en esta tesis los valores
de δ15N aumentan cuando hay una mayor cobertura agriacutecola o pastizales A su vez
tiende a disminuir cuando aumenta la cobertura arboacuterea independiente si esta es
por plantaciones forestales o por bosque nativo
121
Por otra parte la dinaacutemica del lago tambieacuten imprime caracteriacutesticas
especiales en el POM observaacutendose variaciones estacionales en los valores
δ15NPOM Durante la estacioacuten caacutelida y seca se observan valores maacutes bajos que
durante la estacioacuten friacutea y huacutemeda posiblemente relacionado con el incremento de
la contribucioacuten de especies fijadoras de N (Lago Vichuqueacuten Fig 9 Cap 2) durante
el verano En la estacioacuten friacutea y huacutemeda el DIN seriacutea la principal fuente de N Las
mayores entradas de MO y N terrestre debidos a un incremento del lavado de la
cuenca como consecuencia de las precipitaciones y la mineralizacioacuten de la MO
podriacutean estimular la produccioacuten de MO lacustre por crecimiento del fitoplancton
Como consecuencia se observan tendencias decrecientes de los valores de
δ15N en la MO sedimentaria cuando la productividad en el lago estaacute relacionada
con especies fijadoras de N mientras que el δ15N muestra aumentos cuando la
productividad del lago estaacute asociada principalmente al consumo del DIN pero
tambieacuten cuando predominan procesos de perdida de N (eg desnitrificacioacuten Fig
1)
Hemos identificado tres fases en la dinaacutemica del δ15N para ambos lagos
Entre 1800 y 1940 CE la cuenca de laguna de Matanzas estaba dominada por
actividades agriacutecolas y ganaderas (δ15Nsuelo gt 4permil Cap 1 Fig 4 y 6) Las entradas
de sedimentos y MO desde la cuenca son altas (δ13C lt -209permil ZrTi ~029 AlTi
~013) y coincide con un clima maacutes huacutemedo y friacuteo (Von Gunten et al 2009
Christiansen et al 2011) El lago teniacutea un nivel de agua maacutes profundo dominado
por condiciones anoacutexicas (MnFe ~0013) que contribuiriacutean a mayores valores de
δ15N en la columna de agua y por ende de los sedimentos acumulados en el fondo
debido a la peacuterdida diferencial de 14N viacutea desnitrificacioacuten (Fig 7 Cap1) La baja
122
produccioacuten de MO lacustre (BrTi ~002 TOC~17) coincide con un bajo contenido
de NT (~478 μg) y valores maacutes positivos de δ15N (gt 4permil)
La actividad agriacutecola tambieacuten dominaba en la cuenca del Lago Vichuqueacuten
durante este periacuteodo (δ15Nsuelo gt4permil Cap2 Fig 6) El aporte sedimentario desde la
cuenca es alto (AlTi ~029 ZrTi ~032) y fue favorecido por mayores
precipitaciones a las actuales (Christiansen et al 2011) El Lago Vichuqueacuten es un
lago estratificado anoacutexico (MnFe ~002) y profundo (~30 m) por lo que la
desnitrificacioacuten juega un rol importante en el enriquecimiento de la MO
sedimentaria La baja produccioacuten de MO (BrTi ~007 TOC ~12) de origen
lacustre (δ13C ~ -268permil) coincide con un bajo contenido de NT (~309μg +6) y
valores maacutes positivos de δ15N (56permil +03)
Durante esta fase en ambos lagos los aportes de N de la cuenca parecen
ser los principales responsables en las fluctuaciones del δ15N Esta sentildeal podriacutea
estar amplificada por bajas temperaturas (que afectan la productividad lacustre) y
altas precipitaciones que favorecen el lavado de la cuenca y aumentan el aporte de
sedimentos y MO desde la cuenca predominantemente agriacutecola
Entre 1940 y 1980 CE en Laguna Matanzas se observa un incremento en
la acumulacioacuten de MO algal (TOC ~2 +1 δ13C ~-230+09permil) El ambiente
deposicional es ligeramente menos turbulento que el anterior (AlTi ~011+001
ZrTi ~03+001) y el lago estaacute en una fase de transicioacuten hacia condiciones maacutes
oacutexicas (MnFe ~002+0004) y maacutes productivas (BrTi ~004+0007) A su vez δ15N
tiende hacia valores maacutes negativos (δ15N ~30permil+03) mientras que el NT aumenta
oscilando en antifase con el δ15N
En Lago Vichuqueacuten en cambio se observa un ligero incremento en la
acumulacioacuten de MO (TOC ~13 +02) de origen terrestre (δ13C ~-27permil +04) La
123
productividad es similar al periacuteodo anterior (BrTi ~007+003) y el ambiente
deposicional es menos turbulento (AlTi ~02 +009 Zrti ~033 +007) El δ15N y el
NT oscilan en torno a valores ligeramente maacutes bajos (δ15N ~51permil +04 NT ~292μg
+6) Este periacuteodo se caracteriza por ser maacutes caacutelido que el anterior lo que
posibilitariacutea un incremento en la productividad lacustre en Laguna Matanzas pero
que no es observada en el Lago Vichuqueacuten
Entre 1980 y la actualidad en Laguna Matanzas habriacutea un incremento de la
acumulacioacuten de MO (TOC ~59 + 3) asociada a un incremento en la productividad
del lago (BrTi ~009+005) y a los aportes de MO terrestres (δ13C ~-24 + 2permil) El
lago se vuelve maacutes oacutexico (Mn ~0003 +0006) y las entradas de sedimento
disminuyen (AlTi~ 009 +002) El δ15N tiende a valores maacutes negativos (δ15N ~13permil
+1) y el NT aumenta de 20 a 55 μg (NT ~346 + 9 μg) tomando valores sin
precedentes en todo el registro En los sedimentos lacustres del Lago Vichuqueacuten
tambieacuten se observa un incremento de la acumulacioacuten de MO (TOC ~17 + 05)
asociada a un incremento en la productividad del lago (BrTi ~01 + 003) y las
entradas de sedimento desde la cuenca disminuyen (AlTi ~005 + 005) El δ15N
(~43 + 05permil) tiende a valores maacutes negativos y el NT incrementa de 20 a 55 μg (NT
~346 + 9 μg) oscilando en antifase
Durante esta fase en ambos lagos se observa un aumento en la
acumulacioacuten de MO sedimentaria un incremento en el NT y valores maacutes negativos
de δ15N que coincide con el incremento de la superficie forestal de las cuencas
(Laguna Matanzas gt17 Lago Vichuqueacuten gt60)
124
Figura 2 Comparacioacuten de los cambios del ciclo del N entre Laguna Matanzas y
Lago Vichuqueacuten con los procesos biogeoquiacutemicos contrastantes en rojo En L
Matanzas las condiciones oacutexicas podriacutean estar favoreciendo la nitrificacioacuten del
amonio En Vichuqueacuten las condiciones anoacutexicas favorecen un aumento en δ15N de
la MO en los sedimentos por desnitrificacioacuten y mineralizacioacuten
Los ambientes mediterraacuteneos en el que los lagos del presente estudio se
encuentran insertos estaacuten caracterizados por una estacioacuten seca prolongada y las
precipitaciones ocurren en eventos puntuales alcanzando altos montos
pluviomeacutetricos en poco tiempo Las lluvias favorecen un lavado de la cuenca la
perdida de N y otros nutrientes Por lo que es esperable que los sedimentos del
lago reflejen mejor los valores isotoacutepicos de los suelos de la cuenca durante los
periacuteodos con altos montos pluviomeacutetricos En el Lago Vichuqueacuten por ejemplo el
125
POM superficial (2 y 5 m de profundidad) despliega valores isotoacutepicos maacutes
positivos en invierno presumiblemente como resultado de mayores aportes de MO
y N de su cuenca (Fig 1b) En ambos lagos los valores maacutes positivos en los
sedimentos lacustres ocurren durante periacuteodos con mayores montos pluviomeacutetricos
(Cap1 Fig 6 y Cap 2 Fig 12)
Ademaacutes del efecto de la precipitacioacuten en el aporte de nutrientes al lago en
esta tesis hemos encontrado que los mayores aportes de N desde la cuenca se dan
cuando la cobertura vegetal del suelo es menor En Laguna Matanzas el desarrollo
de la actividad ganadera desde la llegada de los espantildeoles a la cuenca habriacutea
incrementado los aportes de N al lago Los valores de δ15N en los sedimentos
lacustres depositados durante este periacuteodo son los maacutes positivos de todo el registro
(~4permil) A su vez en Lago Vichuqueacuten los valores isotoacutepicos maacutes positivos se
registran entre 1300 y 1900 CE que coinciden con el mayor desarrollo de
actividades agriacutecola y de extraccioacuten de maderas de la cuenca (Fig 10 Cap 2)
Los recientes LUCC (a partir de 1975) en las cuencas de Laguna Matanzas
y Lago Vichuqueacuten estaacuten caracterizados por un incremento de la superficie forestal
y agriacutecola en reemplazo de la cobertura de pastizal (Laguna Matanzas) y bosque
nativo (Lago Vichuqueacuten) Los valores de δ15N en los testigos sedimentarios fueron
maacutes positivos cuanto mayor la superficie agriacutecola de la cuenca y maacutes negativos
cuanto mayor la superficie forestal Aunque por los alcances de esta tesis no
podemos ser concluyentes respecto del origen del NT (aloacutectonoautoacutectono) parece
ser que esta maacutes ligado a los aportes de la cuenca pese a la disminucioacuten del aporte
sedimentario observado en ambos lagos Las plantaciones forestales a diferencia
del bosque nativo son fertilizadas con una mezcla de N P y K (Toral et al 2005)
Por lo que pese a la disminucioacuten del aporte sedimentario la concentracioacuten de
126
nutrientes en el aporte al lago puede verse incrementada por el tipo de manejo
forestal con respecto al bosque nativo
Los resultados del primer capiacutetulo demuestran que 1) las plantaciones
forestales yo bosque nativo retiene maacutes sedimentos y MO mientras que el uso de
suelo agriacutecola y los pastizales destinados al desarrollo de la actividad ganadera lo
libera 2) Al parecer la desnitrificacioacuten es el proceso natural maacutes importante de
perdida de N en lagos costeros y favorece el enriquecimiento de 15N tanto en la
columna de agua (ie 15N-DIN) como en los sedimentos una vez enterrados y 3) la
desnitrificacioacuten depende de los cambios en las condiciones REDOX en el lago La
oxigenacioacuten en lagos poco profundos -como Laguna Matanzas- es altamente
fluctuante a escalas decadal-centenal depende de la profundidad de la laacutemina de
agua y de la variabilidad de la precipitacioacuten En este sentido en Laguna Matanzas
habriacutea condiciones anoacutexicas en ambientes cuando el lago alcanza niveles maacutes
altos Estas condiciones se observan entre 1820 y 1940 CE y son coetaacuteneas con
episodios maacutes huacutemedos y friacuteos (von Gunten et al 2009 Christie et al 2009) pero
tambieacuten con una fuerte actividad ganadera en la cuenca
Las fuentes variables de N y su fluctuacioacuten en el tiempo hacen necesario
contar con un anaacutelogo moderno que permite usar al δ15N de los sedimentos
lacustres como un indicador indirecto de los cambios en la disponibilidad de N en
el tiempo Por ello en el segundo capiacutetulo integramos muestras de suelo-
vegetacioacuten y un monitoreo de POM de la columna de agua en verano e invierno La
composicioacuten isotoacutepica de POM estariacutea reflejando cambios de la fuente de N (fijacioacuten
vs consumo del DIN) que variacutea durante el antildeo hidroloacutegico Durante el verano la
mayor temperatura y horas-luz favoreceriacutean las entradas de N al lago viacutea fijacioacuten
bioloacutegica de N2 atmosfeacuterico resultando en un incremento de la MO con un valor
127
isotoacutepico bajo (POM verano~5permil Fig 1b) El N2 (δ15N = 0permil) es fijado praacutecticamente
sin fraccionamiento isotoacutepico generando un δ15N de la OM cercano a 0permil (Leng et
al 2006) En invierno las precipitaciones pueden estar favoreciendo un incremento
en la entrada de nutrientes al lago El DIN de la columna de agua llega a contener
valores de δ15N maacutes altos reflejando estos mayores aportes de la cuenca y el POM
del Lago Vichuqueacuten queda enriquecido en 15N (δ15N ~11permil Cap 2 Fig 9) Estas
variaciones estacionales pueden estar evidenciando un cambio en la composicioacuten
de especies de POM desde especies fijadoras a especies que consumen el N de la
columna de agua Sin embargo para corroborar esta informacioacuten seriacutea deseable
contar con el anaacutelisis de la composicioacuten de especies de las muestras de agua
extraidas
Entre los resultados maacutes importantes estaacuten los valores isotoacutepicos del suelo
y la biomasa representativa de la cuenca que incluye un listado de las especies
nativas de la zona que hasta ahora no se contaba (Cap 2 Tabla en material
suplementario) Las especies colectadas despliegan valores de δ15Nfoliar maacutes
positivos (plantaciones forestales y pastizal ~89permil) que el suelo (35permil) salvo por
las especies nativas las que oscilan en torno a valores cercanos al atmosfeacuterico
(~14permil) La razoacuten isotoacutepica 15N14N refleja la dinaacutemica de intercambio de N entre la
vegetacioacuten y el suelo (Koba et al 2002) La discriminacioacuten es positiva en la mayoriacutea
de los sistemas bioloacutegicos por lo que las plantas tienen valores de δ15N maacutes bajos
que el suelo (Evans et al 2001) como sucede con las especies nativas en Lago
Vichuqueacuten En consecuencia las plantas quedan empobrecidas en 15N y conducen
a un enriquecimiento en 15N del suelo sobretodo si no hay reposicioacuten de N por otras
viacuteas (eg fijacioacuten de N) Por lo tanto los valores maacutes negativos de δsup1⁵Nfoliar de las
especies nativas pueden estar relacionados con el consumo preferencial de 14N del
128
suelo donde la discriminacioacuten isotoacutepica de la vegetacioacuten contra el 15N conduce a
valores del δsup1⁵Nsuelo maacutes altos (Houmlgberg 1997) Por el contrario los valores maacutes
positivos de δsup1⁵Nfoliar de las plantaciones forestales y pastos con respecto al suelo
puede deberse por una parte que el suelo no cuenta con mecanismos naturales de
reposicioacuten de N salvo por la descomposicioacuten de la hojarasca que puede ser maacutes
lenta en Pinus radiata que en bosque nativo (Lusk et al 2001) y por otra por el alto
impacto de los aportes de N (y otros nutrientes) derivado de las actividades
humanas (eg uso de fertilizantes) en el suelo
El alcance maacutes significativo de esta tesis se relaciona con un cambio en la
tendencia maacutes negativa de δsup1⁵N y el incremento del NT a partir de 1940 y a partir
de 1970 CE en ambos lagos La causa maacutes probable de esta tendencia es el
reemplazo de la cobertura natural o preexistente a 1940 CE por plantaciones
forestales
En la figura 2 se observa una siacutentesis de los principales procesos que
afectan la sentildeal isotoacutepica de la MO en los sedimentos lacustres de L Matanzas y
L Vichuqueacuten La entrada de N y MO desde la cuenca depende del tipo de cobertura
Las plantaciones forestales y el bosque nativo retienen maacutes nutrientes y sedimentos
en la cuenca mientras que la agricultura los favorece Sin embargo las praacutecticas
de manejo que incluyen la aplicacioacuten de N (y otros nutrientes) pueden aportar maacutes
nutrientes al lago que la cobertra de bosque nativo Cuando las actividades
forestales cubren una mayor superficie de la cuenca (1940 en adelante) δsup1⁵N oscila
en valores maacutes negativos y el NT tiende a incrementarse en los sedimentos
lacustres Cabe destacar que los valores de δsup1⁵N observados en los testigos
sedimentarios a fines del s XX no tienen precedente durante la conquista y colonia
espantildeola o durante el resto del periodo de la Repuacuteblica
129
Figura 2 Modelo de transferencia de N entre los ecosistemas terrestres y
acuaacuteticos El δ15N de los sedimentos lacustres depende de la interaccioacuten entre los
aportes de la cuenca y porcesos ldquoin-lakerdquo Flechas indican la direccioacuten del flujo de
N En azul las salidasentradas de 14N en rojo las salidasentradas de 15N δ15N de
la interaccioacuten plantas-suelo depende del tipo de cobertura de suelo
130
CONCLUSIONES GENERALES
La transferencia de N entre cuencas y lagos es un factor de control del ciclo
del N en los lagos y queda reflejado en la sentildeal isotoacutepica de los sedimentos
lacustres (Leng et al 2006 McLauchlan et al 2007) En Laguna Matanzas el
suelo de las especies nativas y las plantaciones forestales despliegan valores de
δ15N maacutes bajos (~1permil) que los de Lago Vichuqueacuten (~35permil) Coincidentemente los
sedimentos lacustres de Laguna Matanzas oscilan con valores de δ15N maacutes bajos
(-15 a +45permil Fig 1) que los de Lago Vichuqueacuten (+3 a +8permil)
Por otra parte el estudio de la MO de los sedimentos lacustres ha permitido
reconstruir los cambios en los aportes de MO y N desde la cuenca Aunque no es
posible afirmar queacute tipo de N estaacute entrando al lago (eg Nitroacutegeno orgaacutenico e
inorganico) si podemos decir que los cambios en las fluctuaciones de δ15N son
coincentes con el crecimiento de la actividad forestal (1980 - hasta 2016 L
Matanzas 0-17 y L Vichuqueacuten 1-66) El δ15N fluctuacutea hacia valores maacutes
negativos Ademaacutes se observa un incremento del NT en los sedimentos lacustres
cuanto mayor es la superficie forestal
Entre 1800-1940 las cuencas eran fundamentalmente agriacutecolas y
ganaderas (Cap1 Fig 7 y Cap 2 Fig 12) el δ15N de los sedimentos lacustres
oscila con valores maacutes positivos (L Matanzas +40 plusmn 05permil y L Vichuqueacuten 56 plusmn
033permil Fig 1) hay una baja bioproductividad en el lago (BrTi ~002 TOC~17)
lo que coincide con un bajo contenido de NT (~478 μg) Este es un periacuteodo de altas
precipitaciones (Christie et al 2009) que tienden a favorecer el lavado de la cuenca
131
y con ello el aporte de MO sedimentos y nutrientes desde la cuenca tiende a verse
favorecido Aunque las principales actividades humanas en estas cuencas son
diferentes (ganaderiacutea en Laguna Matanzas Contreras-Lopez et al 2014
agricultura en Lago Vichuqueacuten Odone 1997) en ambos casos hubo un reemplazo
de la cobertura natural de bosques nativo que favorecioacute mayores aportes de N y
sedimentos desde la cuenca en un efecto sumado con el aumento de las
precipitaciones
A partir de 1980 CE en ambos lagos se observa una disminucioacuten en los
valores de δ15N y un incremento del NT que no tiene precedentes en todo el registro
y pese a que ambos lagos son limnologicamente muy diferentes En Lago
Vichuqueacuten el δ15N tiende a oscilar en antifase con el NT (Fig 1) En el caso de
Laguna Matanzas se observa una tendencia hacia valores maacutes negativos y a partir
de 1990s a valores maacutes positivos mientras que el NT tiende a incrementarse de
manera sostenida Ello podriacutea estar relacionado con el crecimiento de la actividad
forestal habriacutea una mayor retencioacuten de N y sedimentos en la cuenca ligado al
incremento de la superficie forestal Ademaacutes en el caso de L Matanzas el
incremento de la actividad agriacutecola (sumado al de las plantaciones forestales)
podriacutea expicar el cambio en la tendencia en la deacutecada de los 1990s
En el contexto de Antropoceno esta tesis nos permite identificar un gran
impacto de las actividades humanas en Chile central a partir de la deacutecada de 1940
y una Gran Aceleracoacuten a partir de la deacutecada de 1970s En el registro sedimentario
de los lagos en estudio se observa una gran acumulacioacuten de MO el δ15N oscila
hacia valores maacutes negativos y un gran incremento del NT y el crecimiento de la
actividad forestal parece ser la principal responsable Ademaacutes la sobreexplotacioacuten
132
del agua junto al cambio climaacutetico global ha generado una combinacioacuten letal para
los lagos costeros de Chile central
Figura 1 Comparacioacuten de las fluctuaciones de δ15N durante los uacuteltimos 300
antildeos en Laguna Matanzas y Lago Vichuqueacuten
133
Referencias
Battye W Aneja VP Schlesinger WH 2017 Earthrsquos Future Is nitrogen the next carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia
UC de V 2014 Elementos de la historia natural del An Mus Hist Natulas Vaplaraiso 27 51ndash67
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Evans RD Evans RD 2001 Physiological mechanisms influencing
plant nitrogen isotope composition 6 121ndash126 Fogel ML Cifuentes LA 1993 Isotope fractionation during primary
production Org Geochem 73ndash98 httpsdoiorg101007978-1-4615-2890-6_3 Galloway JN Schlesinger WH Ii HL Schnoor L Tg N 1995
Nitrogen fixation Anthropogenic enhancement-environmental response reactive N Global Biogeochem Cycles 9 235ndash252
Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J
Christie D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Koba K Koyama LA Kohzu A Nadelhoffer KJ 2003 Natural 15 N
Abundance of Plants Coniferous Forest httpsdoiorg101007s10021-002-0132-6
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos
Transactions American Geophysical Union httpsdoiorg1010292007EO070007 McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE
2007 Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
134
Phytologist N Oct N 2006 Tansley Review No 95 15N Natural Abundance in Soil-Plant Systems Peter Hogberg 137 179ndash203
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Turner BL Kasperson RE Matson PA McCarthy JJ Corell RW
Christensen L Eckley N Kasperson JX Luers A Martello ML Polsky C Pulsipher A Schiller A 2003 A framework for vulnerability analysis in sustainability science Proc Natl Acad Sci U S A httpsdoiorg101073pnas1231335100
Vitousek PM Aber JD Howarth RW Likens GE Matson PA
Schindler DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
Vitousek PM Howarth RW 1991 Nitrogen limitation on land and in the
sea How can it occur Biogeochemistry httpsdoiorg101007BF00002772 von Gunten L Grosjean M Rein B Urrutia R Appleby P 2009 A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 Holocene 19 873ndash881 httpsdoiorg1011770959683609336573
Xu R Tian H Pan S Dangal SRS Chen J Chang J Lu Y Maria
Skiba U Tubiello FN Zhang B 2019 Increased nitrogen enrichment and shifted patterns in the worldrsquos grassland 1860-2016 Earth Syst Sci Data httpsdoiorg105194essd-11-175-2019
Zaehle S 2013 Terrestrial nitrogen-carbon cycle interactions at the global
scale Philos Trans R Soc B Biol Sci 368 httpsdoiorg101098rstb20130125
6
AGRADECIMIENTOS
Quiero agradecer a mi tutor y mentor Dr Claudio Latorre por brindarme su
apoyo sin el cual no habriacutea logrado concluir esta tesis de doctorado Claudio tu
apoyo constante incentivo y el fijarme metas que a veces me pareciacutean imposibles
de alcanzar no solo han dado forma a esta tesis sino tambieacuten me ha hecho maacutes
exigente como cientiacutefica Claudio destacas no solo por ser un gran cientiacutefico si no
tambieacuten por tu gran calidad humana eres un gran ejemplo
Quiero agradecer Dr Blas Valero-Garceacutes por nuestras numerosas
conversaciones viacutea Skype que incluiacutean vacaciones y fines de semana para discutir
los resultados de la tesis y que han dado forma a esta investigacioacuten principalmente
al primer capiacutetulo Ademaacutes por haberme acogido como un miembro maacutes en el
laboratorio de Paleoambientes Cuaternarios durante las estancias que he realizado
en el transcurso de estos antildeos Blas eres un ejemplo para miacute conjugas ciencia de
calidad calidez y dedicacioacuten por tus estudiantes
A quienes han financiado mi doctorado la Comisioacuten Nacional de
Investigacioacuten Cientiacutefica y Tecnoloacutegica (CONICYT) con sus becas de manutencioacuten
doctoral (2013) gastos operacionales pasantiacutea (2016) postnatal (2017) y termino
de tesis doctoralrdquo (2013) A FONDECYT a traveacutes del proyecto 1160744 de C
Santoro Al Departamento de InvestigacioacutenAl Instituto de Ecologiacutea y Biodiversidad
(IEB) a traveacutes de del PIA financiamiento basal 170008 la Pontificia Universidad
Catoacutelica de Chile por la beca incentivo para tesis interdisciplinaria para doctorandos
(2015)
Agradezco a mis compantildeeros del laboratorio de Paleoecologiacutea y
Paleoclimatologiacutea Karla Matias Dani Carolina Mauricio y Pancho que han hecho
grato mi tiempo en el laboratorio Agradecimientos especiales a Carolina Matiacuteas y
Leo por acompantildearme a terreno
7
Agradezco a los miembros del laboratorio de Paleoambientes Cuaternarios
(IPE) en especial a Raquel Elena Fernando Ana Penelope Graciela Miguel
Sevilla Mariacutea y Miguel Bartolomeacute
Agradezco a los miembros de la comisioacuten Dr Joseacute Miguel Farintildea Dr Juan
Armesto y Dr Ricardo De Pol-Holz por la gran ayuda durante el desarrollo de mi
doctorado en especial por las correcciones finales de la tesis
Al departamento de Ecologiacutea en especial a Rosario Galaz Edita Luengo
Daniela Mora y Valeria Cavallero por su apoyo
A mis compantildeeros de doctorado Beleacuten Gallardo Pancho Diacuteaz Bryan Bularz
Bian Latorre Renato Borras Juan Carlos Hernaacutendez y Paulina Troncoso con
quienes estudieacute aprendiacute y compartimos muy buenos momentos durante los
primeros antildeos del doctorado
A Pablo Sarricolea mi compantildeero de vida Gracias por tu constante e
incondicional apoyo Sin ti durante este proceso todo habriacutea sido cuesta arriba
A mis padres Arturo y Malena gracias por su carintildeo apoyo y compromiso
Papaacute aunque no esteacutes a mi lado siempre te recuerdo con mucho carintildeo A mi
madre que ha sido un constante apoyo sobre todo en el cuidado de mis hijos xavi
y panchito
A mis hermanos Rodrigo y David por estar presentes durante toda esta
etapa Siempre con carintildeo y hermandad
A Rodrigo David Matiacuteas Jany Paola y Julio por cuidar a mis hijos con carintildeo
siempre que estuve ausente por el doctorado
8
ABREVIATURAS
N Nitroacutegeno (Nitrogen)
DIN Nitroacutegeno Inorgaacutenico Disuelto (Dissolved Inorganic Nitrogen)
C Carbono (Carbon)
TOC Carbono Inorgaacutenico Total (Total Organic Carbon)
TIC Carbono Inorgaacutenico Total (Total inorganic Carbon)
TC Carbono Total (Total Carbon)
TS Azufre Total (Total Sulfur)
LUCC Cambios de UsoCobertura del Suelo (Land Use and Cover Change)
OM Materia Orgaacutenica (Organic Matter)
POM Particulate Organic Matter (materia orgaacutenica particulada)
CE Common Era
BCE Before Common Era
Cal BP Calibrado en antildeos radiocarbono antes de 1950
ie id est (esto es)
e g Exempli gratia (por ejemplo)
9
RESUMEN
El ldquoAntropocenordquo se caracteriza por pertubaciones de origen humano que
conllevan impactos globales Por ejemplo los cambios de usocobertura del suelo
(LUCC) son conocidos por perturbar el ciclo del N a partir de la Revolucioacuten Industrial
pero especialmente desde la Gran Aceleracioacuten (1950 CE) en adelante Sin
embargo existen incertezas asociadas a la magnitud del impacto y su efecto
acoplado con los cambios climaacuteticos actuales (ie megasequiacutea disminucioacuten de las
precipitaciones) y acoplamientos con otros ciclos biogeoquiacutemicos (eg ciclo del
Carbono) Este impacto ha cambiado la disponibilidad de N en los ecosistemas
terrestres y acuaacuteticos y sus enlaces Los sedimentos lacustres contienen
informacioacuten de las condiciones paleoambientales del lago y su cuenca en el
momento en que se depositaron y el anaacutelisis de los isoacutetopos estables de N (δ15N)
en los sedimentos permiten reconstruir los cambios en la disponibilidad de N a
traveacutes del tiempo En este trabajo usamos una aproximacioacuten multiproxy que incluye
10
anaacutelisis sedimentoloacutegicos geoquiacutemicos e isotoacutepicos (δ15N δ13C) de los sedimentos
lacustres columna de agua y suelo-vegetacioacuten de la cuenca y la reconstruccioacuten de
los LUCC entre 1975 y 2016 a partir de imaacutegenes satelitales El objetivo de esta
tesis fue evaluar el rol de LUCC como principal control del ciclo del N en el sistema
cuenca-lago durante las uacuteltimas centurias en Chile central Nuestros principales
resultados muestran que valores maacutes positivos de δ15N en los sedimentos lacustres
estaacuten relacionados con grandes aportes de N de la cuenca que a su vez son
mayores cuanto mayor la superficie agriacutecola yo los pastizales mientras que tanto
las plantaciones forestales como los bosques favorecen la retencioacuten de nutrientes
en la cuenca (lo que se ve reflejado en un δ15N maacutes negativo) Esta tesis plantea
un cambio de estado en la dinaacutemica del N asociado a la introduccioacuten y expansioacuten
de las plantaciones forestales o bosques los que retienen el Nitroacutegeno en las
cuencas mientras que el clima juega un rol secundario
11
ABSTRACT
The Anthropocene is characterized by human disturbances at the global
scale For example changes in land use are known to disturb the N cycle since the
industrial revolution but especially since the Great Acceleration (1950 CE) onwards
This impact has changed N availability in both terrestrial and aquatic ecosystems
However there are some important uncertainties associated with the extent of this
impact and how it is coupled to ongoing climate change (ie megadroughts rainfall
variability and shifting stationarity) or to other nutrient cycles (e g carbon cycle)
Lake sediments contain paleoenvironmental information regarding the conditions of
the watershed and associated lakes and which the respective sediments are
deposited Nitrogen stable isotope analysis (δ15N) on lake sediments allows us to
reconstruct the changes in N availability through time Here we used a multiproxy
approach that uses sedimentological geochemical and isotopic analyses on
lacustrine sediments water column and soilvegetation from the watershed as well
12
as land use and cover change (LUCC) from 1975 to 2016 obtained from satellite
images The goal of this thesis is to evaluate the role of LUCC as the main driver for
N cycling in a coastal watershed system of central Chile over the last centuries Our
main results show that more positive δ15N values in lake sediments are related to
higher N contributions from the watershed which in turn increase with increased
agricultural andor pasture cover whereas either forest plantations or native forests
can favor nutrient retention in the watershed (δ15N more negative) This thesis
proposes that N dynamics are mainly driven by the introduction and expansion of
forest or tree plantations that retain nitrogen in the watershed whereas climate plays
a secondary role
13
INTRODUCCIOacuteN
El N es un elemento esencial para la vida y limita la productividad en
ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al 1997) Las actividades
humanas han tenido un profundo impacto sobre el ciclo del N global principalmente
a partir la revolucioacuten industrial (IPCC 2013 2007) Las concentraciones de N se
han incrementado por la fijacioacuten industrial de N atmosfeacuterico (viacutea el proceso Haber-
Bosch Erisman et al 2008) por el uso de fertilizantes sobre la base de N para
mejorar el rendimiento de los cultivos la produccioacuten ganadera y ademaacutes de los
cambios en el uso del suelo (abreviado en ingleacutes- LUCC) (Robertson y Vitousek
2009 Galloway et al 2008 Battye et al 2017 Xu et al 2019) Estas actividades
contribuyen significativamente al incremento de la disponibilidad bioloacutegica de N
cuyas consecuencias para los ecosistemas incluye la perdida de diversidad
modificacioacuten de la disponibilidad de nutrientes y acidificacioacuten de los suelos entre
otras Sin embargo se desconoce la cantidad destino y el alcance que ha tenido
14
el N movilizado entre los ecosistemas generado por la influencia de las actividades
humanas (Vitousek et al 1997)
La entrada natural de N en los ecosistemas terrestres y acuaacuteticos es viacutea
fijacioacuten atmosfeacuterica la que es llevada a cabo por microorganismos muchos de ellos
en relaciones simbioacuteticas (eg Rhizobium en leguminosas) y las algas (Vitousek et
al 1997 Battye et al 2017) Las principales salidas estaacuten relacionadas con la
desnitrificacioacuten volatilizacioacuten del amoniaco y por escurrimiento superficial y
subterraacuteneo (Galloway et al 2008 Diacuteaz et al 2016) En el caso de los sistemas
lacustres ademaacutes estaacuten la resuspencioacuten de N del fondo del lago los aportes desde
la cuenca (entradas de N) La depositacioacuten de MO en el fondo del lago que marca
la salida de N de la columna de agua Estas relaciones de intercambio de N tienen
un control geograacutefico (eg latitud exposicioacuten relieve etc) climaacutetico y biotico
(McLauchlan et al 2013) La alteracioacuten provocada por los LUCC no solo genera
las peacuterdidas de nutrientes del suelo por erosioacuten y escurrimiento superficial sino que
tambieacuten modifica la disponibilidad y transferencia de N entre los ecosistemas
terrestres y acuaacuteticos (Elbert et al 2012 McLauchlan et al 2013) Por ejemplo el
reemplazo de la vegetacioacuten nativa por la agricultura yo plantaciones forestales
altera el ciclo del N debido al despeje de los terrenos viacutea quema de la vegetacioacuten
pero tambieacuten debido al reemplazo de la vegetacioacuten preexistente por un
monocultivo (eg trigo pino) y el uso ineficiente de fertilizantes para mejorar el
rendimiento agriacutecola Estas actividades favorecen un incremento de los aportes de
N (y otros nutrientes) viacutea sistemas de drenajes a lagos los que actuacutean como
sumideros El incremento del N derivado de las actividades humanas tanto en los
ecosistemas terrestres como acuaacuteticos genera incertezas relacionados con la
trayectoria y la magnitud de la disponibilidad de N en ambos ecosistemas (Batye et
15
al 2017 Mclauchlan et al 2017) Por lo tanto se requiere de una liacutenea base de
largo plazo para saber coacutemo estas perturbaciones impactan la disponibilidad de N
en las uacuteltimas deacutecadassiglos en los ecosistemas lacustres y por lo tanto el alcance
real que los LUCC han tenido en el ciclo del N
Los ecosistemas mediterraacuteneos y el ciclo del N
Los ecosistemas mediterraacuteneos son especialmente sensibles a los LUCC
pues estos se encuentran sometidos a un estreacutes hiacutedrico durante largas temporadas
estivales y las precipitaciones se concentran en eventos puntuales y a veces con
altos montos pluviomeacutetricos Las precipitaciones tienen un alto potencial de arrastre
de sedimentos y nutrientes los que actuacutean como fertilizantes al llegar a los
ecosistemas lacustres A su vez un incremento en la disponibilidad de N puede
generar un desbalance con otros nutrientes (eg Fosforo) con efectos sobre la
productividad y diversidad de los lagos (Pentildeuelas et al 2012 Sardans et al 2012
McLauchlan 2013) En este sentido en Chile central se observa una disminucioacuten
de las precipitaciones especialmente fuerte en las uacuteltimas deacutecadas por lo que se ha
denoacuteminado como Megasequiacutea (Garreaud et al 2017) y el efecto de las
precipitaciones esporaacutedicas pueden generar un arrastre de nutrientes que no ha
sido evaluado
Los ecosistemas mediterraacuteneos concentran el 20 de la biodiversidad global
(Myers et al 2000) pero existe una escasez de conocimiento respecto a los
efectos del incremento de N en los cuerpos de agua como consecuencia de las
actividades humanas en el pasado histoacuterico (Ochoa-Hueso et al 2011) y coacutemo la
disponibilidad de N ha variado en el tiempo Los LUCC han aumentado el flujo de
N en las cuencas hidrograacuteficas viacutea escurrimiento superficial y lixiviacioacuten
16
favoreciendo la peacuterdida de sedimentos y nutrientes del suelo en el mundo entero
(Elser et al 2011 McLauchlan et al 2013 Xu et al 2017 y 2019) Ello ha
contribuido a la acidificacioacuten y eutrofizacioacuten generalizada de riacuteos y lagos
(McLauchlan et al 2013 Schindler et al 2008)
El ecosistema mediterraacuteneo de Chile central es una regioacuten ampliamente
intervenida desde al menos la colonizacioacuten espantildeola (1600-1800 CE) Los LUCC
han tenido efectos negativos en la disponibilidad de agua especialmente
observados con el desarrollo de las actividades minera (Prieto et al 2019) Aunque
se sabe de los impactos en los ecosistemas de bosque en teacuterminos de cobertura
debidos al desarrollo silvo-agropecuarias (Armesto et al 2010) se desconoce el
impacto de estas actividades sobre el ciclo del N en los cuerpos de agua o en queacute
momento cobran fuerza suficiente para alterar el flujo de N hacia los lagos de Chile
Central Por lo tanto en esta tesis nos centramos en entender coacutemo los LUCC han
afectado la disponibilidad de N en los lagos costeros de Laguna Matanzas y Lago
Vichuqueacuten desde la llegada de los espantildeoles a Chile hasta el presente
Los lagos como sensores ambientales
Los sedimentos lacustres son buenos sensores de cambios en los aportes
de nutrientes contaminacioacuten y eutroficacioacuten de los cuerpos de agua ya que son
capaces de preservar sentildeales quiacutemicas y fiacutesicas al momento de su formacioacuten y
ademaacutes con alta resolucioacuten y a largo plazo (Leng et al 2006) Por lo tanto
constituyen una herramienta uacutetil para reconstruir el flujo de N entre ecosistemas
terrestres y acuaacuteticos en el pasado sus consecuencias (eg incremento de la
productividad lacustre) y el rol del clima y los LUCC en el pasado (McLauchlan et
al 2013 von Gunten et al 2009) Por ejemplo el cultivo de maiacutez por parte de los
17
nativos iroqueses en Canadaacute provocoacute la eutrofizacioacuten del Lago Crawford (43degN)
durante los siglos XII y XV Entre las consecuencias registradas se documentoacute un
claro incremento de la productividad primaria y cambios en la estructura comunitaria
de diatomeas foacutesiles (incremento de Stephanodiscus complex y disminucioacuten de
Cyclotella bodanica en Ekdahl et al 2007) Otro ejemplo demuestra como las
actividades agriacutecolas en la cuenca del riacuteo Mississippi modificaron los patrones de
sedimentacioacuten y aporte de nutrientes al lago Pepin EE UU (44degN) a partir del
asentamiento europeo en 1840 CE y principalmente desde 1940 CE (Engstrom et
al 2009) Para Chile von Gunten et al (2009) a partir de indicadores
limnogeoloacutegicos evidencioacute la contaminacioacuten y eutroficacioacuten de cinco lagos cercanos
a la ciudad de Santiago (33ordmS) como consecuencia de la deposicioacuten atmosfeacuterica
de metales pesados (especialmente Cu) quema de combustible foacutesil y flujo de
nutrientes durante los uacuteltimos 200 antildeos
Caracteriacutesticas limnoloacutegicas de los lagos
Los procesos limnoloacutegicos afectan la distribucioacuten y abundancia de los
organismos en los lagos Estaacuten influenciados por forzamientos externos por
ejemplo la precipitacioacuten o la penetracioacuten de la luz y la morfometriacutea del lago En este
sentido el nivel del lago dependeraacute del balance entre las entradas y salidas de agua
(precipitaciones escurrimiento superficial y subterraacuteneo y la evaporacioacuten) y la forma
de la cuenca (profundidad pendiente aacuterea del espejo de agua)
En funcioacuten de la profundidad de penetracioacuten de la luz es posible diferenciar
dos aacutereas que determinan la distribucioacuten de los organismos la zona foacutetica (donde
penetra la luz y permite la fotosiacutentesis) y la zona afoacutetica (subyacente a la zona
foacutetica) Al depender de la radiacioacuten la zona foacutetica variacutea estacionalmente Ademaacutes
18
puede variar por el grado de turbidez la morfometriacutea del lago y la produccioacuten de
materia orgaacutenica en la columna de agua
Otro factor que influye en la productividad es el reacutegimen de mezcla de la
columna de agua pues favorece la oxigenacioacuten y el reciclado de nutrientes La
mezcla ocurre en condiciones de gran inestabilidad usualmente relacionado con el
reacutegimen de viento Por el contrario un lago estratificado resulta de grandes
diferencias en temperatura o salinidad entre la superficie (epilimnion) y el fondo del
lago (hipolimnion) que separa las masas de agua superficial y de fondo por una
termoclina (si la estratificacioacuten estaacute relacionada con diferencias de temperatura de
las masas de agua) o haloclina (diferencias de salinidad) En funcioacuten del reacutegimen
de mezcla los lagos se pueden clasificar en (Lewis 1983)
1 Amiacutecticos no hay mezcla vertical de la columna de agua
2 Monomiacutecticos friacuteos y caacutelidos se mezcla una vez al antildeo
3 Dimiacutecticos la columna de agua se mezcla dos veces al antildeo
4 Oligomiacutecticos Lagos estratificados la mayor parte del tiempo pero se mezclan a
intervalos irregulares mayores a 1 antildeo
5 Polimiacutecticos friacuteos y caacutelidos la columna de agua se mezcla varias veces al antildeo
El ciclo del N en lagos
Al estar presente en aacutecidos nucleicos aminoaacutecidos y proteiacutenas el N es un
nutriente esencial de los organismos Por lo tanto su disponibilidad en la columna
de agua es clave para la produccioacuten composicioacuten y acumulacioacuten de la MO a traveacutes
del tiempo (Olson 1998 Talbot 2002) Aunque el principal reservorio de N estaacute en
19
la atmosfera (N2) solo unos pocos organismos (eg cianobacterias) pueden fijarlo
directamente Asiacute el Nitroacutegeno Inorgaacutenico Disuelto (DIN) representa la principal
fuente de N disponible para la produccioacuten primaria en los ecosistemas acuaacuteticos
(Talbot et al 2002) El DIN estaacute compuesto por nitrato (NO3-) nitrito (N02
-) y amonio
(NH4+) y los cambios en su disponibilidad condicionan el tipo de produccioacuten primaria
(eg fijadores vs asimiladores de N ver Scott y Grant 2013 Scott 2008)
La Figura 1 resume los principales componentes en lagos del ciclo del N y
sus interacciones La principal entrada natural de N es viacutea fijacioacuten de N atmosfeacuterico
y es llevado a cabo principalmente por bacterias y algas verde-azules capaces de
romper el triple enlace entre los dos aacutetomos de N a un alto costo energeacutetico (Torres
et al 2012) El N fijado es reducido a NH3 Tras la muerte de los organismos el N
es liberado de sus tejidos por hongos y bacterias implicados en la descomposicioacuten
de la MO Una parte se deposita en el fondo del lago y la otra queda disponible para
ser asimilada por el fitoplancton como amonio mediante el proceso de
amonificacioacuten (Talbot 2002) La amonificacioacuten resulta de la reduccioacuten bacteriana
del amoniaco y se da preferentemente en condiciones anoacutexicas La volatizacioacuten del
amoniaco es un proceso por medio este se incorpora a la atmoacutesfera Este proceso
se da preferentemente en lagos aacutecidos (pH gt 85) El nitrato es otra forma de N
bioloacutegicamente disponible para el fitoplancton En los lagos el NO3 del DIN estaacute
compuesto por los aportes desde la cuenca hidrograacutefica en la que se circunscriben
por la resuspensioacuten desde el fondo del lago favorecido por procesos de mezcla
(descrito en seccioacuten anterior) y por los nitratos de la columna de agua Estos uacuteltimos
son el resultado de la nitrificacioacuten proceso realizado por bacterias aeroacutebicas
mediante el cual el amonio se oxida para formar nitrito y nitrato El ciclo se completa
20
con la desnitrificacioacuten que es la reduccioacuten bacteriana de nitrato al N atmosfeacuterico
Este proceso se da preferentemente en condiciones anoacutexicas
Figura 1 Materia Orgaacutenica sedimentaria y su relacioacuten con el ciclo del N y las
variaciones de δ15N (elaborado a partir de Talbot et al 2002) En la figura se
representan los factores clave en la acumulacioacuten de la MO sedimentaria y su
relacioacuten con el ciclo del N El δ15N de la MO es resultado de los aportes de MO
desde la cuenca MO de macroacutefitas y de la productividad bioloacutegica La productividad
en ecosistemas lacustres estaacute condicionada por DIN y la fijacioacuten de N atmosfeacuterico
El δ15N del pool del DIN disponible se va enriqueciendo por la asimilacioacuten
preferencial de 14N por parte del fitoplancton Ademaacutes el DIN remanente se va
enriqueciendo por accioacuten microbiana (amonificacioacuten nitrificacioacuten desnitrificacioacuten)
Reconstruyendo el ciclo del N a partir de variaciones en δ15N
La sentildeal isotoacutepica de N (δ15N) en los sedimentos lacustres puede ser usada
para reconstruir los cambios pasados del ciclo N la transferencia de N entre
ecosistemas terrestres y acuaacuteticos y el estado troacutefico del lago (Bruland y Mackenzie
2015 McLauchlan et al 2013 Woodward et al 2012 von Gunten et al 2009
Leng et al 2006 Meyers y Teranes 2001) La Figura 1 resume los principales
procesos y fuentes que afectan la sentildeal isotoacutepica δ15N (total o ldquobulkrdquo) en la MO de
21
los sedimentos lacustres Entre los que destacan las fuentes de MO (aloacutectona vs
autoacutectona) la bioproductividad lacustre y las entradas de N (ie fijacioacuten bioloacutegica
de N2) (Torres et al 2012) Estos procesos conducen a un fraccionamiento
isotoacutepico cineacutetico que favorece la disociacioacuten entre sus isotopos ldquopesadosrdquo (15N) y
ldquoligerosrdquo (14N) Asiacute por ejemplo los valores de δ15N de la MO se enriquecen en 15N
en la medida que los sedimentos lacustres estaacuten sometidos a perdidas de 14N viacutea
desnitrificacioacuten (Bruland y Mackenzie 2015 Diaz et al 2016 Talbot et al 2002)
Por otra parte el fitoplancton en la medida que aumenta su biomasa (eg
durante la estacioacuten caacutelida) puede llegar a asimilar el DIN hasta agotarse En este
caso el N remanente se va empobreciendo en 14N si no hay reposicioacuten de N (eg
aporte de NO3 de la cuenca) y la MO queda enriquecida en 15N Esta disminucioacuten
induce a una limitacioacuten de fitoplancton por N y los organismos diztroacuteficos pueden
verse estimulados por lo que se incrementa la entrada de N viacutea fijacioacuten de N2 (Scott
y Grantsz 2013) como resultado la MO oscila en valores maacutes bajos (en torno a 0permil)
La cantidad de MO que se deposita en el fondo del lago depende del
predominio de contribuciones provenientes desde la cuenca (aloacutectona) versus las
producidas dentro del mismo lago (autoacutectona) (Torres et al 2012) Aunque en
general los lagos reciben permanentemente aportes de sedimentos y MO desde
su cuenca los lagos pequentildeos tienden a reflejar maacutes los procesos que ocurren
solamente en su cuenca directa y reflejan los valores isotoacutepicos de la misma (Gu et
al 2006) Woodward et al (2012) realizaron un estudio en 50 lagos en Irlanda que
les permitioacute correlacionar diferentes patrones de LUCC (bosques de coniacuteferas
agricultura ganaderiacutea entre otros) con los valores de δ15N (y δ13C) en los
sedimentos superficiales de los lagos Encontraron que los valores de δ15N son maacutes
negativos en cuencas poco impactadas y maacutes positivos en aquellas con alto
22
impacto humano (eg usos de suelo agriacutecola y ganadero) McLauchlan et al (2007)
encontraron valores maacutes positivos de δ15N en los sedimentos del Mirror Lake New
Hampshire (29ordm N) a partir de la desforestacioacuten de su cuenca en 1790 CE (inicio
del asentamiento europeo en el lago) para el desarrollo de la actividad agriacutecola
Estos valores se volvieron maacutes negativos hacia valores similares al pre-
asentamiento europeo despueacutes del cese de la agricultura en la cuenca y la
recuperacioacuten del bosque a partir de 1929 CE
El anaacutelisis de δ15N en los sedimentos lacustres es una herramienta casi sin
explorar en Chile central Proponemos reconstruir la dinaacutemica de transferencia de
N cuenca-lago como consecuencia de los LUCC desde la colonizacioacuten espantildeola en
los lagos costeros de Laguna Matanzas (34ordmS) y Lago Vichuqueacuten (36ordmS) Como son
muacuteltiples los procesos que pueden afectar la sentildeal isotoacutepica de N (medido como
δ15N) en los sedimentos lacustres existen muchos problemas para su
interpretacioacuten paleoambiental (Leng et al 2006) Por ello en esta tesis utilizamos
un enfoque multiproxy que consiste en integrar el anaacutelisis limnoloacutegico y geoquiacutemico
de los sedimentos lacustres con los valores isotoacutepicos modernos de la columna de
agua (mediante monitoreo del POM) la relacioacuten vegetacioacuten-suelo de la cuenca y la
reconstruccioacuten de los principales usos de suelo de las cuencas desde 1975 CE
mediante el uso de imaacutegenes satelitales Considerando estas muacuteltiples liacuteneas de
evidencia nos planteamos utilizar las variaciones del δ15N para reconstruir los
cambios en la disponibilidad de N en los lagos costeros de Chile central y establecer
coacutemo y desde cuaacutendo las actividades humanas en la cuenca son lo suficientemente
importantes como para modificar el ciclo del N en los lagos desde la colonizacioacuten
espantildeola (siglo XVII)
23
Como hipoacutetesis planteamos que la dinaacutemica de transferencia de sedimentos
y nutrientes entre la cuenca y el lago tiene controles geoloacutegicos climaacuteticos y
bioloacutegicos que interactuacutean en el tiempo registraacutendose en la acumulacioacuten de MO de
los sedimentos lacustres (Fig1) Bajo este marco los LUCC modifican esta
dinaacutemica alterando la transferencia de N a los lagos ya que las actividades agriacutecolas
y ganaderas tienden a aportar maacutes N (aumentando δ15N) que la actividad forestal
(la que disminuye δ15N)
En el primer capiacutetulo titulado ldquoA combined approach to establishing the timing
and magnitude of anthropogenic nutrient alteration in a mediterranean coastal lake-
watershed systemrdquo se evaluoacute el rol de los LUCC en el control de la descarga de N
y sedimentos a Laguna Matanzas en un sistema cuenca-lago desde el siglo XVII
Para ello se realizoacute un anaacutelisis de la dinaacutemica sedimentaria lacustre mediante el
anaacutelisis de muacuteltiples proxys sedimentoloacutegicos (ie minerales y textura)
geoquiacutemicos (eg geoquiacutemica orgaacutenica e isotoacutepica) y bioloacutegicos (ie diatomeas) de
Laguna Matanzas que cubren los uacuteltimos 200 antildeos Ademaacutes se realizoacute una
reconstruccioacuten de los usos de suelo entre 1975 y 2016 a partir de imaacutegenes de
sateacutelites y se colectaron muestras de suelo de las principales coberturas de la
cuenca a los cuales se midioacute el δ15N
Entre los principales resultados obtenidos se destaca la influencia de la
ganaderiacutea y la denitrificacioacuten lo que explica los altos valores de δ15N evidenciados
por el registro lacustre durante la colonizacioacuten espantildeola e inicios de la repuacuteblica A
partir de la Gran Aceleracioacuten (1950 CE) la desforestacioacuten y el reemplazo de la
ganaderiacutea por plantaciones forestales tienen un correlato en el registro
sedimentario con valores de δ15N maacutes bajos Estos resultados demuestran que los
LUCC son el factor de primer orden para explicar los cambios observados en
24
nuestro registro de δ15N y a su vez implica un rol secundario del clima como posible
control de la disponibilidad de N en Laguna Matanzas Este capiacutetulo fue sometido
a la revista Scientific Reports y estaacute en revisioacuten desde el 26 de marzo de 2019 En
la elaboracioacuten del mauscrito han colaborado Claudio Latorre Blas Valero-Garceacutes
Matiacuteas Frugone Pablo Sarricolea Santiago Giralt Manuel Contreras-Lopez
Ricardo Prego y Patricia Bernardez
El segundo capiacutetulo se titula ldquoStable isotopes track land use and cover
changes in a mediterranean lake in central Chile over the last 600 yearsrdquo Aquiacute
evaluamos la dinaacutemica del N actual en el sistema cuenca-lago considerando los
valores isotoacutepicos modernos tanto de la vegetacioacuten como del suelo ademaacutes de los
cambios estacionales del POM de la columna de agua Esta informacioacuten se utiliza
como un anaacutelogo moderno para conocer el rol de los LUCC en la disponibilidad de
N en los uacuteltimos 600 antildeos Para ello se colectaron muestras filtradas de la columna
de agua a 2 5 y 20 m dos veces por antildeo entre noviembre de 2015 y julio de 2018
y se obtuvo el δsup1⁵N del POM Ademaacutes se colectoacute muestras de vegetacioacuten y suelo
de la cuenca diferenciando entre especies nativas plantaciones forestales y
vegetacioacuten herbaacutecea Esta informacioacuten se analizoacute en conjunto con la reconstruccioacuten
de los LUCC entre 1975 y 2014 y el anaacutelisis limnoloacutegico del lago Ello nos permitioacute
evaluar coacutemo el δ15N de los sedimentos lacustres refleja los valores δ15N de la
cuenca en el Lago Vichuqueacuten y el control que ejerce el clima en la sentildeal isotoacutepica
de POM Este capiacutetulo seraacute sometido a la revista Science of total environmet
proximamente En la elaboracioacuten del mauscrito han colaborado Claudio Latorre
Blas Valero-Garceacutes Matiacuteas Frugone Pablo Sarricolea Leonardo Villacis y M Laura
Carrevedo
25
Entre los principales resultados encontramos que el δ15N en los sedimentos
lacustres es maacutes positivo con presencia de agricultura en la cuenca y maacutes negativo
cuando esta es reemplazada por cobertura arboacuterea bien sea plantaciones
forestales bien sea bosque nativo A su vez durante la estacioacuten seca y caacutelida la
mayor entrada al lago es viacutea fijacioacuten de N atmosfeacuterico (bajos valores de δ15NPOM)
Durante la estacioacuten huacutemeda en cambio prevalecen los aportes de la cuenca con
altos valores de δ15NPOM Ello estariacutea evidenciando un cambio estacional en la
composicioacuten de especies en verano ligado a especies fijadoras de N y en invierno
las algas y microorganismos que consumen el DIN de la columna de agua
Referencias
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea AM Marquet PA 2010 From the Holocene to the Anthropocene A historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the
next carbon Earth rsquo s Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409 httpsdoiorg102134jeq20090005
Diacuteaz FP Frugone M Gutieacuterrez RA Latorre C 2016 Nitrogen cycling in an
extreme hyperarid environment inferred from δ15 N analyses of plants soils and herbivore diet Sci Rep 6 1ndash11 httpsdoiorg101038srep22226
Ekdahl EJ Teranes JL Wittkop CA Stoermer EF Reavie ED Smol JP
2007 Diatom assemblage response to Iroquoian and Euro-Canadian eutrophication of Crawford Lake Ontario Canada J Paleolimnol 37 233ndash246 httpsdoiorg101007s10933-006-9016-7
Elbert W Weber B Burrows S Steinkamp J Buumldel B Andreae MO
Poumlschl U 2012 Contribution of cryptogamic covers to the global cycles of carbon and nitrogen Nat Geosci 5 459ndash462
26
httpsdoiorg101038ngeo1486 Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506
httpsdoiorg101126science1215567 ARTICLE Engstrom DR Almendinger JE Wolin JA 2009 Historical changes in
sediment and phosphorus loading to the upper Mississippi River Mass-balance reconstructions from the sediments of Lake Pepin J Paleolimnol 41 563ndash588 httpsdoiorg101007s10933-008-9292-5
Erisman J W Sutton M A Galloway J Klimont Z amp Winiwarter W (2008)
How a century of ammonia synthesis changed the world Nature Geoscience 1(10) 636
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892
httpsdoiorg101126science1136674 Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J Christie
D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592 Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
IPCC 2013 Climate Change 2013 The Physical Science BasisContribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change Cambridge University Press Cambridge United Kingdom and New York NY USA
httpsdoiorg101017CBO9781107415324 Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007 Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32iumliquestfrac12S 3050iumliquestfrac12miumliquestfrac12asl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470
27
httpsdoiorg101073pnas0701779104 McLauchlan KK Gerhart LM Battles JJ Craine JM Elmore AJ Higuera
PE Mack MC McNeil BE Nelson DM Pederson N Perakis SS 2017 Centennial-scale reductions in nitrogen availability in temperate forests of the United States Sci Rep 7 1ndash7 httpsdoiorg101038s41598-017-08170-z
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Myers N Mittermeler RA Mittermeler CG Da Fonseca GAB Kent J
2000 Biodiversity hotspots for conservation priorities Nature httpsdoiorg10103835002501
Pentildeuelas J Poulter B Sardans J Ciais P Van Der Velde M Bopp L
Boucher O Godderis Y Hinsinger P Llusia J Nardin E Vicca S Obersteiner M Janssens IA 2013 Human-induced nitrogen-phosphorus imbalances alter natural and managed ecosystems across the globe Nat Commun 4 httpsdoiorg101038ncomms3934
Prieto M Salazar D Valenzuela MJ 2019 The dispossession of the San
Pedro de Inacaliri river Political Ecology extractivism and archaeology Extr Ind Soc httpsdoiorg101016jexis201902004
Robertson GP Vitousek P 2010 Nitrogen in Agriculture Balancing the Cost of
an Essential Resource Ssrn 34 97ndash125 httpsdoiorg101146annurevenviron032108105046
Sardans J Rivas-Ubach A Pentildeuelas J 2012 The CNP stoichiometry of
organisms and ecosystems in a changing world A review and perspectives Perspect Plant Ecol Evol Syst 14 33ndash47 httpsdoiorg101016jppees201108002
Schindler DW Howarth RW Vitousek PM Tilman DG Schlesinger WH
Matson PA Likens GE Aber JD 2007 Human Alteration of the Global Nitrogen Cycle Sources and Consequences Ecol Appl 7 737ndash750 httpsdoiorg1018901051-0761(1997)007[0737haotgn]20co2
Scott E and EM Grantzs 2013 N2 fixation exceeds internal nitrogen loading as
a phytoplankton nutrient source in perpetually nitrogen-limited reservoirs Freshwater Science 2013 32(3)849ndash861 10189912-1901
28
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Talbot M R (2002) Nitrogen isotopes in palaeolimnology In Tracking
environmental change using lake sediments (pp 401-439) Springer Dordrecht
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable
isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K
Yeager KM 2016 Different responses of sedimentary δ15N to climatic changes and anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
29
CAPIacuteTULO 1 A COMBINED APPROACH TO ESTABLISHING THE TIMING
AND MAGNITUDE OF ANTHROPOGENIC NUTRIENT ALTERATION IN A
MEDITERRANEAN COASTAL LAKE- WATERSHED SYSTEM
30
A combined approach to establishing the timing and magnitude of anthropogenic
nutrient alteration in a mediterranean coastal lake- watershed system
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugoneabc Pablo
Sarricolead Santiago Giralte Manuel Contreras-Lopezf Ricardo Prego g Patricia
Bernaacuterdez g Blas Valero-Garceacutesch
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Institute of Earth Sciences Jaume Almera (ICTJA-CSIC) CLluis Solegrave Sabaris sn Barcelona E-
08028 Spain
f Facultad de Ingenieriacutea y Centro de Estudios Avanzados Universidad de Playa Ancha Traslavintildea
450 Vintildea del Mar Chile
g Instituto de Investigaciones Marinas (CSIC) CEduardo Cabello 6 36208 Vigo Spain
h Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding author
E-mail address
clatorrebiopuccl magdalenafuentealbagmailcom
Abstract
Since the industrial revolution and especially during the Great Acceleration (1950
CE) human activities have profoundly altered the global nutrient cycle through land
use and cover changes (LUCC) However the timing and intensity of recent N
variability together with the extent of its impact in terrestrial and aquatic ecosystems
and coupled effects of regional LUCC and climate are not well understood Here
we used a multiproxy approach (sedimentological geochemical and isotopic
31
analyses historical records climate data and satellite images) to evaluate the role
of LUCC as the main control for N cycling in a coastal watershed system of central
Chile during the last few centuries The largest changes in N dynamics occurred in
the mid-1970s associated with the replacement of native forests and grasslands for
livestock grazing by plantations of introduced Monterey pine (Pinus radiata) and
eucalyptus (Eucalyptus globulus) Lake productivity increased and is evidenced by
an increase trend in δ15N values Our study shows that anthropogenic land
usecover changes are key in controlling nutrient supply and N availability in
Mediterranean watershed ndash lake systems and that large-scale forestry
developments during the mid-1970s likely caused the largest changes in central
Chile
Keywords Anthropocene Organic geochemistry watershedndashlake system Stable
Isotope Analyses Land usecover change Nitrogen cycle Mediterranean
ecosystems central Chile
1 INTRODUCTION
Human activities have become the most important driver of the nutrient cycles in
terrestrial and aquatic ecosystems since the industrial revolution (Gruber and
Galloway 2008 Galloway et al 2008 Holtgrieve et al 2011 Fowler et al 2013
Goyette et al 2016) Among these N is a common nutrient that limits productivity
in terrestrial and aquatic ecosystems (Vitousek and Howarth 1991 McLauchlan et
al 2013) With the advent of the Haber-Bosch industrial N fixation process in the
early 20th century total N fluxes have surpassed previous planetary boundaries
32
(Galloway et al 2008 Elser 2011) reaching unprecedented values (ie tipping
points) in the Earth system especially during what is now termed the Great
Acceleration (which began in the 1950s) which intensified in the 1970s (Howarth
2004 Steffen et al 2015) Yet dramatic changes in LUCC have occurred in the last
few centuries mainly linked to forestry and agro-pastoral activities (Poraj-Goacuterska et
al 2017 Ge et al 2019 Cheng et al 2019) These have had large impacts on N
(and Carbon -C-) availability in terrestrial and aquatic ecosystems but the synergic
effect with climate change and global N dynamics has not been established
(Vitousek et al 1997 Mclauchlan et al 2007 2013 Pongratz et al 2010
Woodward et al 2012 Mclauchlan et al 2017)
The onset of the Anthropocene poses significant challenges in mediterranean
regions that have a strong seasonality of hydrological regimes and an annual water
deficit (Stocker et al 2013) Mediterranean climates occur in all continents
(California central Chile Australia South Africa circum-Mediterranean regions)
providing a unique opportunity to investigate global change processes during the
Anthropocene in similar climate settings but with variable geographic and cultural
contexts The effects of global change in mediterranean watersheds have been
analyzed from different perspectives hydrology (Giorgi 2006 Arnell and Gosling
2013 Prudhomme et al 2014) vegetation dynamics (Lenihan et al 2003 Vicente-
Serrano et al 2013 Matesanz and Valladares 2014) sediment dynamics (Garciacutea-
Ruiz et al 2013 Garciacutea-Ruiz 2010 Syvitski et al 2005) changes in
biogeochemical cycles (Thomas et al 2013 Fowler et al 2013 Frank et al 2015)
carbon storage (Muntildeoz-Rojas et al 2015) and biodiversity (Hooper et al 2012) A
recent review showed an extraordinarily high variability of erosion rates in
mediterranean watersheds positive relationships with slope and annual
33
precipitation and the paramount effect of land use (Garciacutea-Ruiz et al 2015)
However the temporal context and effect of LUCC on nutrient supply to
mediterranean lakes has not been analyzed in much detail
Major LUCC in central Chile occurred during the Spanish Colonial period
(17th-18th centuries) (Armesto 1994 Gurnell et al 2001 Figueroa et al 2004
Martiacuten-Foreacutes et al 2012 Contreras-Loacutepez et al 2014) with the onset of
industrialization and mostly during the mid to late 20th century (von Gunten et al
2009 Gayo et al 2012) Recent copper pollution caused by 20th century mining
and industrial smelters has been documented in cores throughout the Andes
(Laguna el Ocho and Laguna Ensuentildeo von Gunten et al 2009) and also from our
surveys in the central valley (Batuco wetland) coastal range (Cordillera de Name)
and along the coast itself (Bucalemito and Colejuda) (Valero-Garceacutes et al 2010
unpublished data)
Paleolimnological studies have shown how these systems respond to
climate LUCC and anthropogenic impacts during the last millennia (Jenny et al
2002 2003 Villa Martiacutenez et al 2002 Frugone-Aacutelvarez et al 2017 Contreras et
al 2018) Furthermore changes in sediment and nutrient cycles have also been
identified in associated terrestrial ecosystems dating as far back as the Spanish
Conquest and related to fire clearance and wood extraction practices of the native
forests (Armesto 1994 Contreras-Loacutepez et al 2014) Nevertheless pollen and
limnological evidence argue for a more recent timing of the largest anthropogenic
impacts in central Chile For example paleo records show that during the mid-20th
century increased soil erosion followed replacement of native forest by Pinus
radiata and Eucalyptus globulus plantations at Laguna Matanzas Aculeo and
34
Vichuqueacuten lakes (Jenny et al 2002 Villa-Martiacutenez et al 2002 2003 Frugone-
Aacutelvarez et al 2017)
Lakes are a central component of the global carbon cycle Lakes act as a
sink of the carbon cycle both by mineralizing terrestrially derived organic matter and
by storing substantial amounts of organic carbon (OC) in their sediments (Anderson
et al 2009) Paleolimnological studies have shown a large increase in OC burial
rates during the last century (Heathcote et al 2015) however the rates and
controls on OC burial by lakes remain uncertain as do the possible effects of future
global change and the coupled effect with the N cycle LUCC intensification of
agriculture and associated nutrient loading together with atmospheric N-deposition
are expected to enhance OC sequestration by lakes Climate change has been
mainly responsible for the increased algal productivity since the end of the 19th
century and during the late 20th century in lakes from both the northern (Ruumlhland et
al2015) and southern hemispheres (Michelutti et al 2015 Carrevedo et al 2015)
but many studies suggest a complex interaction of global warming and
anthropogenic influences and it remains to be proven if climate is indeed the only
factor controlling these transitions (Catalaacuten et al 2013) Alternative causes for
recent N (Galloway et al 2008) increases in high altitude lakes such as catchment
mediated processes cannot be ruled out (Camarero and Catalan 2012 Fritz and
Anderson 2013) Few lake-watershed systems have robust enough chronologies of
recent changes to compare variations in C and N with regional and local processes
and even fewer of these are from the southern hemisphere (McLauchlan et al
2007 Holtgrieve et al 2011)
In this paper we present a multiproxy lake-watershed study including N and
C stable isotope analyses on a series of short cores from Laguna Matanzas in
35
central Chile focused in the last 200 years We complemented our record with land
use surveys satellite and aerial photograph studies Our major objectives are 1) to
reconstruct the dynamics among climate human activities and changes in the N
cycle over the last two centuries 2) to evaluate how human activities have altered
the N cycle during the Great Acceleration (since the mid-20th century)
2 STUDY SITE
Laguna Matanzas (33ordm45rsquoS 71ordm 40acuteW 7 m asl Fig 1) is a coastal lake located
in central Chile near to a large populated area (Santiago gt6106 inhabitants) The
lake has a surface area of 15 km2 with a max depth of 3 m and a watershed of 30
km2 The lake basin is emplaced over Pleistocene-Holocene aeolian and alluvial fan
deposits (SERNAGEOMIN 2003) A recent phase of dune activity occurred from the
mid to late Holocene which mostly sealed off the basin from the ocean (Villa-
Martinez 2002) Climate in Laguna Matanzas is characterized by cool-wet winters
and hot-dry summers with annual precipitation of ~510 mm and a mean annual
temperature of 12ordmC Central Chile is in the transition zone between the southern
hemisphere mid-latitude westerlies belt and the South Pacific Anticyclone (or SPA)
(Garreaud et al 2009) In winter precipitation is modulated by the north-west
displacement of the SPA the northward shift of the westerlies wind belt and an
increased frequency of storm fronts stemming off the Southern Hemisphere
Westerly Winds (SWW) (Frugone-Aacutelvarez et al 2017) Austral summers are
typically dry and warm as a strong SPA blocks the northward migration of storm
tracks stemming off the SWW
36
Historic land cover changes started after the Spanish conquest with a Jesuit
settlement in 1627 CE near El Convento village and the development of a livestock
ranch that included the Matanzas watershed After the Jesuits were expelled from
South America in 1778 CE the farm was bought by Pedro Balmaceda and had more
than 40000 head of cattle around 1800 CE (Contreras-Lopez et al 2014) The first
Pinus radiata and Eucalyptus globulus trees were planted during the second half of
the 19th century and were mostly used for dune stabilization (Albert 1900 Gibson
1972) However the main plantation phase occurred 60 years ago (Villa-Martinez
2002) as a response to the application of Chilean Forestry Laws promulgated in
1931 and 1974 and associated state subsidies
Major land cover changes occurred recently from 1975 to 2008 as shrublands
were replaced by more intensive land uses practices such as farmland and tree
plantations (Schulz et al 2010) Laguna Matanzas is part of the Reserva Nacional
Humedal El Yali protected area (Ramsar Ndeg 878) Despite this protected status the
lake and its watershed have been heavily affected by intense agricultural and
farming activities during the last decades The main inlet ldquoLas Rosasrdquo has been
diverted for crop irrigation causing a significant loss of water input to the lake
Consequently the flooded area of the lake has greatly decreased in the last couple
of decades (Fig 1b) Exotic tree species cover a large surface area of the
watershed Recently other activities such as farms for intensive chicken production
have been emplaced in the watershed
37
Figure 1 Laguna Matanzas a) Digital Elevation Model showing the watershed and
the surface hydrologic connection with Colejuda and Cabildo lakes b) Climograph
depicting the warm dry season in austral summer c) Annual precipitation from
1965-2015 (arrow shows start of the most recent ldquomega-droughtrdquo- see Garreaud et
al 2017) d) A decline of lake area occurs between 2007 and 2018 Lake surface
area decreased first along the western sector (in 2007) followed by more inland
areas (in 2018)
38
3 RESULTS
31 Age Model
The age model for the Matanzas sequence was developed using Bacon software
to establish the deposition rates and age uncertainty (Blaauw and Christen 2011)
It is based on 210Pb and 137Cs dates and two 14C dates (Table 2) According to this
age model the lake sequence spans the last 1000 years (Fig 2) A major breccia
layer (unit 3b) was deposited during the early 18th century which agrees with
historic documents indicating that a tsunami impacted Laguna Matanzas and its
watershed in 1730 CE (Contreras-Loacutepez et al 2014) Here we focus in the last 200
years were the most important changes occurred in terms of LUCC (after the
sedimentary hiatus caused by the tsunami) The Spanish colonial period (17th -18th
century) brought new forms of territorial management along with an intensification
of watershed use which remained relatively unchanged until the 1900s
39
Figure 2 a) Age-depth model obtained for the Laguna de Matanzas sedimentary
sequence A hiatus occurs at 80 cm depth (unit 3b) The section of core used for our
analysis is highlighted in a red rectangle b) Close up of the age model used for
analysis of recent anthropogenic influences on the N cycle c) Information regarding
the 14C dates used to construct age model
Lab code Sample ID
Depth (cm) Material Fraction of modern C
Radiocarbon age
Pmc Error BP Error
D-AMS 021579
MAT11-6A 104-105 Bulk Sediment
8843 041 988 37
D-AMS 001132
MAT11-6A 1345-1355
Bulk Sediment
8482 024 1268 21
POZ-57285
MAT13-12 DIC Water column 10454 035 Modern
Table 2 Laguna Matanzas radiocarbon dates
32 The sediment sequence
Laguna Matanzas sediments consist of massive to banded mud with some silt
intercalations They are composed of silicate minerals (plagioclase quartz and clay
minerals) with relatively high TOC content (Fig 3) Pyrite is a common mineral
indicating dominant anoxic conditions in the lake sediments whereas aragonite
occurs only in the uppermost section Mineralogical analyses visual descriptions
texture and geochemical composition were used to characterize five main facies
(Fig 3) F1 (organic-rich mud) represents baseline sedimentation in a shallow well-
mixed brackish highly productive lake and F1acute (dark orange) is a less organic facies
than F1 (more details see table in the supplementary material) F2 (massive to
banded silty mud) indicates periods of higher clastic input into the lake but finer
(mostly clay minerals) likely from suspension deposition associated with flooding
40
events Aragonite (up to 15 ) occurs in both facies but only in samples from the
uppermost 30 cm and it is interpreted as endogenic linked to higher MgFe waters
and elevated biologic productivity
Figure 3 Sedimentary facies and units mineralogy grain size elemental geochemical
and isotopic composition of Laguna Matanzas core Values highlighted in gray indicate
that these are above average
The banded to laminated fining upward silty clay layers (F3) reflect
deposition by high energy turbidity currents The presence of aragonite suggests
that littoral sediments were incorporated by these currents Non-graded laminated
coarse silt layers (F4) do not have aragonite indicating a dominant watershed
41
sediment source Both facies are interpreted as more energetic flood deposits but
with different sediment sources A unique breccia layer with coarse silt matrix and
cm-long soft clasts (F5) is interpreted as a high-energy event (ie a tsunami)
capable of eroding the littoral zone and depositing coarse clastic material in the
distal zone of the lake Similar coarse breccia layers have been found at several
coastal sites in Chile and interpreted as tsunami-related deposits (Vargas et al
2005 Le Roux et al 2008)
33 Sedimentary units
Three main units and six subunits have been defined (Fig 3) based on
sedimentary facies and sediment composition We use ZrTi as an indicator of the
mineral fraction transported from the watershed (Marzecoacuteva et al 2011) as higher
ZrTi (F3 and F4) is commonly associated with coarser sediments (Cuven et al
2010) A high correlation among Br BrTi and TOC (r = 046-087 p value=0)
supports the use of BrTi as an indicator of lake productivity (Marzecoacuteva et al 2011
Frugone-Aacutelvarez et al 2017) The MnFe ratio is indicative of lake bottom
oxygenation (Naecher et al 2013) as under reducing conditions Mn mobilizes more
than Fe leading to a decreased MnFe ratio (Marzecoacuteva et al 2011) SrTi indicates
periods of increased aragonite formation as Sr is preferentially included in the
aragonite mineral structure (Veizer et al 1971) (See supplementary material)
The basal unit (3c 129 to 99 cm) is relatively organic-rich (TOC mean=26
BioSi mean=5) and composed by F1 (without aragonite) with some coarser F4
flood layers (ZrTi mean=025) Unit 3b (98 to 80 cm) is interpreted as a tsunami or
storm surge deposit (breccia F5 grading into massive to banded silt F2) Unit 3a
(79 to 73 cm) is characterized by relatively low productivity (TOC=2 BrTi=002
42
BioSi=4) under anoxic conditions (MnFe=001) Unit 2a (72 to 31cm) has
relatively less organic content and more intercalated clastic facies F3 and F4 The
top of this unit (43-30 cm) has elevated TS values The Subunit 1b (30-20 cm)
shows increasing TOC BioSi and BrTi values (TOC mean = 29 BioSi mean =
54 BrTi mean=004) and the upper subunit 1a (19-0 cm) has the highest TOC
(mean = 64) and BioSi (mean = 56) high BrTi mean = 010 and the presence
of aragonite More frequent anoxic conditions (MnFe lower than 001) during units
3 and 2 shifted towards more oxic episodes (MnFe = 003) in unit 1 (Fig 3)
34 Isotopic signatures
Figure 4 shows the isotopic signature from soil samples of the major land
usescover present in the Laguna Matanzas used as an end member in comparison
with the lacustrine sedimentary units δ15N from cropland samples exhibit the
highest values whereas grassland and soil samples from lake shore areas have
intermediate values (Fig 4) Tree plantations and native forests have similarly low
δ15N values (+11 permil SD=24) All samples (except those from the lake shore)
exhibit low δ13C values (from ndash285permil to ndash298permil) CNmolar from agriculture land
lakeshore area and non-vegetation areas samples display the lowest values (about
18) CNmolar from tree plantations and native forest have the highest values (383
and 267 respectively)
43
Figure 4 C-N stable isotope plot showing a comparison of lake sediments grouped
by sedimentary units (MAT11-6A) with the soil end members of present-day (lake
shore and land usecover) from Laguna Matanzas
The δ15N values from sediment samples (MAT11-6A) range from ndash15 and
+53permil (mean= +35 SD=05) δ13C values range from ndash266permil to ndash202permil (mean=
ndash240 SD=14) In unit 3c δ15N and δ13C show relatively high values (mean=
+41permil and ndash233permil respectively) δ15N and δ13C from unit 3b and 3a fluctuate at
slightly lower values than in 3c (mean δ15N from 3a= +38permil and 3b mean= +39permil
mean δ13C from 3a= ndash242permil and 3b mean= ndash244permil) In unit 2a δ15N values are
relatively high (mean= +38permil) but show a slightly decreasing trend (from +52 to
+24permil at 66 cm and 36 cm respectively) δ13C also decreases (mean= ndash242permil)
reaching minimum values at 45 cm (ndash266permil) with a sharp increase towards the top
of this unit with maximum values ca ndash210permil Unit 1 exhibits the lowest δ15N values
(1a mean= +11permil 1b mean= +28permil) with negative values in the uppermost
44
sediments (ndash04permil at 14 cm) δ13C values show a decreasing trend over most of
subunit 1b and increase only near the very top of this unit
35 Recent land use changes in the Laguna Matanzas watershed
Major LUCC from 1975 (unit 1b) to 2016 CE in the Laguna Matanzasacutes
watershed is summarized in Figure 5 The watershed has a surface area of 30 km2
of which native forest (36) and grassland areas (44) represented 80 of the
total surface in 1975 The area occupied by agriculture was only 02 and tree
plantations were absent Isolated burned areas (33) were located mostly in the
northern part of the watershed By 1989 tree plantations surface area had increased
to 5 burned areas to 17 and agricultural fields to 9 (a 45-fold increase) and
native forest and grassland sectors decreased to 23 and 27 respectively By
2016 agricultural land and tree plantations have increased to 17 of the total area
whereas native forests decreased to 21
45
Figure 5 Land uses and cover changes from 1975 to 2016 in Laguna Matanzas
watershed from natural cover and areas for livestock grazing (grassland) to the
expansion of agriculture and forest plantation
4 DISCUSSION
41 N and C dynamics in Laguna Matanzas
Small lakes with relatively large watersheds such as Laguna Matanzas would
be expected to have relatively high contributions of allochthonous C to the sediment
OM (Gu et al 2006) Terrestrial C3 plants (δ13C =ndash26 to ndash28permil Meyers and Terranes
2001 Ku et al 2007) are dominant in the Laguna Matanzas watershed Likewise
our soil samples ranged across similar although slightly more negative values
46
(δ13C = ndash30 to ndash28permil Fig 4) to those proposed by Meyers and Terranes (2001)
and are used here as terrestrial end members oil samples were taken from the lake
shore (δ13C = ndash22permil) and POM from surface water (δ13C = ndash24permil) were more
positive than the terrestrial end member and are used as lacustrine end members
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from terrestrial vegetation and more positive δ13C values have increased
aquatic OM (Gu et al 2006 Torres et al 2012) Phytoplankton preferentially uptake
12C leaving the DIC pool enriched in 13C (Gu et al 2006) especially when there are
no important external sources of C (eg decreased C input from the watershed)
Therefore during events of elevated primary productivity the phytoplankton uptakes
12C until its depletion and are then obligated to use the heavier isotope resulting in
an increase in δ13C Changes in lake productivity thus greatly affect the C isotope
signal (Torres et al 2012) with high productivity leading to elevated δ13C values
(Torres et al 2012 Gu et al 2006)
In a similar fashion the N isotope signatures in Laguna Matanzas reflect a
combination of factors including different N sources (autochthonousallochthonous)
and lake processes such as productivity isotope fractionation in the water column
and sediment denitrification Elevated δ15N values from a POM sample (+22permil) and
average values from the lake shore (mean=+34permil SD=028) are used as aquatic
end members whereas terrestrial samples have values from +10 +24 (tree species)
to +77permil +35 (agriculture) which we use as terrestrial end members (Fig 4)
Autochthonous OM in aquatic ecosystems typically displays low δ15N values
when the OM comes from N-fixing species Atmospheric fixation of N2 by
cyanobacteria ends in OM with δ15N values close to 0permil (Leng et al 2006)
Phytoplankton preferentially uptake 14N from Dissolved Inorganic Nitrogen (DIN) in
47
the water column and derived OM typically have δ15N values lower than DIN values
When productivity increases the remaining DIN becomes depleted in 14N which in
turn increases the δ15N values of phytoplankton over time especially if the N not
replenished (Torres et al 2012) Thus high POM δ15N values from Laguna
Matanzas reflect elevated phytoplankton productivity with a 14N- depleted DIN In
addition N-watershed inputs also contribute to high δ15N values Heavily impacted
watersheds by human activities are often reflected in isotope values due to land use
changes and associated modified N fluxes For example the input of N runoff
derived from the use of inorganic fertilizers leads to the presence of elevated δ15N
(between ndash4 to +4permil) values in the water bodies (Elliott and Brush 2006 Diebel and
Vander Zanden 2009) Widory et al (2004) reported a direct relationship between
elevated δ15N values and increased nitrate concentration from manure in the
groundwater of Brittany (France) Terranes and Bernasconi (2000) found a good
correlation between augmented nutrient loading and a progressive increase in δ15N
values of sedimentary OM related to agricultural land use
Post-depositional diagenetic processes can further affect C and N isotope
signatures Several studies have shown a decrease in δ13C values of OM in anoxic
environments particularly during the first years of burial related to the selective
preservation of OM depleted in 13C (Hollander and Smith 2001 Lehmann et al
2002 Gӓlman et al 2009 Torres et al 2012) Diagenetic processes can also lead
to post-burial N isotope enrichment of the sediments Indeed 14N is consumed more
rapidly than 15N by denitrifying bacteria which intensifies under anoxic conditions
(Diebel and Vander Zanden 2009) Thus the remaining OM pool will be enriched
in 15N leaving to elevated δ15N values (Granger et al 2008 Menzel et al 2013)
48
In summary the relatively high δ15N values in sediments of Laguna Matanzas
reflect N input from an agriculturegrassland watershed with positive synergetic
effects from increased lake productivity enrichment of DIN in the water column and
most likely denitrification The increase of algal productivity associated with
increased N terrestrial input andor recycling of lake nutrients (and lesser extent
fixing atmospheric N) and denitrification under anoxic conditions can all increase
δ15N values (Fig 3) In addition elevated lake productivity without C replenishing
(eg by terrestrial C input) produces shifts towards positive δ13C values whereas C
input from the watershed generates more negative δ13C values
42 Recent evolution of the Laguna Matanzas watershed
Sedimentological compositional and geochemical indicators show three
depositional phases in the lake evolution under the human influence in the Laguna
Matanzas over the last two hundred years Although the record is longer (around
1000 years) we analyzed the last two centuries (unit 2 and 1) to provide a recent
historical context for the large changes detected during the 20th century
The first phase lasted from the beginning of the 19th century until ca 1940
(unit 2a Fig 3 4) and was characterized by moderate productivity with elevated
sediment input from the watershed as indicated by our geochemical proxies (BrTi
= 004 AlTi gt 01 Fig 6) Higher lake levels and dominant anoxic bottom conditions
(MnFe = 001 Fig3) can be related to increased precipitation (mean= 557 mmyr)
and lower temperatures (summer annual temperature lt19ordmC) During the Spanish
colonial period the Laguna Matanzas watershed was used as a livestock farm
(Jesuits 1627-1780 unit 3 followed by the Pedro Balmaceda era 1780-1810- unit
2a) (Contreras-Loacutepez et al 2014) The main buildings and farm facilities at the El
49
Convento village During this period livestock grazing and lumber extraction for
mining would have involved extensive deforestation and loss of native vegetation
(eg Armesto et al 1994 2010) However the Matanzas pollen record does not
show any significant regional deforestation during this period (Villa Martiacutenez 2002)
suggesting that the impact may have been highly localized
Lake productivity sediment input and elevated precipitation (Fig 6) all
suggest that N availability was related to this increased input from the watershed
The N from cow manure and soil particles would have led to higher δ15N values
(Elliot and Brush 2016) In addition anoxic lake bottom conditions would lead to
even further enrichment of buried sediment N The δ13C values lend further support
to our interpretation of increased sediment input -and N- from the watershed
Decreased δ13C values reach their lowest values in the entire record (ndash270permil) at
ca 1910 CE (Fig 4 6)
During most of the 19th century human activities in Laguna Matanzas were
similar to those during the Spanish Colonial period However the appearance of
Pinus radiata and Eucalyptus globulus pollen ca 1890 CE (Gibson 1972) the dune
stabilization-afforestation program which began ca 1900 CE (Albert 1900) and the
application of the Chilean forestry law of 1931 (DFL nordm265) contributed to an
increased capacity of the surrounding vegetation to retain nutrients and sediments
The law subsidized forest plantations in areas devoid of vegetation and prohibited
the cutting of forest on slopes greater than 45ordm These land use changes were coeval
with decreased sediment inputs (AlTi trend) from the watershed slightly increased
lake productivity (BrTi from 001 to 003) and a decrease in annual precipitation
(Fig 6) N isotope values become more negative during this period although they
remained high (from +49permil to +37permil) whereas the δ13C trend towards more
50
positive values reflects changes in the N source from watershed to in-lake dynamics
(e g increased endogenic productivity)
The second phase started after 1940 and is clearly marked by an abrupt
change in the general trend of δ15N that oscillates around +41permil (SD = 04permil) during
the first phase to +24permil (SD = 14permil) during the second These δ15N trends reflect
the lowest watershed nutrient and sediment inputs (based on the AlTi record)
decreased precipitation (mean = 318 mm year) and a slight increase in lake
productivity (increased BrTI) Depositional dynamics in the lake likely crossed a
threshold as human activity intensified throughout the watershed and lake levels
decreased
During the Great Acceleration δ15N values shifted towards higher values to
ca 3permil with an increase in δ13C values that are not reflected either in lake
productivity or lake level As the sediment input from the watershed increased and
precipitation remained as low as the previous decade δ15N values during this period
are likely related to watershed clearance which would have increased both nutrient
and sediment input into the lake
The δ13C trend to more positive values reaching the peaks in the 1960s (ndash
212permil) at the same time as the peaks in temperature (196 ordmC mean annual) with a
downward trend in precipitation A shift in OM origin from macrophytes and
watershed input influences to increased lake productivity could explain this trend
(Fig 4 1b)
In the 1970s the Laguna Matanzasacute watershed was mostly covered by native
forest (36) and grassland areas were intended for livestock grazing (44 Fig 5)
Both soils have δ15NSoil lt 5permil (Fig 4) Farming fields occupied a very small area and
tree plantations were almost nonexistent The decreasing trend in δ15N values seen
51
in our record is interrupted by several large peaks that occurred between ca 1975
and ca 1989 when the native forest and grassland areas fell by 23 and 27
respectively largely due to fires affecting 17 of the forests Agriculture fields
increased by 4 and forest plantations increased by 9 (unit 1a) Concomitantly
sediment ndash and likely N - inputs from the watershed decreased (as indicated by the
trend in AlTi) the precipitation was still relatively low (Fig 6) These changes are
likely related to the increase of vegetation cover especially of tree plantations (which
have more negative δ15N values) The small increase in productivity in the lake could
have been favored by increased temperature (von Gunten et al 2009) After 1989
the increase in agriculture land (17 in 2016) correlates with increasing δ15N δ13C
and TOC trends in spite of declining rainfall The increase of forest plantations was
mostly in response to the implementation of the Law Decree of Forestry
Development (DL 701 of 1974) that subsidized forest plantation After 1989 the
increase in agricultural land (17 in 2016) is synchronous with increasing δ15N
δ13C TOC and MnFe trends despite the decline in rainfall and overall lower lake
levels as more water is used for irrigation
The third phase started c 1990 CE (unit 1a) when OM accumulation rates
increase and δ13C δ15N decreased reaching their lowest values in the sequence
around 2000 CE Afterward during the 21st century δ13C and δ15N values again
began to increase The onset of unit 1 is marked by increased lake productivity and
decreased sediment input (AlTi lt 02) synchronous with intensive farming replacing
forestry and extensive agriculture (Fig 5 6)
A change in the general trend of δ15N values which decreased until 1990
(unit 1a) as the native forest and grassland area fell to 23 and 27 respectively
is most likely due to deforestation and fires Agriculture surface increased to 4 and
52
forest plantations increased by 9 (Fig 5) Concomitantly sediment ndash and likely N
ndash inputs from the watershed decreased probably related to the low precipitation (Fig
1b) and the increase of vegetation cover in the watershed in particularly by tree
plantations (with more negative δ15N Fig 4)
At present agriculture and tree plantations occupy around 34 of the
watershed surface whereas native forests and grassland cover 21 and 25
respectively Lake productivity as indicated by BrTi (Fig6) is higher and generates
OM with lower δ15N and higher δ13C values (about +2permil and ndash20permil ca 2000 CE
respectively) due to in-lake processes (ie biological N fixation and nutrient
recycling) and driven by changes in the arboreal cover which diminishes nutrient
flux into the lake (Fig6)
53
Figure 6 Anthropogenic and climatic forcing and lake dynamics response
(productivity sediment input N and C cycles) at Matanzas Lake over the last two
54
centuries Mean annual precipitation reconstructed and temperatures (von Gunten
et al 2009) Vertical gray bars indicate mega-droughts
5 CONCLUSIONS
Human activities have been the main factor controlling the N and C cycle in
the Laguna Matanzas during the last two centuries The N isotope signature in the
lake sediments reflects changes in the watershed fluxes to the lake but also in-lake
processes such as productivity and post-depositional changes Denitrification could
have been a dominant process during periods of increased anoxic conditions which
were more frequent prior to Great Acceleration (1950 CE) Higher δ15N and lower
δ13C values are associated with increased nutrient input from the watershed due to
increased livestock grazing and agriculture pressure (phase 1 Fig 7a) whereas
lower isotope values occurred during periods of increased forest plantations (phase
3 Fig 7c) During periods of increased lake productivity - such as in the last few
decades - δ15N values increased significantly
The most important change in C and N dynamics in the lake occurred after the
1950s (Phase 2 and 3) as tree plantations dominated the watershed Recent
changes in N dynamics can be explained by the higher nutrient contribution
associated with intensive agriculture (i e fertilizers) since the 1990s Although the
replacement of livestock activities with forestry and farming seems to have reduced
nutrient and soil export from the watershed to the lake the inefficient use of fertilizer
(by agriculture) can be the ultimate responsible for lake productivity increase during
the last decades
55
Figure 7 Schematic diagrams illustrating the main factors controlling the
isotope N signal in sediment OM of Laguna Matanzas N input from watershed
depends on human activities and land cover type Agriculture practices and cattle
(grassland development) contribute more N to the lake than native forest and
plantations Periods of higher productivity tend to deplete the dissolved inorganic N
in 14N resulting in higher δ15N (OM) The denitrification processes are more effective
in anoxic conditions associated with higher lake levels
6 METHODS
Short sediment cores were recovered from Laguna Matanzas using an Uwitec
gravity piston corer in 2011 and 2013 (Fig 1a) Sediment cores (MAT11-6A 129 cm
MAT13-2A 149 cm MAT13-3A 33 cm MAT13-4A 99 cm) were split
photographed sub-sampled and stored at the Pyrenean Institute of Ecology (IPE-
CSIC Spain) Core MAT11-6A was obtained from the central sector of the lake and
56
was selected for detailed multiproxy analyses (including elemental geochemistry C
and N isotope analyses XRF and 14C dating)
The isotope analyses (δ13C and δ15N) were performed at the Laboratory of
Biogeochemistry and Applied Stable Isotopes (LABASI-PUC Chile) using a Delta
V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via a
Conflo IV interface Isotope results are expressed in standard delta notation (δ) in
per mil (permil) relative to the standards Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Sediment samples
for δ13Corg were pre-treated with 50 ml of HCl and 50 ml of deionized water and
dried at 60ordmC for 4 hr to remove carbonates (Harris et al 2011)
Total Carbon (TC) Total Organic Carbon (TOC) Total Inorganic Carbon (TIC)
and Total Sulphur (TS) were measured every cm with a LECO SC 144 DR at IPE-
CSIC XRF measurements were carried out every 4 mm in MAT11-1A-1G core using
an AVAATECH X-ray Fluorescence II core scanner at the University of Barcelona
(Spain) Results are expressed as element intensities in counts per second (cps)
Tube voltage was operated at 30 kV and 10 kV to obtain the abundances of 15
elements (Al Si S Cl K Ca Ti V Mn Fe Br Rb Sr Y Zr) with an average at
least of 1600 cps (less for Br=1000)
Biogenic silica content mineralogy and grain size were measured every 4
cm Biogenic silica was measured following Mortlock and Froelich (1989) and
Bernaacuterdez et al (2005) using an Auto Analyzer Technicon AAII for dissolved silicate
analysis Mineralogy was analyzed with a Siemens D-500 X-ray diffractometer (Cu
kα 40 kV 30 mA graphite monochromator) at the ICTJA-CSIC (Spain) Grain size
analyses were performed in a Beckmann Coulter LS 13 320 Particle Size Analyzer
57
at the IPE-CSIC The samples were classified according to textural classes as
follows clay (lt2 μm) silt (20-2 microm) and sand (gt2 microm) fractions
The age-depth model for the Laguna Matanzas sedimentary sequence was
constructed using 210Pb and 137Cs dating techniques (MAT13-4A) as well as two 14C
AMS dates on bulk sediment samples (MAT11-6A fig 2) We dated the dissolved
inorganic carbon (DIC) in the water column and no significant reservoir effect is
present in the modern-day water column (10454 + 035 pcmc Table 2) An age-
depth model was obtained with the Bacon R package to estimate the deposition
rates and associated age uncertainties along the core (Blaauw and Christen 2011)
To estimate LUCC of Laguna Matanzasacutes watershed we use satellite images
Landsat MSS for 1975 Landsat TM for 1989 and Landsat OLI for 2016 all taken in
summer or autumn (Table 1) We performed supervised classification of land uses
(maximum likelihood algorithm) for each year (1975 1989 and 2016) and the results
were mapped using software ArcGIS 102 in 2017
Satellite Images Acquisition Date
Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat OLI 20160404 30 m
Table 1 Landsat imagery
Surface water samples were filtered for obtained particulate organic matter In
addition soil samples from the main land usecover present in the Laguna Matanzas
watershed were collected Elemental C N and their corresponding isotopes from
POM and soil were obtained at the LABASI and used here as end members
Daily precipitation at Santo Domingo (33ordm39`S 71ordm3 6`W)- the nearest weather
station to Laguna Matanzasndash was compiled using the redPrec R package (Fig1 d
Serrano-Notivoli et al 2017 Sarricolea et al 2019) To extend our precipitation
58
reconstruction back to 1824 we correlated this dataset with that available for
Santiago The Santiago data was compiled from data published in the Anales of
Universidad of Chile (1850) for the years 1824 to 1849 Almeyda (1940) for the years
1849 to 1864 and the Quinta Normal series from 1866 to the present (Direccioacuten
Meteoroloacutegica de Chile) We generated a linear regression model between the
presentday Santo Domingo station and the compiled Santiago data with a Pearson
coefficient of 087 and p-valuelt 001
Acknowledgments This research was funded by grants CONICYT AFB170008
to the Institute of Ecology and Biodiversity (IEB) and FONDECYT 1150763 (to CL)
Doctoral grant Becas Chile 21150224 MEDLANT (Spanish Ministry of Economy
and Competitiveness grant CGL2016-76215-R) Additional funding was provided
by the Laboratorio Internacional de Cambio Global (LINCGlobal PUC-CSIC) We
thank R Lopez E Royo and M Gallegos for help with sample analyses We thank
the Laboratory of Biogeochemistry and Applied Stable Isotopes (LABASI) of the
Department of Ecology (PUC) for sample analyses
References
Albert F 1900 Las dunas o sean arenas volantes voladeros arenas muertas invasion de las arenas playas I meacutedanos del centro de Chile comprendiendo el litoral desde el liacutemite norte de la provincia de Aconcagua hasta el liacutemite sur de la de Arauco Memorias cientiacuteficas I Lit
Anderson NJ W D Fritz SC 2009 Holocene carbon burial by lakes in SW
Greenland Glob Chang Biol httpsdoiorg101111j1365-2486200901942x
Armesto J Villagraacuten C Donoso C 1994 La historia del bosque templado
chileno Ambient y Desarro 66 66ndash72 httpsdoiorg101007BF00385244
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A
59
historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Arnell NW Gosling SN 2013 The impacts of climate change on river flow
regimes at the global scale J Hydrol 486 351ndash364 httpsdoiorg101016JJHYDROL201302010
Bernaacuterdez P Prego R Franceacutes G Gonzaacutelez-Aacutelvarez R 2005 Opal content in
the Riacutea de Vigo and Galician continental shelf Biogenic silica in the muddy fraction as an accurate paleoproductivity proxy Cont Shelf Res httpsdoiorg101016jcsr200412009
Blaauw M Christen JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474 httpsdoiorg10121411-BA618
Brush GS 2009 Historical land use nitrogen and coastal eutrophication A
paleoecological perspective Estuaries and Coasts 32 18ndash28 httpsdoiorg101007s12237-008-9106-z
Camarero L Catalan J 2012 Atmospheric phosphorus deposition may cause
lakes to revert from phosphorus limitation back to nitrogen limitation Nat Commun httpsdoiorg101038ncomms2125
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego
R Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and environmental change from a high Andean lake Laguna del Maule central Chile (36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Catalan J Pla-Rabeacutes S Wolfe AP Smol JP Ruumlhland KM Anderson NJ
Kopaacuteček J Stuchliacutek E Schmidt R Koinig KA Camarero L Flower RJ Heiri O Kamenik C Korhola A Leavitt PR Psenner R Renberg I 2013 Global change revealed by palaeolimnological records from remote lakes A review J Paleolimnol httpsdoiorg101007s10933-013-9681-2
Chen Y Y Huang W Wang W H Juang J Y Hong J S Kato T amp
Luyssaert S (2019) Reconstructing Taiwanrsquos land cover changes between 1904 and 2015 from historical maps and satellite images Scientific Reports 9(1) 3643 httpsdoiorg101038s41598-019-40063-1
Contreras S Werne J P Araneda A Urrutia R amp Conejero C A (2018)
Organic matter geochemical signatures (TOC TN CN ratio δ 13 C and δ 15 N) of surface sediment from lakes distributed along a climatological gradient on the western side of the southern Andes Science of The Total Environment 630 878-888 httpsdoiorg101016jscitotenv201802225
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia UC
60
de V 2014 ELEMENTOS DE LA HISTORIA NATURAL DEL An Mus Hist Natulas Vaplaraiso 27 51ndash67
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-010-9453-1
Diebel M Vander Zanden MJ 201209 Nitrogen stable isotopes in streams
stable isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19 1127ndash1134 httpsdoiorg10189008-03271
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506 Espizua LE Pitte P 2009 The Little Ice Age glacier advance in the Central
Andes (35degS) Argentina Palaeogeogr Palaeoclimatol Palaeoecol 281 345ndash350 httpsdoiorg101016JPALAEO200810032
Figueroa JA Castro S A Marquet P A Jaksic FM 2004 Exotic plant
invasions to the mediterranean region of Chile causes history and impacts Invasioacuten de plantas exoacuteticas en la regioacuten mediterraacutenea de Chile causas historia e impactos Rev Chil Hist Nat 465ndash483 httpsdoiorg104067S0716-078X2004000300006
Fowler D Coyle M Skiba U Sutton MA Cape JN Reis S Sheppard
LJ Jenkins A Grizzetti B Galloway N Vitousek P Leach A Bouwman AF Butterbach-bahl K Dentener F Stevenson D Amann M Voss M 2013 The global nitrogen cycle in the twenty- first century Philisophical Trans R Soc B httpsdoiorghttpdxdoiorg101098rstb20130164
Frank D Reichstein M Bahn M Thonicke K Frank D Mahecha MD
Smith P van der Velde M Vicca S Babst F Beer C Buchmann N Canadell JG Ciais P Cramer W Ibrom A Miglietta F Poulter B Rammig A Seneviratne SI Walz A Wattenbach M Zavala MA Zscheischler J 2015 Effects of climate extremes on the terrestrial carbon cycle Concepts processes and potential future impacts Glob Chang Biol httpsdoiorg101111gcb12916
Fritz SC Anderson NJ 2013 The relative influences of climate and catchment
processes on Holocene lake development in glaciated regions J Paleolimnol httpsdoiorg101007s10933-013-9684-z
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
61
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS) implications for past sea level and environmental variability J Quat Sci 32 830ndash844 httpsdoiorg101002jqs2936
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892 httpsdoiorg101126science1136674
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924 httpsdoiorg104319lo20095430917
Garciacutea-Ruiz JM 2010 The effects of land uses on soil erosion in Spain A
review Catena httpsdoiorg101016jcatena201001001
Garciacutea-Ruiz JM Loacutepez-Moreno JI Lasanta T Vicente-Serrano SM
Gonzaacutelez-Sampeacuteriz P Valero-Garceacutes BL Sanjuaacuten Y Begueriacutea S Nadal-Romero E Lana-Renault N Goacutemez-Villar A 2015 Los efectos geoecoloacutegicos del cambio global en el Pirineo Central espantildeol una revisioacuten a distintas escalas espaciales y temporales Pirineos httpsdoiorg103989Pirineos2015170005
Garciacutea-Ruiz JM Nadal-Romero E Lana-Renault N Begueriacutea S 2013
Erosion in Mediterranean landscapes Changes and future challenges Geomorphology 198 20ndash36 httpsdoiorg101016jgeomorph201305023
Garreaud RD Vuille M Compagnucci R Marengo J 2009 Present-day
South American climate Palaeogeogr Palaeoclimatol Palaeoecol httpsdoiorg101016jpalaeo200710032
Gayo EM Latorre C Jordan TE Nester PL Estay SA Ojeda KF
Santoro CM 2012 Late Quaternary hydrological and ecological changes in the hyperarid core of the northern Atacama Desert (~21degS) Earth-Science Rev httpsdoiorg101016jearscirev201204003
Ge Y Zhang K amp Yang X (2019) A 110-year pollen record of land use and land
cover changes in an anthropogenic watershed landscape eastern China Understanding past human-environment interactions Science of The Total Environment 650 2906-2918 httpsdoiorg101016jscitotenv201810058
Goyette J Bennett EM Howarth RW Maranger R 2016 Global
Biogeochemical Cycles 1000ndash1014 httpsdoiorg1010022016GB005384Received
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and
oxygen isotope fractionation during dissimilatory nitrate reduction by
62
denitrifying bacteria 53 2533ndash2545 httpsdoiorg10230740058342
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
Gurnell AM Petts GE Hannah DM Smith BPG Edwards PJ Kollmann
J Ward J V Tockner K 2001 Effects of historical land use on sediment yield from a lacustrine watershed in central Chile Earth Surf Process Landforms 26 63ndash76 httpsdoiorg1010021096-9837(200101)261lt63AID-ESP157gt30CO2-J
Heathcote A J et al Large increases in carbon burial in northern lakes during the
Anthropocene Nat Commun 610016 doi 101038ncomms10016 (2015) Hollander DJ Smith MA 2001 Microbially mediated carbon cycling as a
control on the δ13C of sedimentary carbon in eutrophic Lake Mendota (USA) New models for interpreting isotopic excursions in the sedimentary record Geochim Cosmochim Acta httpsdoiorg101016S0016-7037(00)00506-8
Holtgrieve GW Schindler DE Hobbs WO Leavitt PR Ward EJ Bunting
L Chen G Finney BP Gregory-Eaves I Holmgren S Lisac MJ Lisi PJ Nydick K Rogers LA Saros JE Selbie DT Shapley MD Walsh PB Wolfe AP 2011 A coherent signature of anthropogenic nitrogen deposition to remote watersheds of the Northern Hemisphere Science (80) 334 1545ndash1548 httpsdoiorg101126science1212267
Hooper DU Adair EC Cardinale BJ Byrnes JEK Hungate BA Matulich
KL Gonzalez A Duffy JE Gamfeldt L Connor MI 2012 A global synthesis reveals biodiversity loss as a major driver of ecosystem change Nature httpsdoiorg101038nature11118
Horvatinčić N Sironić A Barešić J Sondi I Krajcar Bronić I Borković D
2016 Mineralogical organic and isotopic composition as palaeoenvironmental records in the lake sediments of two lakes the Plitvice Lakes Croatia Quat Int httpsdoiorg101016jquaint201701022
Howarth RW 2004 Human acceleration of the nitrogen cycle Drivers
consequences and steps toward solutions Water Sci Technol httpsdoiorg1010382Fscientificamerican0490-56
Jenny B Valero-Garceacutes BL Urrutia R Kelts K Veit H Appleby PG Geyh
M 2002 Moisture changes and fluctuations of the Westerlies in Mediterranean Central Chile during the last 2000 years The Laguna Aculeo record (33deg50primeS) Quat Int 87 3ndash18 httpsdoiorg101016S1040-
63
6182(01)00058-1
Jenny B Wilhelm D Valero-Garceacutes BL 2003 The Southern Westerlies in
Central Chile Holocene precipitation estimates based on a water balance model for Laguna Aculeo (33deg50primeS) Clim Dyn 20 269ndash280 httpsdoiorg101007s00382-002-0267-3
Leng M J Lamb A L Heaton T H Marshall J D Wolfe B B Jones M D
amp Arrowsmith C 2006 Isotopes in lake sediments In Isotopes in palaeoenvironmental research (pp 147-184) Springer Dordrecht
Le Roux JP Nielsen SN Kemnitz H Henriquez Aacute 2008 A Pliocene mega-
tsunami deposit and associated features in the Ranquil Formation southern Chile Sediment Geol 203 164ndash180 httpsdoiorg101016jsedgeo200712002
Lehmann M Bernasconi S Barbieri a McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66 3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Lenihan JM Drapek R Bachelet D Neilson RP 2003 Climate change
effects on vegetation distribution carbon and fire in California Ecol Appl httpsdoiorg101890025295
McLauchlan K K Gerhart L M Battles J J Craine J M Elmore A J
Higuera P E amp Perakis S S (2017) Centennial-scale reductions in nitrogen availability in temperate forests of the United States Scientific reports 7(1) 7856 httpsdoi101038s41598-017-08170-z
Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32ordmS 3050 masl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
Martiacuten-Foreacutes I Casado MA Castro I Ovalle C Pozo A Del Acosta-Gallo
B Saacutenchez-Jardoacuten L Miguel JM De 2012 Flora of the mediterranean basin in the chilean ESPINALES Evidence of colonisation Pastos 42 137ndash160
Marzecovaacute A Mikomaumlgi a Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54 httpsdoiorg103176eco2011105
Matesanz S Valladares F 2014 Ecological and evolutionary responses of
Mediterranean plants to global change Environ Exp Bot httpsdoiorg101016jenvexpbot201309004
64
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Menzel P Gaye B Wiesner MG Prasad S Stebich M Das BK Anoop A
Riedel N Basavaiah N 2013 Influence of bottom water anoxia on nitrogen isotopic ratios and amino acid contributions of recent sediments from small eutrophic Lonar Lake central India Limnol Oceanogr 58 1061ndash1074 httpsdoiorg104319lo20135831061
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Michelutti N Wolfe AP Cooke CA Hobbs WO Vuille M Smol JP 2015
Climate change forces new ecological states in tropical Andean lakes PLoS One httpsdoiorg101371journalpone0115338
Muntildeoz-Rojas M De la Rosa D Zavala LM Jordaacuten A Anaya-Romero M
2011 Changes in land cover and vegetation carbon stocks in Andalusia Southern Spain (1956-2007) Sci Total Environ httpsdoiorg101016jscitotenv201104009
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Poraj-Goacuterska A I Żarczyński M J Ahrens A Enters D Weisbrodt D amp
Tylmann W (2017) Impact of historical land use changes on lacustrine sedimentation recorded in varved sediments of Lake Jaczno northeastern Poland Catena 153 182-193 httpdxdoiorg101016jcatena201702007
Pongratz J Reick CH Raddatz T Claussen M 2010 Biogeophysical versus
biogeochemical climate response to historical anthropogenic land cover change Geophys Res Lett 37 1ndash5 httpsdoiorg1010292010GL043010
Prudhomme C Giuntoli I Robinson EL Clark DB Arnell NW Dankers R
Fekete BM Franssen W Gerten D Gosling SN Hagemann S Hannah DM Kim H Masaki Y Satoh Y Stacke T Wada Y Wisser D 2014 Hydrological droughts in the 21st century hotspots and uncertainties from a global multimodel ensemble experiment Proc Natl Acad Sci httpsdoiorg101073pnas1222473110
65
Ruumlhland KM Paterson AM Smol JP 2015 Lake diatom responses to
warming reviewing the evidence J Paleolimnol httpsdoiorg101007s10933-
015-9837-3
Sarricolea P Meseguer-Ruiz Oacute Serrano-Notivoli R Soto M V amp Martin-Vide
J (2019) Trends of daily precipitation concentration in Central-Southern Chile Atmospheric research 215 85-98 httpsdoiorg101016jatmosres201809005
Sernageomin 2003 Mapa geologico de chile version digital Publ Geol Digit Serrano-Notivoli R de Luis M Begueriumlia S 2017 An R package for daily
precipitation climate series reconstruction Environ Model Softw httpsdoiorg101016jenvsoft201611005
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Stine S 1994 Extreme and persistent drought in California and Patagonia during
mediaeval time Nature 369 546ndash549 httpsdoiorg101038369546a0
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL
Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Syvitski JPM Voumlroumlsmarty CJ Kettner AJ Green P 2005 Impact of humans
on the flux of terrestrial sediment to the global coastal ocean Science 308(5720) 376-380 httpsdoiorg101126science1109454
Thomas RQ Zaehle S Templer PH Goodale CL 2013 Global patterns of
nitrogen limitation Confronting two global biogeochemical models with observations Glob Chang Biol httpsdoiorg101111gcb12281
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Urbina Carrasco MX Gorigoitiacutea Abbott N Cisternas Vega M 2016 Aportes a
la historia siacutesmica de Chile el caso del gran terremoto de 1730 Anuario de Estudios Americanos httpsdoiorg103989aeamer2016211
Vargas G Ortlieb L Chapron E Valdes J Marquardt C 2005 Paleoseismic
inferences from a high-resolution marine sedimentary record in northern Chile
66
(23degS) Tectonophysics httpsdoiorg101016jtecto200412031 399(1-4) 381-398httpsdoiorg101016jtecto200412031
Valero-Garceacutes BL Latorre C Morelliacuten M Corella P Maldonado A Frugone M Moreno A 2010 Recent Climate variability and human impact from lacustrine cores in central Chile 2nd International LOTRED-South America Symposium Reconstructing Climate Variations in South America and the Antarctic Peninsula over the last 2000 years Valdivia p 76 (httpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-yearshttpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-years
Vicente-Serrano SM Gouveia C Camarero JJ Begueria S Trigo R
Lopez-Moreno JI Azorin-Molina C Pasho E Lorenzo-Lacruz J Revuelto J Moran-Tejeda E Sanchez-Lorenzo A 2013 Response of vegetation to drought time-scales across global land biomes Proc Natl Acad Sci httpsdoiorg101073pnas1207068110
Villa Martiacutenez R P 2002 Historia del clima y la vegetacioacuten de Chile Central
durante el Holoceno Una reconstruccioacuten basada en el anaacutelisis de polen sedimentos microalgas y carboacuten PhD
Villa-Martiacutenez R Villagraacuten C Jenny B 2003 Pollen evidence for late -
Holocene climatic variability at laguna de Aculeo Central Chile (lat 34o S) The Holocene 3 361ndash367
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Human alteration of the global nitrogen cycle sources and consequences Ecological applications 7(3) 737-750 httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Rein B Urrutia R amp Appleby P (2009) A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 The Holocene 19(6) 873-881 httpsdoiorg1011770959683609336573
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Widory D Kloppmann W Chery L Bonnin J Rochdi H Guinamant JL
2004 Nitrate in groundwater An isotopic multi-tracer approach J Contam Hydrol 72 165ndash188 httpsdoiorg101016jjconhyd200310010
67
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
68
Supplementary material
Facie Name Description Depositional Environment
F1 Organic-rich
mud
Massive to banded black
organic - rich (TOC up to 14 )
mud with aragonite in dm - thick
layers Slightly banded intervals
contain less OM (TOClt4) and
aragonite than massive
intervals High MnFe (oxic
bottom conditions) High CaTi
BrTi and BioSi (up to 5)
Distal low energy environment
high productivity well oxygenated
and brackish waters and relative
low lake level
F2 Massive to
banded silty clay
to fine silt
cm-thick layers mostly
composed by silicates
(plagioclase quartz cristobalite
up to 65 TOC mean=23)
Some layers have relatively high
pyrite content (up to 25) No
carbonates CaTi BrTi and
BioSi (mean=48) are lower
than F1 higher ZrTi (coarser
grain size)
Deposition during periods of
higher sediment input from the
watershed
69
F3 Banded to
laminated light
brown silty clay
cm-thick layers mostly
composed of clay minerals
quartz and plagioclase (up to
42) low organic matter
(TOC mean=13) low pyrite
and BioSi content
(mean=46) and some
aragonite
Flooding events reworking
coastal deposits
F4 Laminated
coarse silts
Thin massive layers (lt2mm)
dominated by silicates Low
TOC (mean=214 ) BrTi
(mean=002) MnFe (lt02)
TIC (lt034) BioSi
(mean=46) and TS values
(lt064) and high ZrTi
Rapid flooding events
transporting material mostly
from within the watershed
F5 Breccia with
coarse silt
matrix
A 17 cm thick (80-97 cm
depth) layer composed by
irregular mm to cm-long ldquosoft-
clastsrdquo of silty sediment
fragments in a coarse silt
matrix Low CaTi BrTi and
MnFe ratios and BioSi
Rapid high energy flood
events
70
(mean=43) and high ZrTi
(gt018)
Table Sedimentological and compositional characteristics of Laguna Matanzas
facies
71
CAPIacuteTULO 2 STABLE ISOTOPES TRACK LAND USE AND COVER
CHANGES IN A MEDITERRANEAN LAKE IN CENTRAL CHILE OVER THE
LAST 600 YEARS
72
Stable isotopes track land use and cover changes in a mediterranean lake in
central Chile over the last 600 years
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugone-Aacutelvarezabc Pablo
Sarricolead Leonardo Villaciacutese M Laura Carrevedob Blas Valero-Garceacutescg
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Departamento de Ecologiacutea Universidad de ChileLas Palmeras 3425 Nuntildeoa Santiago Chile
f Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding authors E-mail addresses clatorrebiopuccl magdalenafuentealbagmailcom
Key words Anthropocene lake sediment δsup1⁵N δsup1sup3C Great Acceleration organic
geochemistry watershedndashlake system Stable Isotope Analyses land usecover
change Nitrogen cycle mediterranean ecosystems central Chile
73
Abstract
Nutrient fluxes in many aquatic ecosystems are currently being overridden by
anthropic controls especially since the industrial revolution (mid-1800s) and the
Great Acceleration (mid-1900s) Land use and cover changes (LUCC) alter the
availability and fluxes of nutrients such as nitrogen that are transferred via runoff
and groundwater into lakes By altering lake productivity and trophic status these
changes are often preserved in the sedimentary record Here we use
geolimnological proxies and stable isotope analyses (δsup1⁵N δsup13C) on lake sediments
to reconstruct the lake-watershed nutrient transfer in a central Chilean lake (Lago
Vichuqueacuten) over the past 600 years We compare our multiproxy record from recent
lake sediments to the soilvegetation relationship across the watershed as well as
land usecover changes from 1975 to 2014 derived from satellite images Our results
show that lake sediment δsup1⁵N values increased with meadow cover but decreased
with tree plantations suggesting increased nitrogen retention when trees dominate
the watershed Our δsup1⁵N record from central Chile can thus be interpreted as a proxy
for nutrient availability over the last 600 years mainly derived from land use changes
coupled with climate drivers Although variable sources of organic matter and in situ
fractionation often hinder straightforward environmental interpretations of stable N
isotope signatures our study shows that lake sediment δsup1⁵N can be a useful tool for
assessing the contribution of past human activities in nutrient and nitrogen cycling
1 Introduction
Nitrogen (N) is a limiting nutrient in terrestrial and aquatic ecosystems (Vitousek
et al 1997) Changes in its availability can drive eutrophication and increase
pollution in these ecosystems (McLauchlan et al 2013) Although recent human
74
impacts on the global N cycle have been significant the consequences of increased
anthropogenic N on these ecosystems are unclear (Gerhart and Mclauchlan 2014
Battye et al 2017) Isotopic signatures are key to understand N dynamics in lakes
nevertheless in situ andor diagenetic fractionation along with multiple sources of
organic matter (OM) often hinder straightforward environmental interpretations from
isotopes Monitoring δ15N and δ13C values as components of the N cycle
specifically those related to the link between terrestrial and aquatic ecosystems can
help differentiate between effects from processes versus sources in stable isotope
values (eg from Particulate Organic Matter -POM- soil and vegetation) and
improve how we interpret variations in δ15N (and δ13C) values at longer temporal
scales
The main processes controlling stable N isotopes in bulk lake OM are source
lake productivity diagenesis and denitrification (Talbot 2001 Leng et al 2006
Fariacuteas et al 2009 Torres et al 2012 Brahney et al 2014) N sources depend on
contributions from the watershed (ie soil and biomass) the transfer of atmospheric
N to the lake (ie atmospheric N2 fixation) and nutrient recycling (Leng et al 2006)
Atmospheric N2 (isotopically defined as δ15N =0) is fixed by cyanobacteria with
minimal fractionation resulting in δ15NOM values that fluctuate from -2 to +2permil (Fogel
and Cifuentes 1993 Talbot 2001 Elliot and Brush 2006) Besides N2 fixation by
cyanobacteria phytoplankton species assimilate dissolved inorganic nitrogen (DIN)
and values typically range from +7 to +10permil (Xu et al 2016 Meyers et al 2003) In
addition seasonal changes in POM occur in the lake water column Gu et al (2006)
sampled the water column of Lake Wauberg (Florida USA) during a hydrologic year
and found a higher development of N fixing species during the summer A major
factor behind this increase are human activities in the watershed which control the
75
inputs of the sediment and nutrients to the lake (Woodward et al 2011) Some
studies have shown higher δ15N values in lake sediments from watersheds that are
highly affected by human activities (Bruland and Mackenzie 2010 Woodward et al
2011) Syntenic fertilizers displayed δ15N values between -4 to +4permil cattle manure
around +8 to +22permil and surface and groundwater OM between 0 to +10permil (Elliott
and Brush 2006 Leng et al 2006) Although relatively low δ15N values from
fertilizers constitute major N input to human-altered watersheds the elevated loss
of 14N via volatilization of ammonia and denitrification leaves the remaining total N
input enriched in 15N (Bruland and Mackenzie 2010)
In addition to the different sources and variations in lake productivity early
diagenesis at the sedimentndashwater interface in the sediment can further alter
sediment OM isotope values (Brahney et al 2014 Galman et al 2009) During
diagenetic loss of N in lacustrine systems kinetic fractionation occurs and the
remaining N is enriched in 15N leading to higher δ15N values (Galman et al 2006
Talbot 2001) Denitrification tends to enrich lake sediments in 15N due to the
assimilation of nitrates by denitrifying bacteria (Granger et al 2008) and is more
prevalent in anoxic environments (Brahney et al 2016 Lehman et al 2002)
Carbon isotopes in lake sediments can also provide useful information about
paleoenvironmental changes OM origin and depositional processes (Meyers et al
2003) Allochthonous organic sources (high CN ratios) produce isotope values
similar to values from catchment vegetation Autochthonous organic matter (low CN
ratio) is influenced by fractionation both in the lake and the watershed leading up to
carbon assimilation by lacustrine biota (Woodward et al 2011) Changes in
productivity greatly affect δ13COM values (Torres et al 2012) as during C uptake
plankton preferentially assimilate 12C leaving the dissolved inorganic carbon (DIC)
76
pool enriched in 13C (Gu et al 2006) with phytoplankton averaging ca 20 permil lower
than the respective δ13CDIC values (Leng et al 2006) Under conditions of low to
moderate primary productivity plankton preferentially uptake the lighter 12C
resulting in low δ13C values displayed by the OM (Torres et al 2012) Conversely
during high primary productivity phytoplankton will uptake 12C until its depletion and
is then forced to assimilate the heavier isotope resulting in an increase in δ13C
values Higher productivity in C-limited lakes due to slow water-atmosphere
exchange of CO2 also results in high δ13C values (Galman et al 2009) In these
cases algae are forced to uptake dissolved bicarbonate with δ13C values between
7 to 10permil higher than those of the atmospheric CO2 dissolved in water (Sun et al
2016 Torres et al 2012 Galman et al 2009)
Stable isotope analyses from lake sediments are thus useful tools to
reconstruct shifts in lake-watershed dynamics caused by changes in limnological
parameters and LUCC Our knowledge of the current processes that can affect
stable isotope signals in a watershed-lake system is limited however as monitoring
studies are scarce Besides in order to use stable isotope signatures to reconstruct
past environmental changes we require a multiproxy approach to understand the
role of the different variables in controlling these values Hence in this study we
carried out a detailed survey of current N dynamics in a coastal central Chilean lake
(Lago Vichuqueacuten) and a multiproxy study of a sediment record spanning the last
600 years The characterization of the recent changes in the watershed since 1970s
is based on satellite images to compare recent changes in the lake and assess how
these are related with climate variability and an ever increasing human footprint
(Lara et al 2012 Gallardo et al 2015 Steffen et al 2015) Our main goal is to
investigate how stable isotope values from lake sediment reflect changes in the lake
77
ndash watershed system during periods of high watershed disruption (eg Spanish
Conquest late XIX century Great Acceleration) and recent climate change (eg
Little Ice Age and current global warming)
2 Study Site
Lago Vichuqueacuten (34ordm49`S 72ordm04W 4 m asl 30 m of depth Fig 1) is a
mesotrophic to eutrophic lake with dominant anoxic bottom conditions that is
stratified year around (Frugone-Aacutelvarez et al 2017) The main inlets are the
Vichuqueacuten and Huintildee rivers and the only outlet is the Llico estuary that drains into
the Pacific Ocean High tides can sporadically shift the flow direction of the Llico
estuary which increases the marine influence in the lake Dune accretion gradually
limited ocean-lake connectivity until the estuary was almost completely closed off
by ca 10 ky BP and the current lake regime formed (Frugone-Aacutelvarez et al 2017)
The area is characterized by a mediterranean climate with cold-wet winters and
hot-dry summers and an annual precipitation of ~650 mm and a mean annual
temperature of 15ordmC During the austral winter months (June - August) precipitation
is modulated by north-west shifts of the South Pacific Anticyclone (SPA) followed by
an increased frequency of storm fronts stemming off the South Westerly Winds
(SWW) A strengthened SPA during austral summers (December - March) which
are typically dry and warm blocks the northward migration of storm tracks stemming
off the SWW (Fig1 Frugone-Aacutelvarez et al 2017)
78
Figure 1 a) The location of the Lago Vichuqueacuten in central Chile b) Map of land
uses and cover in 2016 c) Ombrothermic diagram mediterranean climates are
characterized by cold-wet winters with surplus moisture from June to August and
hot-dry summers d) Lake bathymetry showing location of cores and water sampling
sites used in this study
Although major land cover changes in the area have occurred since 1975 to the
present as the native forests were replaced by tree (Monterey pine and eucalyptus)
plantations the region was settled before the Spanish conquest (Frugone-Alvarez
et al 2018) Historical documents indicate that Vichuqueacuten was occupied by a
Mitimaes colony as part of the Inka empire (Odone 1998) Unlike other Andean
areas during the Inka period the introduction of corn and quinoa in the Vichuqueacuten
watershed do not seem to have intensified land use The Spanish colonial period in
Chile lasted from 1542 CE to the independence in 1810 CE The first historical
document (1550 CE) shows that the areas around Vichuqueacuten were settled by the
Spanish early on as the Vichuqueacuten watershed was given as an ldquoencomiendardquo
system to Juan Cuevas The ldquoencomiendardquo system provided the owners with land
79
and indigenous people to work but also the introduction of wheat wine cattle
grazing and logging of native forests for lumber extraction and increasing land for
agricultural activities (Odone et al 1998 Armesto et al 2010) During the 19th
century (the Republic) the export of wheat to Australia and Canada generated
intensive changes in land cover use The town of Vichuqueacuten became the regional
capital and plans to build a maritime port in the bay of Vichuqueacuten were drawn
However the fall of international markets in 1880 paralyzed these plans During the
20th century the plantation of trees (Pinus radiata and Eucalyptus globulus) in areas
cleared of native forests was subsidized by two national laws DFL nordm265 (1931) and
DFL nordm 701 (1974) both of which provided funds for such plantations During the last
decades the urbanization with summer vacation homes along the shorelines of
Lago Vichuqueacuten has greatly increased Extensive lake eutrophication is currently a
large environmental problem (EULA 2008)
3 Methods
Coring campaigns were organized from 2011 to 2015 (Fig 1d) We recovered
12 short cores and 4 long cores (up to 14 m long) at several sites using a hammer-
modified UWITEC gravity corer Sediment cores (VIC11-2A 170 cm VIC13-2B 170
cm VIC13-2D 1200 cm and VIC15-1A 119 cm) were split photographed sub-
sampled and stored in the Pyrenean Institute of Ecology (IPE-CSIC Spain) Core
VIC13-2B was selected for detailed multiproxy analyses (including elemental
geochemistry C and N isotopes XRF and 14C dating) Stable isotope analyses
(δ13C δ15N) were performed at the Laboratory of Biogeochemistry and Applied
Stable Isotopes (LABASI-PUC Chile) Sediment samples for δ13Corg were pre-
treated with 50 ml of HCl and 50 ml of deionized water and dried at 60ordmC for 4 hr to
remove carbonates (Harris et al 2011) Isotope analyses were conducted using a
80
Delta V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via
a Conflo IV interface Isotope results are expressed in standard delta notation (δ)
and expressed in per mil (permil) relative to the Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Total carbon (TC)
Total Organic Carbon (TOC) Total Inorganic Carbon (TIC) and Total Sulfur (TS)
were measured every 2 cm with a LECO SC 144 DR at the IPE-CSIC
An AVAATECH X-Ray Fluorescence II core scanner with a Rh X-ray tube from
the University of Barcelona was used to obtain XRF logs every 4 mm of resolution
Results are expressed as element intensities in counts per second (cps) Tube
voltage was operated at 30 kV and 10 kV to obtain the abundances of 18 elements
(Pb Al Si S Cl K Ca Ti V Mn Fe Co Zn Br Rb Sr Y Zr) with an average of
at least of 1800 cps (less for Pb=437 and Zn=934 cps) Pb was included due to
similar behavior with Co and Fe Element ratios were calculated to describe changes
in redox conditions (MnFe) endogenic productivity (BrTi) carbonate formation
(SrCa and CaTi) and sediment input (ZrTi) (Barreiro-Lostres et al 2014
Carrevedo et al 2015 Frugone-Aacutelvarez et al 2017 Kylander et al 2011 Moreno
et al 2007a)
Several campaigns were carried out to sample the POM from the water column
two per hydrologic year from November 2015 to August 2018 A liter of water was
recovered in three sites through to the lake two are from the shallower areas (with
samples taken at 2 and 5 m depth at each site) and one in the deeper central portion
(at 2 5 and 20 m depth) of the lake (Fig 1d) The samples were filtered using glass
fiber filters (25 m) and then frozen to obtain the stable carbon and nitrogen isotope
signal of lacustrine POM Additionally soil and vegetation samples from the
following communities native species meadow hydrophytic vegetation and
81
Monterrey pine (Pinus radiata) were taken for stable isotope analyses (see in
supplementary material)
The age model for the complete Lago Vichuqueacuten sedimentary sequence is
based on Frugone-Aacutelvarez et al (2017) with the last 1000 years based on
210Pb137Cs dating techniques in VIC15-1A and 14C AMS dates on bulk sediment
samples (Supplementary Table S1) The 14C measurements of lake water DIC show
a moderate reservoir effect of 180 plusmn 25 14C yr) The revised age-depth model used
here includes three more 14C AMS dates performed with the program Bacon to
establish the deposition rates along the core (Blaauw and Christen 2011 Fig 2)
The age-depth model indicates that average resolution between 0 to 87 cm is lt2
cm per year and from 88 to 170 cm it is lt47 cm per year
82
Figure 2 Chronological age-depth model for the Lago Vichuqueacuten sedimentary
sequence spanning the last 1000 yr (see also Frugone-Alvarez et al 2017)
To estimate land use changes in the watershed we use Landsat MSS images
for 1975 and Landsat TM for 1989 and 2014 all taken in either summer or autumn
(Table 1) We performed supervised classification of land uses (maximum likelihood
83
algorithm) for each year (1975 1989 and 2014) and results were mapped using
ArcGIS 102
Table 1 Images using for LUCC reconstruction
Source of LUCC
Acquisition
Date Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat TM 19991226 30 m
CONAF 2009 30 m
Land cover Chile 2014 30 m
CONAF 2016 30 m
Previous Work on Lago Vichuqueacuten sedimentary sequence
The sediments are organic-poor dark brown to brown laminated silt with some
intercalated thin coarser clastic layers Lacustrine facies have been classified
according to elemental composition (TOC TS TIC and TN) grain size and
sedimentary textures (Frugone-Aacutelvarez et al 2017) Four main types of lacustrine
facies were identified in this short core Facies L1 is a laminated (1cm) black to dark
brown diatom-poor silt with relatively higher TOC percentages (TOC about 185)
TS (about of 18) and TOCTN (about 92) Facies L2 is characterized by a
homogeneous black organic-poor silty clay with low TOC (mean= 13) TS (mean=
13) TOCTN (mean= 88) values Facies L3 is a laminated (1cm) black organic-
poor (TOC mean 14) silts with higher TS (mean= 21) and TOCTN ratios
(mean= 88) Facies L3 is interpreted as ldquobaselinerdquo deposition on the distal areas
84
of Vichuqueacuten lake Mineralogy is dominated by mica (muscovite) clay minerals
(chlorite) and plagioclase (albite) Quartz and calcite occur at the top and pyrite
occurs in the lower part of the sequence Facies T is composed by massive banded
sediments 4 cm thick and coarse grain size It is interpreted as an instantaneous
depositional event (for more detail see Frugone-Aacutelvarez et al 2017) In this work
we identified four subunits based on geochemical and stable isotope signals
4 Results
41 Geochemistry and PCA analysis
High positive correlations exist between Al Si K and Ti (r = 078 ndash 096
supplementary Fig S1) and between Zn Zr Rb and Y (r = 072 - 096) which reflect
the silicate content of the watershed-derived sediments (Marzecoacuteva et al 2011) Zr
is commonly associated with minerals more abundant in coarser deposits Thus the
ZrTi profile could reflect grain size (Cuven et al 2010) showing higher variability
in the upper part of the Lake Vichuqueacuten sequence and in the alternation between
facies L1 and L2 Another group of elements made up of Co Fe and Pb displayed
positive correlations (r = 067ndash 097) and represents the input of heavy metals Br
Cl and Ca show a positive correlation (r = 049 ndash 06 p-value = 00001) BrTi ratio
is interpreted as a productivity indicator due to Br having a strong affinity with humic
and organic compounds (Marzecoacuteva et al 2011 Frugone-Aacutelvarez et al 2017) In
our Lago Vichuqueacuten sequence BrTi values are high from 170 to 132 cm and from
36 cm to the top of core where it shows several sharps peaks (Fig 3) The MnFe
ratio is indicative of lake bottom oxygenation (Naecher et al 2013) as under
reducing conditions Mn tends to become more mobile than Fe leading to a
decreased MnFe (Marzecoacuteva et al 2011) Several sharp peaks in MnFe occurred
85
from 127 to 120 cm and from 57 to 30 cm (MgFegt03) The silicate group and the
Br Cl Ca Mn group are negatively correlated (r= -012 and -066)
Principal Component Analysis (PCA) was undertaken on the XRF
geochemical data to investigate the main factors controlling sediment deposition in
Lago Vichuqueacuten (Fig 3) The four main eigenvectors explain 801 of the variance
(supplementary material Table S2) The principal component (PC1) explain 437
of the total of variance and grouped elements are associated with terrigenous input
to the lake Positive values of the biplot have been attributed to higher heavy metals
deposition (Pb Co and Fe) and negative values to higher silicate input (Al K Ti and
Si) in a high energy catchment environment (Zr) The PC2 explains 198 of the
total of variance and highlights the endogenic productivity in the lake The positive
loading is compound by Br Cl Ca and Sr and to a lesser extent by Y Zn Rb and
Mn The PC2 may explain both the precipitation of carbonates (Ca Sr) and biological
production (Br)
86
Figure 3 Principal Component Analysis of XRF geochemical measurements in
VIC13-2B Lago Vichuqueacuten lake sediments
42 Sedimentary units
Based on geochemical and stable isotope analysis we identified four
lithological subunits in the short core sedimentary sequence Our PCA analyses and
Pearson correlations pointed out which variables were better for characterizing the
subunits (Fig 4) in terms of endogenic productivity heavy metals and terrestrial
input The unit 2c (170-131 cm) is characterized by the occurrence of some clastic
layers (L2 from 160 and 150 cm) high δsup1⁵N values (mean= +59 plusmn 04permil) with
Magnetic Susceptibility (MS) BrTi ZrTi and SrCa values increase towards the top
Also it has relatively low δsup1sup3C values (mean= -265 plusmn 11permil) and CNmolar ratios
(mean=78 plusmn 04) The ZrTi show upwards increasing values reaching the highest
values of the sequence at the top of this unit suggesting a coarsening upward trend
and relatively higher depositional energy The MS trend also indicates higher
erosion in the watershed and enhanced delivery of ferromagnetic minerals likely
from soil particles particularly from 150 to 160 cm (Frugone-Aacutelvarez at al 2017)
The subunit 2b (130-118 cm) is also composed of black silts but it has the
lowest MS values of the whole sequence and its onset is marked by a sharp
decrease in ZrTi This subunit is marked by sharp peaks of MnFe (at 126 and 120
cmgt027) SrCa (at 125 and 120 cmgt033) CaTi (at 124 and 120 cmgt199) TOC
(about 26 at 130 124 122 and 118 cm) and δsup1⁵N (gt+67permil at 126 and 120 cm)
BrTi oscillates around low values (about 01) whereas the δsup1sup3C values range
between -262 and -282permil
87
The unit 2a (58-117 cm) shows increasing and then decreasing MS values
and displayed sharps peaks of δsup1⁵N (gt64 at 102 to 99 87 and 75 cm) and CN
(about 10 at 86 74 66 and 60 cm) SrCa (mean = 04 plusmn 013) BrTi (mean = 008
plusmn 002) MnFe (mean = 002 plusmn 001) and δsup1sup3C (mean = -260 plusmn 07permil) oscillate in
low values The upper part of this unit (72 ndash 57 cm) shows an increase of SrCa
(from 03 to 05)
The upper unit 1a (0 - 57 cm) starts with a fine clastic layer (T at 57 to 54
cm) interpreted as deposition during a high-energy event It is characterized by
lower BrTi (mean = 003 plusmn 002) TOC (mean = 12 plusmn 03) and δsup1sup3C (mean= -
266 plusmn 06permil) higher δsup1⁵N (mean = +55 plusmn 08 permil) This unit is made of alternating
fine (lt 1cm) black and dark brown laminae with carbonate (L1 facies) Recently
deposited sediments show decreasing MS values increasing TOC (mean= 15 plusmn
04 ) sharp peaks of BrTi (gt015 at 33 21 5 and 1 cm) and the lowest δsup1⁵N values
of the entire record (mean= 45 plusmn 06 permil) and δsup1sup3C values (mean= -277 plusmn 09 permil)
Heavy metal content increases abruptly in the upper section of the core (2 - 20 cm)
(peaks of FeTi CoTi and PbTi)
88
Figure 4 Sedimentary units of Lago Vichuqueacuten sedimentary sequence Selected
variables illustrate watershed input (Magnetic Susceptibility ZrTi CoTi and MnFe)
endogenic productivity (SrCa CaTi TS) organic matter source (BrTi TOC
CNmolar and stable isotope records (δ13Corg and δ15Nbulk)
43 Recent seasonal changes of particulate organic matter on water column
The average of δ15NPOM from the three sites across to the lake was +86 plusmn 58
permil oscillating widely from -14 to +193permil (Fig 5) Important seasonal differences
occur at 2 and 5 m depth with more negative values during summer (~53 plusmn 52 permil)
than winter (~122 plusmn 40permil) At 20 m depth δ15NPOM values exhibit small seasonal
ranges (mean = +10permil for winter and +87permil for summer) The δ13CPOM average was
-296 plusmn 33permil with slightly seasonal and water column depth differences However
more positive δ13CPOM values occurred during winter (mean=-290 plusmn 41permil) than in
summer (mean= -301 plusmn 19 permil) Like the δ15NPOM values CNPOM (mean= 83 plusmn 17)
displayed important seasonal and water depth differences Lower CNPOM ratios
89
occurred during summer at 2 and 5 m depth (mean= 95 plusmn 17) and higher and more
constant values during winter (mean = 73 plusmn 09) at the same depth At 20 m CNPOM
shows similar values in both winter (70) and summer (74)
Figure 5 Seasonal POM averages from samples taken of the Lago Vichuqueacuten
water column Samples were taken from 2015 to 2018 at 2 (n=16) 5 (n=14) and 20
(n=8) meters depth
44 Stable isotope values across the Lake Vichuqueacuten watershed
Figure 6 shows modern vegetation soil and sediment isotope values found for
the watershed Huintildee tributary and Vichuqueacuten lake The δsup1⁵NFoliar values from
meadow plantations and macrophytes have similar range values with a mean of
+89 plusmn50 permil (n=18) +74 plusmn 75 permil (n=5) +82 plusmn 36 permil (n=6) respectively Native
vegetation has overall lower δsup1⁵NFoliar values with a mean of 14 plusmn 21 permil (n=28 see
Table 2 in supplementary material) The δsup1sup3CFoliar from terrestrial samples exhibit
similar values across the different plant communities (tree plantation mean=-274 plusmn
13permil meadow mean = -275 plusmn 23 permil native species mean=-272 plusmn 14permil) whereas
macrophytes display slightly more negative values with a mean of -287 plusmn 23permil
Macrophytes substrate displayed the highest δsup1⁵N values with a mean of +60 plusmn
14permil (n=3) Similar and relatively high δsup1⁵N values for soil of plantation (mean= +54
plusmn 13permil n=4) meadow (mean= +49 plusmn 04permil n=8) and surface river sediment
90
(mean= +52 plusmn 17permil n=10) Native forest soils oscillate around slightly more
negative values with a mean of +45 plusmn 12permil (n=4) Slightly more positive soil δsup1sup3C
values occur both underneath native forests and in tree plantations with means of -
284 plusmn 23permil and -298 plusmn 13permil respectively compared to either meadow soils
(mean= -302 plusmn 11permil) aquatic sediments underneath macrophytes (-311 plusmn 10permil)
or from surface river sediments (mean= -312 plusmn 10permil)
Figure 6 Stable isotope content of modern soil river sediment and foliar vegetation
used as end members in the sedimentary sequence of Lago Vichuqueacuten a)
Scatterplot of foliar δsup1⁵N and δsup13C values for vegetation from Lago Vichuqueacuten
watershed (plantation meadow and native species) and macrophytes on Lake
Vichuqueacuten See supplementary material for more detail of vegetation types b)
Scatterplot of soil δsup1⁵N and δsup13C values across the watershed the sediments of the
Huintildee and Vichuqueacuten rivers and the fluvial sediment corresponding to the
macrophyte vegetation
45 Land use and cover change from 1975 to 2014
Major land use changes between 1975 CE and 2016 CE in the Lago
Vichuqueacuten watershed are summarized in Figure 7 The watershed has a surface
area of 535329 km2 of which native vegetation (26) and shrublands (53)
represent 79 of the total surface in 1975 Meadows are confined to the valley and
91
represent 17 of watershed surface Tree plantations initially occupied 1 of the
watershed and were first located along the lake periphery By 1989 the areas of
native forests shrublands and meadows had decreased to 22 31 and 14
respectively whereas tree plantations had expanded to 30 These trends
continued almost invariably until 2016 when shrublands and meadows reached 17
and 5 of the total areas while tree plantations increased to 66 Native forests
had practically disappeared by 1989 and then increased up to 7 of the total area
in 2016 (Fig 6) preferentially occupying the southeastern sector of the watershed
Figure 7 Changes in land use and cover between 1975 and 2016 in the Lago
Vichuqueacuten watershed as measured from satellite images The major change is
represented by the replacement of native forest shrubland and meadows by
plantations of Monterrey pine (Pinus radiata)
Figure 8 shows correlations between lake sediment stable isotope values and
changes in the soil cover from 1975 to 2013 Positive relationships occurred
between sediment δsup1⁵N and δsup13C values from core VIC13-2B (unit 1) and the
92
percentage of native forest (r=089 for δsup1⁵N 082 for δsup13C) shrublands (r = 071 for
δsup1⁵N 057 for δsup13C) and meadow cover (r = 088 for δsup1⁵N 081 for δsup13C) All of these
correlations are significant (p value lt 0001) In contrast significant negative
correlations (p lt0001) occurred between tree plantation cover and lake sediment
stable isotope values (δsup1⁵N r = -082 and δsup13C r = -067) native forest (r = -097)
meadows (r = -086) and shrubland (r =-093)
Figure 8 Correlation plots of land use and cover change versus lake sediment
stable isotope values The δsup1⁵N values are positively correlated with native forests
agricultural fields and meadow cover across the watershed Total Plantation area
increases are negatively correlated with native forest meadow and shrubland total
area Significance levels are indicated by the symbols p-values (0 0001 001
005 01 1) lt=gt symbols ( )
93
5 Discussion
51 Seasonal variability of POM in the water column
The stable isotope values of POM can vary during the annual cycle due to
climate and biologic controls namely temperature and length of the photoperiod
which affect phytoplankton growth rates and isotope fractionation in the water
column (Gu et al 2006 Leng et al 2006) POM from Lago Vichuqueacuten surface
samples (2 and 5 m depth) displayed lower δ13CPOM values during the summer than
in winter During C uptake phytoplankton preferentially utilize 12C leaving the
DICpool enriched in 13C Therefore as temperature increases during the summer
phytoplankton growth generates OM enriched in 12C until this becomes depleted
and then the biomas come to enriched u At the onset of winter the DICpool is now
enriched in 13C and despite an overall decrease in phytoplankton production the
OM produced will also be enriched in 13C In contrast δ13CPOM values at 20 m depth
did not reflect these seasonal differences probably due to water-column
stratification that maintains similar temperatures and biological activity throughout
the year
Lake N availability depends on N sources including inputs from the
watershed and the atmosphere (ie deposition of N compounds and fixation of
atmospheric N2) which varies during the hydrologic year The fixation of atmospheric
N2 is an important natural source of N to the lake occurring mainly during the
summer season associated with higher temperature and light (Gu et al 2006)
Because the δ15NPOM of N2 fixers is close to 0permil with almost no overall isotope
fractionation (Leng et al 2006) when N2 is the major N source δ15NPOM values are
typically low However when DIN concentrations are high or alternatively when little
94
N2 fixation occurs δ15NPOM values will be higher (Gu et al 2006) δ15NPOM values
from summer Lago Vichuqueacuten samples were lower than those from winter with large
differences between the seasons (~5permil Fig 5 and 9) Furthermore δ15NPOM values
were high when monthly average temperature was low and monthly precipitation
was high (Fig 9) This pattern is consistent with the dominance of high N2 fixation
by cyanobacteria associated with increased summer temperatures This correlation
of δ15NPOM values with temperature further suggests a functional group shift i e
from N fixers to phytoplankton that uptake DIN The correlation between wetter
months and higher δ15NPOM values could be caused by increased N input from the
watershed due to increased runoff during the winter season The lack of data of the
δ15NPOM between the 30 mm and the 160 mm of precipitation is due to the
mediterranean-type climate that concentrates precipitations in the winter months
Finally Lago Vichuqueacuten is characterized by pronounced seasonality leading to
higher phytoplankton biomass in summer characterized by low δ15NPOM In winter
low biomass production and increased input from watershed is associated to high
δ15NPOM
95
Figure 9 Seasonal changes can drive δ15NPOM values in Lago Vichuqueacuten Data
correspond to average monthly temperature and total monthly precipitation for the
months when the water samples were taken (years 2015 - 2018) P-valuelt005
52 Stable isotope signatures in the Lake Vichuqueacuten watershed
The natural abundance of 15N14N isotopes of soil and vegetation samples
from the Lago Vichuqueacuten watershed appear to result from a combination of factors
isotope fractionation different N sources for plants and soil microorganisms (eg N2
fixation Nitrates) depth in the soil where N uptake by vegetation occurs and N loss
mechanisms (ie denitrification leaching and ammonia volatilization Hogberg
1997) The lowest δsup1⁵Nfoliar values are associated with native species and are
probably due to the contribution of N-fixing species (eg Sophora) (Fig6 and for
more detail see Table S3 in supplementary material) The number native N-fixers
species present in the Chilean mediterranean vegetation are not well known
however so there could be other N-fixing species in this group Alternatively δsup1⁵Nfoliar
values reflect soil N uptake (Kahmen et al 2008) In environments limited by N
plants have δsup1⁵N values similar to the soil δsup1⁵N values however the denitrification
and volatilization of ammonia can lead to the remain N of soil to come enriched in
15N (Cifuentes et al 1989 Craine et al 2009 Bruland and Mckenzie 2010) Soil N
isotope samples from native species communities tends to display relatively high
δsup1⁵N values respect to foliar samples due to loss of N-soil
The higher foliar and soil δsup1⁵N values obtained from samples of meadows
aquatic macrophytes and tree plantations can be attributed to the presence of
greater amounts of nitrate available for plant uptake (Fig 6) Houmlgberg (1997)
suggests that the availability of different N sources in soils (ie nitrates versus
96
ammonia) with different residence times can also explain these δsup1⁵NFoliar values
Indeed Feigin et al (1974) described differences of up to 20permil between ammonia
and nitrates sources Denitrification and nitrification discriminate much more against
15N than N2 fixation does (Craine et al 2009) leading to soils (and plants after
uptake) enriched in 14N
In general multiple processes that affect the isotopic signal result in similar
δsup1⁵N values between the soil of the watershed and the sediments of the river
However POM isotope fluctuations allow to say that more negative δsup1⁵N values are
associated to lake productivity while more positive δsup1⁵N values are associated with
N input from the watershed
δ13C values in Lago Vichuqueacuten watershed reflect CO2 gas exchange between
C3 plants and algae with the atmosphere During photosynthesis plants
discriminate against the heavier stable isotope of carbon (13C) in favor of the lighter
isotope 12C which results in δ13CFoliar values between -220permil and -320permil (Browman
and Cook 2002 Liu et al 2007) Likewise the δ13CFoliar values from Lago Vichuqueacuten
oscillate around -27permil (Fig 6) Plants transform atmospheric CO2 into soil organic
carbon (C) which in turn reflects this initial discrimination against 13C during C
uptake and post photosynthetic fractionation (Farquhar et al 1982 1989 Badeck
et al 2005 Pausch and Kuzyakov 2017) Slightly more positive δ13CFoliar values
(about 15permil) were measured in comparison with their δ13CSoil values This may be
reflecting the C transference from plants to the soil but also a soil-atmosphere
interchange The preferential assimilation of the light isotopes (12C) during soil
respiration carried by the roots and the microbial biomass that is associated with the
decomposition of litter roots and soil organic matter explain this differential
(Berhardt et al 2006 Blagodatskaya et al 2010 Midwood and Millard 2011)
97
In general the δ13CSoil values from the Lago Vichuqueacuten watershed oscillated
around -290permil and did not vary with our plant classification types Here we use
these values as terrestrial-end members to track changes in source OM (Fig 6)
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from the terrestrial watershed By the other hand more positive δ13C
values most likely reflect an increased aquatic OM component as indicated by POM
isotope fluctuations (Fig 9)
53 Recently land use and cover change and its influences on N inputs to the lake
Tree plantations in the Lago Vichuqueacuten watershed have increased greatly in
the last few decades (from 1 in 1975 to 66 in 2016 Fig 7) replacing previous
native forests (from 26 to 7) meadows (17 to 5) and shrublands (53 to
17) In 1975 tree plantations were confined to the lake perimeter with discrete
patches distributed throughout the watershed The ldquoForest Lawrdquo (DFL 701- passed
in 1974) allocated state funding to afforestation efforts and management of tree
plantations which greatly favored the replacement native forests by introduced trees
This increase is marked by a sharp and steady decrease in lake sediment δ15N and
δ13C values because tree plantations function as a nutrient sink whereas other land
uses such as meadows favor nutrient loss from the watershed (Fig8) Bruland and
Mackenzie (2014) noted a decrease in wetland δ15N values when watershed
forested cover increased and concluded that N inputs to the wetlands are lower from
the forested areas as they generally do not export as much N as agricultural lands
A positive correlation between native vegetation and δ15Ncore values can be
explained by the relatively scarce arboreal cover in the watershed in 1975 when
native forest occupied just 26 of the watershed surface whereas shrublands and
98
meadows occupied more than the 70 of the surface of the watershed with the
concomitant elevated loss of N (Fig 7 and Fig 8)
54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years
Sedimentological compositional and geochemical indicators all show changes in
the N cycle associated to LUCC lake dynamics and climate changes (Fig 10) From
the pre-Columbian indigenous settlement including the Spanish colonial period up
to the start of the Republic (1300 - 1800 CE) the introduction of crops such as
quinoa and wheat but also the clearing of land for extensive agriculture would have
favored the entry of N into the lake Conversely major changes observed during the
last century were characterized by a sharp decrease of N input that were coeval
with the increase of tree plantations
From 1365 until 1550 CE (unit 2c 2b and half of 2a Fig 3 and Fig 10) pre-
Columbian agricultural societies planted mostly crops of corn and quinoa (Ramirez
and Vidal 1985) This period is characterized by large peaks seen in the δsup1⁵N record
(eg 1403 1452 1476 and 1505) with overall high δsup1⁵N values (up to 5permil) indicating
that N input from watershed was elevated and oscillating to the beat of the NT These
positive δsup1⁵N peaks could be due to several causes including a) the clearing of land
for farming b) N loss via denitrification which would be generally augmented in
anoxic environments (Diebel et al 2009) such as Lago Vichuqueacuten (low MnFe
values about 0001 Fig 4) or c) climate especially cold-wet winters or hot-dry
summers can also exert control on the δsup1⁵N record Indeed tree-ring records and
summer temperature reconstructions show overall wetcold conditions during this
period (Christie et al 2009 Von Gunten et al 2009 Fig 10) Increased
precipitation would bring more sediment (and nutrients) from the watershed into the
99
lake and increase lake productivity which is also detected by the geochemical
proxies for productivity (peaks of BrTi SrCa and TOC Fig 4 and Fig 10) (see also
Frugone-Alvarez et al 2017)
Figure 10 Changes in the N availability during the last six centuries in Lago
Vichuqueacuten a) lake sediment δsup1⁵N values displayed more positive values during the
prehistoric period Spanish Colony and the starting 19th century which is associated
with enhanced N input from the watershed by extensive clearing and crop
plantations The inset shows this relationship between sediment δsup1⁵N and
100
percentage of meadow cover over the last 30 years b) Summer temperature
reconstruction from central Chile (von Gunten et al 2009) showing a
correspondence with cold phases and high δsup1⁵N values at Vichuqueacuten except for the
last 100 years c) Palmer Drought Severity Index (PDSI) is an interannual moisture
variability reconstruction for late springndashearly summer during the last six centuries
(Christie et al 2009) Grey shadow indicating higher precipitation periods
From 1550 and 1810 CE (unit 2a) and during the Colonial period several peaks
of δ15N could be further related not only to high precipitation (Fig 10 grey shadow)
but also pulses of enhanced N input from the watershed linked to human land use
In 1550 CE Juan Cuevas was granted lands and indigenous workers under the
encomienda system for agricultural and mining development of the Vichuqueacuten
village (Ramiacuterez and Vidal 1985) Historical documents show that around 1579 CE
the Vichuqueacuten watershed was occupied by indigenous communities dedicated to
wheat plantations and vineyards wood extraction and gold mining (Odone 1998)
The introduction of the Spanish agricultural system implied not just a change in the
types of crops used (from quinoa to vineyards and wheat) but also a clearing of
native species for the continuous increase of agricultural surface and wood
extraction During the Colonial period wheat from Vichuqueacuten was exported to Peru
(Ramiacuterez and Vidal 1985) Armesto et al (2009) indicate that during the XVIII and
XIX centuries the extraction of wood for mining operations was important enough to
cause extensive loss of native forests The independence and instauration of the
Chilean Republic did not change this prevailing system Increases in the
contributions of N to the lake during the second half of the XIX century (peaks in
δsup1⁵N ca 1870 CE) could be due to expanding farm land dedicated for wheat
101
production and increased commercial trade with California and Canada (Ramiacuterez
and Vidal 1985)
In contrast LUCC in the last century are clearly related to the development of
large-scale tree plantations (Fig 7) The δsup1⁵N record reaches the lowest values of
the entire sequence in the last few decades (Fig 10) A marked increase in lake
productivity NT concentration and decreasing sediment input is synchronous (unit
1 Fig 4) with trees replacing meadows shrublands and areas with native forests
(Fig 7 and 8) The implementation of the lsquoForest Lawrsquo in 1974 had a stronger impact
on the landscape and lake ecosystem dynamics than the impacts of ongoing climate
change in the region which is much more recent (Garreaud et al 2018) although
the prevalence of hot dry summers seen over the last decade would also be
associated with low δsup1⁵N values and increase of NT (Fig 10) Heavy metals ratios
(FeTi CoTi and PbTi) show notable increases in lake sediments from 1992 to 2011
CE (Fig 4) Although this could be related to mining in the El Maule region the
closest mines are 60 Km away (Pencahue and Romeral) so local factors related to
shoreline urbanization for the summer homes and an increase in tourist activity
could also be a major factor
6 Conclusions
The N isotope signal in the watershed depends on the rates of exchange
between vegetation and soil cover (Fig 11) Plants preferentially uptake 14N and the
underlying soils become enriched in 15N especially when the terrestrial ecosystem
is N-limited andor significant N loss occurs (ie denitrification andor ammonia
volatilization) LUCC in the Lago Vichuqueacuten watershed have heavily modified the
links between terrestrial and aquatic ecosystems with agriculture practices
102
contributing more N to the lake than tree plantations or native forests In situ lake
processes can also fractionate N isotopes An increase of N-fixing species results
in OM depleted in 15N which results in POM with lower δsup1⁵N values during these
periods During winter phytoplankton is typically enriched in 15N due to the
decreased abundance of N-fixing species and increased N input from the
watershed This in turn generates the highest δsup1⁵NPOM values in Lago Vichuqueacuten
Periods of elevated productivity tend to deplete the dissolved inorganic N of 14N
resulting in even higher δ15N values
Our study illustrates the usefulness of δsup1⁵N as a tool for reconstructing the past
influence of LUCC on N availability in lake ecosystems To constrain the relative
roles of the diverse forcing mechanisms that can alter N cycling in mediterranean
ecosystems all main components of the N cycle should be monitored seasonally
(or monthly) including the measurements of δ15N values in land samples
(vegetation-soil) as well as POM
103
Figure 11 Summary of human and environmental factors controlling the δ15N
values of lake sediments Particulate organic matter(POM) δ15N values in
mediterranean lakes are driven by N input from the watershed that in turn depend
on land use and cover changes (ie forest plantation agriculture) andor seasonal
changes in N sources andor lake ecosystem processes (ie bioproductivity redox
condition denitrification) Arrows indicate the direction of N fluxes (inputoutput from
the N cycle) N cycle processes that deplete lake sediments of 15N are shown in
blue whereas those that enrich sediments in 15N are shown in red
104
Supplementary material
Figure S1 Pearson correlate coefficient between geochemical variables in core
VIC13-2B Positive and large correlations are in blue whereas negative and small
correlations are in red (p valuelt0001)
Figure S2 Principal Component Analysis of geochemical elements from core
VIC13-2B
105
Table S1 Lago Vichuqueacuten radiocarbon samples
RADIOCARBON
LAB CODE
SAMPLE
CODE
DEPTH
(m)
MATERIAL
DATED
14C AGE ERROR
D-AMS 029287
VIC13-2B-
1 043 Bulk 1520 24
D-AMS 029285
VIC13-2B-
2 085 Bulk 1700 22
D-AMS 029286
VIC13-2B-
2 124 Bulk 1100 29
Poz-63883 Chill-2D-1 191 Bulk 945 30
D-AMS 001133
VIC11-2A-
2 201 Bulk 1150 44
Poz-63884
Chill-2D-
1U 299 Bulk 1935 30
Poz-64089
VIC13-2D-
2U 463 Bulk 1845 30
Poz-64090
VIC13-2A-
3U 469 Bulk 1830 35
D-AMS 010068
VIC13-2D-
4U 667 Bulk 2831 25
Poz-63886
VIC13-2D-
4U 719 Bulk 3375 35
106
D-AMS 010069
VIC13-2D-
5U 775 Bulk 3143 27
Poz-64088
VIC13-2D-
5U 807 Bulk 3835 35
D-AMS-010066
VIC13-2D-
7U 1075 Bulk 6174 31
Poz-63885
VIC13-2D-
7U 1197 Bulk 6440 40
Poz-5782 VIC13-15 DIC 180 25
Table S2 Loadings of the trace chemical elements used in the PCA
Elementos PC1 PC2 PC3 PC4
Zr 0922 0025 -0108 -0007
Zn 0913 -0124 -0212 0001
Rb 0898 -0057 -0228 0016
K 0843 0459 0108 0113
Ti 0827 0497 0060 -0029
Al 0806 0467 0080 0107
Si 0803 0474 0133 0136
Y 0784 -0293 -0174 0262
V 0766 0455 0090 -0057
Br 0422 -0716 -0045 0226
Ca 0316 -0429 0577 0489
Sr 0164 -0420 0342 -0182
Cl 0151 -0781 -0397 0162
107
Mn -0121 -0091 0859 0095
S -0174 -0179 -0051 0714
Pb -0349 0414 -0282 0500
Fe -0700 0584 -0023 0280
Co -0704 0564 -0107 0250
Table S3 Stable Isotope values from vegetation of Lago Vichuqueacutenacutes watershed
Taxa Classification δsup1⁵N δsup1sup3C CN
molar
Poaceae Meadow 1216 -2589 3602
Juncacea Meadow 1404 -2450 3855
Cyperaceae Meadow 1031 -2596 1711
Taraxacum
officinale Meadow 836 -2400 2035
Poaceae Meadow 660 -2779 1583
Poaceae Meadow 453 -2813 1401
Poaceae Meadow 966 -2908 4010
Juncus Meadow 1247 -2418 3892
Poaceae Meadow 747 -3177 6992
Poaceae Meadow 942 -2764 3147
Poaceae Meadow 1479 -2634 2895
Poaceae Meadow 1113 -2776 1795
Poaceae Meadow 2215 -2737 7971
Poaceae Meadow 1121 -2944 2934
Poaceae Meadow 638 -3206 1529
108
Macrophytes Macrophytes 886 -3044 2286
Macrophytes Macrophytes 1056 -2720 2673
Macrophytes Macrophytes 769 -3297 1249
Macrophytes Macrophytes 967 -2763 1442
Macrophytes Macrophytes 959 -2670 2105
Macrophytes Macrophytes 334 -2728 1038
Acacia dealbata
Introduced
species 656 -2696 1296
Acacia dealbata
Introduced
species 487 -2941 1782
Acacia dealbata
Introduced
species 220 -2611 3888
Luma apiculata Native species 433 -2542 4135
Luma apiculata Native species 171 -2664 7634
Luma apiculata Native species -001 -2736 6283
Luma apiculata Native species 029 -2764 6425
Azara sp Native species 159 -2868 8408
Azara sp Native species 101 -2606 2885
Baccharis concava Native species 104 -2699 5779
Baccharis concava Native species 265 -2488 4325
Baccharis concava Native species 287 -2562 7802
Baccharis concava Native species 427 -2781 5204
Baccharis linearis Native species 190 -2610 4414
Baccharis linearis Native species 023 -2825 5647
109
Peumus boldus Native species 042 -2969 6327
Peumus boldus Native species 205 -2746 4110
Peumus boldus Native species 183 -2743 6293
Chusquea quila Native species 482 -2801 4275
Poaceae meadow 217 -2629 7214
Lobelia sp Native species 224 -2645 3963
Lobelia sp Native species -091 -2565 4538
Aristotelia chilensis Native species -035 -2785 5247
Aristotelia chilensis Native species -305 -2889 2305
Aristotelia chilensis Native species 093 -2836 5457
Chusquea quila Native species 173 -2754 3534
Chusquea quila Native species 045 -2950 6739
Quillaja saponaria Native species 223 -2838 9385
Scirpus meadow 018 -2820 7115
Sophora sp Native species -184 -2481 2094
Sophora sp Native species -181 -2717 1721
Pinus radiata
Introduced
trees 1581 -2602 3679
Pinus radiata
Introduced
trees 1431 -2784 4852
Pinus radiata
Introduced
trees -091 -2708 9760
Pinus radiata
Introduced
trees 153 -2568 3470
110
Salix sp
Introduced
trees 632 -2878 1921
LITERATURE CITED
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A historical
framework for land cover change in southwestern South America in the past 15000
years Land use policy 27 148ndash160
httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the next
carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014
httpsdoiorg101002eft2235
Blaauw M Christeny JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474
httpsdoiorg10121411-BA618
Bowman DMJS Cook GD 2002 Can stable carbon isotopes (δ13C) in soil
carbon be used to describe the dynamics of Eucalyptus savanna-rainforest
boundaries in the Australian monsoon tropics Austral Ecol
httpsdoiorg101046j1442-9993200201158x
Brahney J Ballantyne AP Turner BL Spaulding SA Otu M Neff JC 2014
Separating the influences of diagenesis productivity and anthropogenic nitrogen
deposition on sedimentary δ15N variations Org Geochem 75 140ndash150
httpsdoiorg101016jorggeochem201407003
111
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409
httpsdoiorg102134jeq20090005
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego R
Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and
environmental change from a high Andean lake Laguna del Maule central Chile
(36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the
Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from
tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Craine JM Elmore AJ Aidar MPM Dawson TE Hobbie EA Kahmen A
Mack MC Kendra K Michelsen A Nardoto GB Pardo LH Pentildeuelas J
Reich B Schuur EAG Stock WD Templer PH Virginia RA Welker JM
Wright IJ 2009 Global patterns of foliar nitrogen isotopes and their relationships
with cliamte mycorrhizal fungi foliar nutrient concentrations and nitrogen
availability New Phytol 183 980ndash992 httpsdoiorg101111j1469-
8137200902917x
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty
Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-
010-9453-1
Diebel M Vander Zanden MJ 2012 Nitrogen stable isotopes in streams stable
isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19
1127ndash1134 httpsdoiorg10189008-03271
112
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in freshwater
wetlands record long-term changes in watershed nitrogen source and land use SO
- Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash
2916
Fariacuteas L Castro-Gonzales M Cornejo M Charpentier J Faundez J
Boontanon N Yoshida N 2009 Denitrification and nitrous oxide cycling within the
upper oxycline of the oxygen minimum zone off the eastern tropical South Pacific
Limnol Oceanogr 54 132ndash144
Farquhar GD Ehleringer JR Hubick KT 1989 Carbon Isotope Discrimination
and Photosynthesis Annu Rev Plant Physiol Plant Mol Biol
httpsdoiorg101146annurevpp40060189002443
Farquhar GD OrsquoLeary MH Berry JA 1982 On the relationship between
carbon isotope discrimination and the intercellular carbon dioxide concentration in
leaves Aust J Plant Physiol httpsdoiorg101071PP9820121
Fogel ML Cifuentes LA 1993 Isotope fractionation during primary production
Org Geochem httpsdoiorg101007978-1-4615-2890-6_3
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A
Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-
resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS)
implications for past sea level and environmental variability J Quat Sci 32 830ndash
844 httpsdoiorg101002jqs2936
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924
httpsdoiorg104319lo20095430917
113
Gerhart LM McLauchlan KK 2014 Reconstructing terrestrial nutrient cycling
using stable nitrogen isotopes in wood Biogeochemistry 120 1ndash21
httpsdoiorg101007s10533-014-9988-8
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and oxygen
isotope fractionation during dissimilatory nitrate reduction by denitrifying bacteria 53
2533ndash2545 httpsdoiorg10230740058342
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater
eutrophic lake Limnol Oceanogr 51 2837ndash2848
httpsdoiorg104319lo20065162837
Harris D Horwaacuteth WR van Kessel C 2001 Acid fumigation of soils to remove
carbonates prior to total organic carbon or CARBON-13 isotopic analysis Soil Sci
Soc Am J 65 1853 httpsdoiorg102136sssaj20011853
Houmlgberg P 1997 Tansley review no 95 natural abundance in soil-plant systems
New Phytol httpsdoiorg101046j1469-8137199700808x
Kylander ME Ampel L Wohlfarth B Veres D 2011 High-resolution X-ray
fluorescence core scanning analysis of Les Echets (France) sedimentary sequence
New insights from chemical proxies J Quat Sci 26 109ndash117
httpsdoiorg101002jqs1438
Lara A Solari ME Prieto MDR Pentildea MP 2012 Reconstruccioacuten de la
cobertura de la vegetacioacuten y uso del suelo hacia 1550 y sus cambios a 2007 en la
ecorregioacuten de los bosques valdivianos lluviosos de Chile (35o - 43o 30acute S) Bosque
(Valdivia) 33 03ndash04 httpsdoiorg104067s0717-92002012000100002
Lehmann M Bernasconi S Barbieri A McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during
114
simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66
3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007
Liu G Li M An L 2007 The Environmental Significance Of Stable Carbon
Isotope Composition Of Modern Plant Leaves In The Northern Tibetan Plateau
China Arctic Antarct Alp Res httpsdoiorg1016571523-0430(07-505)[liu-
g]20co2
Marzecovaacute A Mikomaumlgi A Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54
httpsdoiorg103176eco2011105
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash
1643 httpsdoiorg1011770959683613496289
Meyers PA 2003 Application of organic geochemistry to paleolimnological
reconstruction a summary of examples from the Laurention Great Lakes Org
Geochem 34 261ndash289 httpsdoiorg101016S0146-6380(02)00168-7
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland
Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Odone C 1998 El Pueblo de Indios de Vichuqueacuten siglos XVI y XVII Rev Hist
Indiacutegena 3 19ndash67
Pausch J Kuzyakov Y 2018 Carbon input by roots into the soil Quantification of
rhizodeposition from root to ecosystem scale Glob Chang Biol
httpsdoiorg101111gcb13850
115
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98
httpsdoiorg1011772053019614564785
Sun W Zhang E Jones RT Liu E Shen J 2016 Biogeochemical processes
and response to climate change recorded in the isotopes of lacustrine organic
matter southeastern Qinghai-Tibetan Plateau China Palaeogeogr Palaeoclimatol
Palaeoecol httpsdoiorg101016jpalaeo201604013
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of
different trophic status J Paleolimnol 47 693ndash706
httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl
httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M 2009 High-resolution quantitative climate
reconstruction over the past 1000 years and pollution history derived from lake
sediments in Central Chile Philos Fak PhD 246
Woodward C Chang J Zawadzki A Shulmeister J Haworth R Collecutt S
Jacobsen G 2011 Evidence against early nineteenth century major European
induced environmental impacts by illegal settlers in the New England Tablelands
south eastern Australia Quat Sci Rev 30 3743ndash3747
httpsdoiorg101016jquascirev201110014
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K Yeager
KM 2016 Different responses of sedimentary δ15N to climatic changes and
116
anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau
J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
117
DISCUSION GENERAL
El Nitroacutegeno (N) es un elemento clave que controla la dinaacutemica y
funcionamiento de ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al
1997) Pese a que el N (N2) es muy abundante en la atmoacutesfera (78) es escaso
en los ecosistemas pues la mayoriacutea de la biota no puede acceder a su forma
molecular (Galloway et al 1995 Zaehle et al 2013) En este sentido la entrada
natural de N en los ecosistemas es viacutea fijacioacuten bioloacutegica del N2 por bacterias que lo
convierten en compuestos bioloacutegicamente disponibles (Battye et al 2017) Debido
a que la fijacioacuten es un proceso energeacuteticamente costoso los ecosistemas
comuacutenmente estaacuten limitados por N (Vitousek and Howarth 1991) Los LUCC
contribuyen al incremento del N disponible y son una de las principales causas de
eutroficacioacuten y contaminacioacuten de los cuerpos de agua (Woodward et al 2012)
En Chile central los LUCC principalmente relacionados con las actividades
agriacutecolas y forestales han causado grandes cambios en el ciclo del N debido al
118
reemplazo de la cobertura natural por un monocultivo y al uso de fertilizantes que
modifican los aportes de MO y N a los cuerpos de agua El programa de
estabilizacioacuten de laderas impulsada por el gobierno (Albert 1900) y la ley forestal
de 1974 han fomentado la forestacioacuten con plantaciones de Pinus radiata y
Eucalyptus globulus y en el caso de Lago Vichuqueacuten estas actualmente cubren maacutes
del 60 de su cuenca (Cap2 Fig 5) Ademaacutes en Laguna Matanzas la
sobreexplotacioacuten del agua en conjunto con el cambio climaacutetico actual el que ha
conducido a una de las megasequiacuteas maacutes severas en las uacuteltimas deacutecadas
(Garreaud et al 2017) generoacute una combinacioacuten letal que secoacute el lago en tan solo
10 antildeos (Cap1 Fig1) Las evidencias obtenidas a partir de estos dos lagos
permiten identificar las huellas del Antropoceno en Chile central basadas en el
registro sedimentario lacustre
La transferencia de N desde los ecosistemas terrestres a acuaacuteticos es un
proceso natural con controles bioacuteticos climaacuteticos y geoloacutegicos El enlace
hidroloacutegico entre los ecosistemas terrestres y acuaacuteticos hace que el estado troacutefico
de los lagos esteacute vinculado al de su cuenca (McLauchlan et al 2013) En Chile
central se sabe poco del impacto de los LUCC sobre el ciclo de nutrientes en los
ecosistemas lacustres aunque al menos desde la colonizacioacuten espantildeola se tienen
registros de influencia humana en las cuencas Durante la colonia espantildeola
Laguna Matanzas fue una hacienda ganadera que exportaba productos caacuternicos al
Peruacute (Contreras-Lopez et al 2014) A su vez en el Lago Vichuqueacuten se realizaban
extensas actividades agriacutecolas hasta comienzos del s XX como por ejemplo
cultivos de vid y trigo ademaacutes de extraccioacuten de madera y mineriacutea de oro (Odone
1997) En las uacuteltimas deacutecadas los LUCC han estado vinculados principalmente con
el desarrollo forestal al compaacutes de una estrategia gubernamental de fomentar con
119
incentivos financieros (Ley de Bosques DFL nordm265 de 1931 y DFL 701 de 1974)
esta actividad El incremento de la superficie forestal es especialmente fuerte en
ambos lagos a partir de la deacutecada de 1980 y hasta 2016 (Laguna de Matanzas 0-
17 Lago Vichuqueacuten 1-66) Este aumento habriacutea sido en detrimento del bosque
nativo y los pastizales (Cap 1 Fig 5 y Cap 2 Fig 8) El incremento de la superficie
forestal generariacutea una disminucioacuten de los aportes de sedimentos y nutrientes al lago
y en este sentido un cambio de estado en los flujos de N (e g tipping points) que
a su vez ha quedado registrado en la sentildeal isotoacutepica (δ15N) y la acumulacioacuten de
MO en los sedimentos lacustres
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central
Los sedimentos lacustres son verdaderos ldquosumiderosrdquo que pueden llegar a
registrar lo que ocurre en cuanto a la historia ambiental de su cuenca En esta tesis
se utilizan los isoacutetopos estables de N (15N y 14N) en sedimentos lacustres para
reconstruir los cambios en la disponibilidad de N en el tiempo y establecer la
magnitud de impacto generado por actividades humanas El fraccionamiento
cineacutetico en los lagos es mediado por procesos bioloacutegicos que mediante la
asimilacioacuten de N genera un producto (eg fitoplancton) maacutes ldquoligerordquo (valores maacutes
bajos de 15N) que su remanente (eg DIN) (Fogel y Cifuentes 1993) Sin embargo
en periacuteodos en que los lagos estaacuten limitados por N el fitoplancton discrimina menos
y la MO resultante queda enriquecida en 15N (Fig 1 a y b) Ademaacutes la
desnitrificacioacuten acentuada en lagos de fondo anoacutexicos (eg Laguna Matanzas
entre 1800 y 1940 Cap1 Fig 6) conduce a un enriquecimiento en 15N de los
sedimentos y de la columna de agua Esta informacioacuten queda reflejada en la MO
120
de los sedimentos lacustres lo que permite conocer la disponibilidad del N en el
tiempo a partir de las variaciones de 15N
En esta tesis analizamos la MO de los sedimentos lacustres para reconstruir
la influencia de los LUCC en los aportes de sedimentos y N al lago En teacuterminos de
asimilacioacuten de N se puede distinguir entre dos grupos principales de productores
primarios que componen el POM (Fig1)
1 Los que asimilan el DIN compuesto por amonio nitratos y nitritos Aquiacute el
δ15N resultante fluctuaraacute en torno a valores maacutes positivos en la medida que
la productividad se incrementa o no hay reposicioacuten del DIN (Fig 1)
2 Los fijadores de N Dado que es un proceso energeacuteticamente costoso en
ambientes que no estaacuten limitados por N muchas veces son excluiacutedas
competitivamente por el resto del fitoplancton Si el DIN queda agotado por
el incremento de la productividad (o bien si el ambiente estaacute limitado ya sea
por N o por otros nutrientes) los fijadores de N pueden florecer y la MO que
se genera a partir de ello tendraacute un δ15N con valores en torno a 0permil
De este modo la MO en los sedimentos lacustres dependeraacute de la
composicioacuten de productores primarios (ie fijadores vs asimiladores del DIN)
ademaacutes de los aportes desde la cuenca tanto de MO como del N de la cuenca que
pasa a formar parte del DIN (eg fertilizantes estieacutercol) (Fig 1)
La MO de los lagos estudiados en esta tesis ha sido analizada a partir de
variables geoquiacutemicas isotoacutepicas frotis y estaacute compuesta principalmente por
diatomeas y MO amorfa A partir de los datos obtenidos en esta tesis los valores
de δ15N aumentan cuando hay una mayor cobertura agriacutecola o pastizales A su vez
tiende a disminuir cuando aumenta la cobertura arboacuterea independiente si esta es
por plantaciones forestales o por bosque nativo
121
Por otra parte la dinaacutemica del lago tambieacuten imprime caracteriacutesticas
especiales en el POM observaacutendose variaciones estacionales en los valores
δ15NPOM Durante la estacioacuten caacutelida y seca se observan valores maacutes bajos que
durante la estacioacuten friacutea y huacutemeda posiblemente relacionado con el incremento de
la contribucioacuten de especies fijadoras de N (Lago Vichuqueacuten Fig 9 Cap 2) durante
el verano En la estacioacuten friacutea y huacutemeda el DIN seriacutea la principal fuente de N Las
mayores entradas de MO y N terrestre debidos a un incremento del lavado de la
cuenca como consecuencia de las precipitaciones y la mineralizacioacuten de la MO
podriacutean estimular la produccioacuten de MO lacustre por crecimiento del fitoplancton
Como consecuencia se observan tendencias decrecientes de los valores de
δ15N en la MO sedimentaria cuando la productividad en el lago estaacute relacionada
con especies fijadoras de N mientras que el δ15N muestra aumentos cuando la
productividad del lago estaacute asociada principalmente al consumo del DIN pero
tambieacuten cuando predominan procesos de perdida de N (eg desnitrificacioacuten Fig
1)
Hemos identificado tres fases en la dinaacutemica del δ15N para ambos lagos
Entre 1800 y 1940 CE la cuenca de laguna de Matanzas estaba dominada por
actividades agriacutecolas y ganaderas (δ15Nsuelo gt 4permil Cap 1 Fig 4 y 6) Las entradas
de sedimentos y MO desde la cuenca son altas (δ13C lt -209permil ZrTi ~029 AlTi
~013) y coincide con un clima maacutes huacutemedo y friacuteo (Von Gunten et al 2009
Christiansen et al 2011) El lago teniacutea un nivel de agua maacutes profundo dominado
por condiciones anoacutexicas (MnFe ~0013) que contribuiriacutean a mayores valores de
δ15N en la columna de agua y por ende de los sedimentos acumulados en el fondo
debido a la peacuterdida diferencial de 14N viacutea desnitrificacioacuten (Fig 7 Cap1) La baja
122
produccioacuten de MO lacustre (BrTi ~002 TOC~17) coincide con un bajo contenido
de NT (~478 μg) y valores maacutes positivos de δ15N (gt 4permil)
La actividad agriacutecola tambieacuten dominaba en la cuenca del Lago Vichuqueacuten
durante este periacuteodo (δ15Nsuelo gt4permil Cap2 Fig 6) El aporte sedimentario desde la
cuenca es alto (AlTi ~029 ZrTi ~032) y fue favorecido por mayores
precipitaciones a las actuales (Christiansen et al 2011) El Lago Vichuqueacuten es un
lago estratificado anoacutexico (MnFe ~002) y profundo (~30 m) por lo que la
desnitrificacioacuten juega un rol importante en el enriquecimiento de la MO
sedimentaria La baja produccioacuten de MO (BrTi ~007 TOC ~12) de origen
lacustre (δ13C ~ -268permil) coincide con un bajo contenido de NT (~309μg +6) y
valores maacutes positivos de δ15N (56permil +03)
Durante esta fase en ambos lagos los aportes de N de la cuenca parecen
ser los principales responsables en las fluctuaciones del δ15N Esta sentildeal podriacutea
estar amplificada por bajas temperaturas (que afectan la productividad lacustre) y
altas precipitaciones que favorecen el lavado de la cuenca y aumentan el aporte de
sedimentos y MO desde la cuenca predominantemente agriacutecola
Entre 1940 y 1980 CE en Laguna Matanzas se observa un incremento en
la acumulacioacuten de MO algal (TOC ~2 +1 δ13C ~-230+09permil) El ambiente
deposicional es ligeramente menos turbulento que el anterior (AlTi ~011+001
ZrTi ~03+001) y el lago estaacute en una fase de transicioacuten hacia condiciones maacutes
oacutexicas (MnFe ~002+0004) y maacutes productivas (BrTi ~004+0007) A su vez δ15N
tiende hacia valores maacutes negativos (δ15N ~30permil+03) mientras que el NT aumenta
oscilando en antifase con el δ15N
En Lago Vichuqueacuten en cambio se observa un ligero incremento en la
acumulacioacuten de MO (TOC ~13 +02) de origen terrestre (δ13C ~-27permil +04) La
123
productividad es similar al periacuteodo anterior (BrTi ~007+003) y el ambiente
deposicional es menos turbulento (AlTi ~02 +009 Zrti ~033 +007) El δ15N y el
NT oscilan en torno a valores ligeramente maacutes bajos (δ15N ~51permil +04 NT ~292μg
+6) Este periacuteodo se caracteriza por ser maacutes caacutelido que el anterior lo que
posibilitariacutea un incremento en la productividad lacustre en Laguna Matanzas pero
que no es observada en el Lago Vichuqueacuten
Entre 1980 y la actualidad en Laguna Matanzas habriacutea un incremento de la
acumulacioacuten de MO (TOC ~59 + 3) asociada a un incremento en la productividad
del lago (BrTi ~009+005) y a los aportes de MO terrestres (δ13C ~-24 + 2permil) El
lago se vuelve maacutes oacutexico (Mn ~0003 +0006) y las entradas de sedimento
disminuyen (AlTi~ 009 +002) El δ15N tiende a valores maacutes negativos (δ15N ~13permil
+1) y el NT aumenta de 20 a 55 μg (NT ~346 + 9 μg) tomando valores sin
precedentes en todo el registro En los sedimentos lacustres del Lago Vichuqueacuten
tambieacuten se observa un incremento de la acumulacioacuten de MO (TOC ~17 + 05)
asociada a un incremento en la productividad del lago (BrTi ~01 + 003) y las
entradas de sedimento desde la cuenca disminuyen (AlTi ~005 + 005) El δ15N
(~43 + 05permil) tiende a valores maacutes negativos y el NT incrementa de 20 a 55 μg (NT
~346 + 9 μg) oscilando en antifase
Durante esta fase en ambos lagos se observa un aumento en la
acumulacioacuten de MO sedimentaria un incremento en el NT y valores maacutes negativos
de δ15N que coincide con el incremento de la superficie forestal de las cuencas
(Laguna Matanzas gt17 Lago Vichuqueacuten gt60)
124
Figura 2 Comparacioacuten de los cambios del ciclo del N entre Laguna Matanzas y
Lago Vichuqueacuten con los procesos biogeoquiacutemicos contrastantes en rojo En L
Matanzas las condiciones oacutexicas podriacutean estar favoreciendo la nitrificacioacuten del
amonio En Vichuqueacuten las condiciones anoacutexicas favorecen un aumento en δ15N de
la MO en los sedimentos por desnitrificacioacuten y mineralizacioacuten
Los ambientes mediterraacuteneos en el que los lagos del presente estudio se
encuentran insertos estaacuten caracterizados por una estacioacuten seca prolongada y las
precipitaciones ocurren en eventos puntuales alcanzando altos montos
pluviomeacutetricos en poco tiempo Las lluvias favorecen un lavado de la cuenca la
perdida de N y otros nutrientes Por lo que es esperable que los sedimentos del
lago reflejen mejor los valores isotoacutepicos de los suelos de la cuenca durante los
periacuteodos con altos montos pluviomeacutetricos En el Lago Vichuqueacuten por ejemplo el
125
POM superficial (2 y 5 m de profundidad) despliega valores isotoacutepicos maacutes
positivos en invierno presumiblemente como resultado de mayores aportes de MO
y N de su cuenca (Fig 1b) En ambos lagos los valores maacutes positivos en los
sedimentos lacustres ocurren durante periacuteodos con mayores montos pluviomeacutetricos
(Cap1 Fig 6 y Cap 2 Fig 12)
Ademaacutes del efecto de la precipitacioacuten en el aporte de nutrientes al lago en
esta tesis hemos encontrado que los mayores aportes de N desde la cuenca se dan
cuando la cobertura vegetal del suelo es menor En Laguna Matanzas el desarrollo
de la actividad ganadera desde la llegada de los espantildeoles a la cuenca habriacutea
incrementado los aportes de N al lago Los valores de δ15N en los sedimentos
lacustres depositados durante este periacuteodo son los maacutes positivos de todo el registro
(~4permil) A su vez en Lago Vichuqueacuten los valores isotoacutepicos maacutes positivos se
registran entre 1300 y 1900 CE que coinciden con el mayor desarrollo de
actividades agriacutecola y de extraccioacuten de maderas de la cuenca (Fig 10 Cap 2)
Los recientes LUCC (a partir de 1975) en las cuencas de Laguna Matanzas
y Lago Vichuqueacuten estaacuten caracterizados por un incremento de la superficie forestal
y agriacutecola en reemplazo de la cobertura de pastizal (Laguna Matanzas) y bosque
nativo (Lago Vichuqueacuten) Los valores de δ15N en los testigos sedimentarios fueron
maacutes positivos cuanto mayor la superficie agriacutecola de la cuenca y maacutes negativos
cuanto mayor la superficie forestal Aunque por los alcances de esta tesis no
podemos ser concluyentes respecto del origen del NT (aloacutectonoautoacutectono) parece
ser que esta maacutes ligado a los aportes de la cuenca pese a la disminucioacuten del aporte
sedimentario observado en ambos lagos Las plantaciones forestales a diferencia
del bosque nativo son fertilizadas con una mezcla de N P y K (Toral et al 2005)
Por lo que pese a la disminucioacuten del aporte sedimentario la concentracioacuten de
126
nutrientes en el aporte al lago puede verse incrementada por el tipo de manejo
forestal con respecto al bosque nativo
Los resultados del primer capiacutetulo demuestran que 1) las plantaciones
forestales yo bosque nativo retiene maacutes sedimentos y MO mientras que el uso de
suelo agriacutecola y los pastizales destinados al desarrollo de la actividad ganadera lo
libera 2) Al parecer la desnitrificacioacuten es el proceso natural maacutes importante de
perdida de N en lagos costeros y favorece el enriquecimiento de 15N tanto en la
columna de agua (ie 15N-DIN) como en los sedimentos una vez enterrados y 3) la
desnitrificacioacuten depende de los cambios en las condiciones REDOX en el lago La
oxigenacioacuten en lagos poco profundos -como Laguna Matanzas- es altamente
fluctuante a escalas decadal-centenal depende de la profundidad de la laacutemina de
agua y de la variabilidad de la precipitacioacuten En este sentido en Laguna Matanzas
habriacutea condiciones anoacutexicas en ambientes cuando el lago alcanza niveles maacutes
altos Estas condiciones se observan entre 1820 y 1940 CE y son coetaacuteneas con
episodios maacutes huacutemedos y friacuteos (von Gunten et al 2009 Christie et al 2009) pero
tambieacuten con una fuerte actividad ganadera en la cuenca
Las fuentes variables de N y su fluctuacioacuten en el tiempo hacen necesario
contar con un anaacutelogo moderno que permite usar al δ15N de los sedimentos
lacustres como un indicador indirecto de los cambios en la disponibilidad de N en
el tiempo Por ello en el segundo capiacutetulo integramos muestras de suelo-
vegetacioacuten y un monitoreo de POM de la columna de agua en verano e invierno La
composicioacuten isotoacutepica de POM estariacutea reflejando cambios de la fuente de N (fijacioacuten
vs consumo del DIN) que variacutea durante el antildeo hidroloacutegico Durante el verano la
mayor temperatura y horas-luz favoreceriacutean las entradas de N al lago viacutea fijacioacuten
bioloacutegica de N2 atmosfeacuterico resultando en un incremento de la MO con un valor
127
isotoacutepico bajo (POM verano~5permil Fig 1b) El N2 (δ15N = 0permil) es fijado praacutecticamente
sin fraccionamiento isotoacutepico generando un δ15N de la OM cercano a 0permil (Leng et
al 2006) En invierno las precipitaciones pueden estar favoreciendo un incremento
en la entrada de nutrientes al lago El DIN de la columna de agua llega a contener
valores de δ15N maacutes altos reflejando estos mayores aportes de la cuenca y el POM
del Lago Vichuqueacuten queda enriquecido en 15N (δ15N ~11permil Cap 2 Fig 9) Estas
variaciones estacionales pueden estar evidenciando un cambio en la composicioacuten
de especies de POM desde especies fijadoras a especies que consumen el N de la
columna de agua Sin embargo para corroborar esta informacioacuten seriacutea deseable
contar con el anaacutelisis de la composicioacuten de especies de las muestras de agua
extraidas
Entre los resultados maacutes importantes estaacuten los valores isotoacutepicos del suelo
y la biomasa representativa de la cuenca que incluye un listado de las especies
nativas de la zona que hasta ahora no se contaba (Cap 2 Tabla en material
suplementario) Las especies colectadas despliegan valores de δ15Nfoliar maacutes
positivos (plantaciones forestales y pastizal ~89permil) que el suelo (35permil) salvo por
las especies nativas las que oscilan en torno a valores cercanos al atmosfeacuterico
(~14permil) La razoacuten isotoacutepica 15N14N refleja la dinaacutemica de intercambio de N entre la
vegetacioacuten y el suelo (Koba et al 2002) La discriminacioacuten es positiva en la mayoriacutea
de los sistemas bioloacutegicos por lo que las plantas tienen valores de δ15N maacutes bajos
que el suelo (Evans et al 2001) como sucede con las especies nativas en Lago
Vichuqueacuten En consecuencia las plantas quedan empobrecidas en 15N y conducen
a un enriquecimiento en 15N del suelo sobretodo si no hay reposicioacuten de N por otras
viacuteas (eg fijacioacuten de N) Por lo tanto los valores maacutes negativos de δsup1⁵Nfoliar de las
especies nativas pueden estar relacionados con el consumo preferencial de 14N del
128
suelo donde la discriminacioacuten isotoacutepica de la vegetacioacuten contra el 15N conduce a
valores del δsup1⁵Nsuelo maacutes altos (Houmlgberg 1997) Por el contrario los valores maacutes
positivos de δsup1⁵Nfoliar de las plantaciones forestales y pastos con respecto al suelo
puede deberse por una parte que el suelo no cuenta con mecanismos naturales de
reposicioacuten de N salvo por la descomposicioacuten de la hojarasca que puede ser maacutes
lenta en Pinus radiata que en bosque nativo (Lusk et al 2001) y por otra por el alto
impacto de los aportes de N (y otros nutrientes) derivado de las actividades
humanas (eg uso de fertilizantes) en el suelo
El alcance maacutes significativo de esta tesis se relaciona con un cambio en la
tendencia maacutes negativa de δsup1⁵N y el incremento del NT a partir de 1940 y a partir
de 1970 CE en ambos lagos La causa maacutes probable de esta tendencia es el
reemplazo de la cobertura natural o preexistente a 1940 CE por plantaciones
forestales
En la figura 2 se observa una siacutentesis de los principales procesos que
afectan la sentildeal isotoacutepica de la MO en los sedimentos lacustres de L Matanzas y
L Vichuqueacuten La entrada de N y MO desde la cuenca depende del tipo de cobertura
Las plantaciones forestales y el bosque nativo retienen maacutes nutrientes y sedimentos
en la cuenca mientras que la agricultura los favorece Sin embargo las praacutecticas
de manejo que incluyen la aplicacioacuten de N (y otros nutrientes) pueden aportar maacutes
nutrientes al lago que la cobertra de bosque nativo Cuando las actividades
forestales cubren una mayor superficie de la cuenca (1940 en adelante) δsup1⁵N oscila
en valores maacutes negativos y el NT tiende a incrementarse en los sedimentos
lacustres Cabe destacar que los valores de δsup1⁵N observados en los testigos
sedimentarios a fines del s XX no tienen precedente durante la conquista y colonia
espantildeola o durante el resto del periodo de la Repuacuteblica
129
Figura 2 Modelo de transferencia de N entre los ecosistemas terrestres y
acuaacuteticos El δ15N de los sedimentos lacustres depende de la interaccioacuten entre los
aportes de la cuenca y porcesos ldquoin-lakerdquo Flechas indican la direccioacuten del flujo de
N En azul las salidasentradas de 14N en rojo las salidasentradas de 15N δ15N de
la interaccioacuten plantas-suelo depende del tipo de cobertura de suelo
130
CONCLUSIONES GENERALES
La transferencia de N entre cuencas y lagos es un factor de control del ciclo
del N en los lagos y queda reflejado en la sentildeal isotoacutepica de los sedimentos
lacustres (Leng et al 2006 McLauchlan et al 2007) En Laguna Matanzas el
suelo de las especies nativas y las plantaciones forestales despliegan valores de
δ15N maacutes bajos (~1permil) que los de Lago Vichuqueacuten (~35permil) Coincidentemente los
sedimentos lacustres de Laguna Matanzas oscilan con valores de δ15N maacutes bajos
(-15 a +45permil Fig 1) que los de Lago Vichuqueacuten (+3 a +8permil)
Por otra parte el estudio de la MO de los sedimentos lacustres ha permitido
reconstruir los cambios en los aportes de MO y N desde la cuenca Aunque no es
posible afirmar queacute tipo de N estaacute entrando al lago (eg Nitroacutegeno orgaacutenico e
inorganico) si podemos decir que los cambios en las fluctuaciones de δ15N son
coincentes con el crecimiento de la actividad forestal (1980 - hasta 2016 L
Matanzas 0-17 y L Vichuqueacuten 1-66) El δ15N fluctuacutea hacia valores maacutes
negativos Ademaacutes se observa un incremento del NT en los sedimentos lacustres
cuanto mayor es la superficie forestal
Entre 1800-1940 las cuencas eran fundamentalmente agriacutecolas y
ganaderas (Cap1 Fig 7 y Cap 2 Fig 12) el δ15N de los sedimentos lacustres
oscila con valores maacutes positivos (L Matanzas +40 plusmn 05permil y L Vichuqueacuten 56 plusmn
033permil Fig 1) hay una baja bioproductividad en el lago (BrTi ~002 TOC~17)
lo que coincide con un bajo contenido de NT (~478 μg) Este es un periacuteodo de altas
precipitaciones (Christie et al 2009) que tienden a favorecer el lavado de la cuenca
131
y con ello el aporte de MO sedimentos y nutrientes desde la cuenca tiende a verse
favorecido Aunque las principales actividades humanas en estas cuencas son
diferentes (ganaderiacutea en Laguna Matanzas Contreras-Lopez et al 2014
agricultura en Lago Vichuqueacuten Odone 1997) en ambos casos hubo un reemplazo
de la cobertura natural de bosques nativo que favorecioacute mayores aportes de N y
sedimentos desde la cuenca en un efecto sumado con el aumento de las
precipitaciones
A partir de 1980 CE en ambos lagos se observa una disminucioacuten en los
valores de δ15N y un incremento del NT que no tiene precedentes en todo el registro
y pese a que ambos lagos son limnologicamente muy diferentes En Lago
Vichuqueacuten el δ15N tiende a oscilar en antifase con el NT (Fig 1) En el caso de
Laguna Matanzas se observa una tendencia hacia valores maacutes negativos y a partir
de 1990s a valores maacutes positivos mientras que el NT tiende a incrementarse de
manera sostenida Ello podriacutea estar relacionado con el crecimiento de la actividad
forestal habriacutea una mayor retencioacuten de N y sedimentos en la cuenca ligado al
incremento de la superficie forestal Ademaacutes en el caso de L Matanzas el
incremento de la actividad agriacutecola (sumado al de las plantaciones forestales)
podriacutea expicar el cambio en la tendencia en la deacutecada de los 1990s
En el contexto de Antropoceno esta tesis nos permite identificar un gran
impacto de las actividades humanas en Chile central a partir de la deacutecada de 1940
y una Gran Aceleracoacuten a partir de la deacutecada de 1970s En el registro sedimentario
de los lagos en estudio se observa una gran acumulacioacuten de MO el δ15N oscila
hacia valores maacutes negativos y un gran incremento del NT y el crecimiento de la
actividad forestal parece ser la principal responsable Ademaacutes la sobreexplotacioacuten
132
del agua junto al cambio climaacutetico global ha generado una combinacioacuten letal para
los lagos costeros de Chile central
Figura 1 Comparacioacuten de las fluctuaciones de δ15N durante los uacuteltimos 300
antildeos en Laguna Matanzas y Lago Vichuqueacuten
133
Referencias
Battye W Aneja VP Schlesinger WH 2017 Earthrsquos Future Is nitrogen the next carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia
UC de V 2014 Elementos de la historia natural del An Mus Hist Natulas Vaplaraiso 27 51ndash67
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Evans RD Evans RD 2001 Physiological mechanisms influencing
plant nitrogen isotope composition 6 121ndash126 Fogel ML Cifuentes LA 1993 Isotope fractionation during primary
production Org Geochem 73ndash98 httpsdoiorg101007978-1-4615-2890-6_3 Galloway JN Schlesinger WH Ii HL Schnoor L Tg N 1995
Nitrogen fixation Anthropogenic enhancement-environmental response reactive N Global Biogeochem Cycles 9 235ndash252
Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J
Christie D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Koba K Koyama LA Kohzu A Nadelhoffer KJ 2003 Natural 15 N
Abundance of Plants Coniferous Forest httpsdoiorg101007s10021-002-0132-6
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos
Transactions American Geophysical Union httpsdoiorg1010292007EO070007 McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE
2007 Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
134
Phytologist N Oct N 2006 Tansley Review No 95 15N Natural Abundance in Soil-Plant Systems Peter Hogberg 137 179ndash203
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Turner BL Kasperson RE Matson PA McCarthy JJ Corell RW
Christensen L Eckley N Kasperson JX Luers A Martello ML Polsky C Pulsipher A Schiller A 2003 A framework for vulnerability analysis in sustainability science Proc Natl Acad Sci U S A httpsdoiorg101073pnas1231335100
Vitousek PM Aber JD Howarth RW Likens GE Matson PA
Schindler DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
Vitousek PM Howarth RW 1991 Nitrogen limitation on land and in the
sea How can it occur Biogeochemistry httpsdoiorg101007BF00002772 von Gunten L Grosjean M Rein B Urrutia R Appleby P 2009 A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 Holocene 19 873ndash881 httpsdoiorg1011770959683609336573
Xu R Tian H Pan S Dangal SRS Chen J Chang J Lu Y Maria
Skiba U Tubiello FN Zhang B 2019 Increased nitrogen enrichment and shifted patterns in the worldrsquos grassland 1860-2016 Earth Syst Sci Data httpsdoiorg105194essd-11-175-2019
Zaehle S 2013 Terrestrial nitrogen-carbon cycle interactions at the global
scale Philos Trans R Soc B Biol Sci 368 httpsdoiorg101098rstb20130125
7
Agradezco a los miembros del laboratorio de Paleoambientes Cuaternarios
(IPE) en especial a Raquel Elena Fernando Ana Penelope Graciela Miguel
Sevilla Mariacutea y Miguel Bartolomeacute
Agradezco a los miembros de la comisioacuten Dr Joseacute Miguel Farintildea Dr Juan
Armesto y Dr Ricardo De Pol-Holz por la gran ayuda durante el desarrollo de mi
doctorado en especial por las correcciones finales de la tesis
Al departamento de Ecologiacutea en especial a Rosario Galaz Edita Luengo
Daniela Mora y Valeria Cavallero por su apoyo
A mis compantildeeros de doctorado Beleacuten Gallardo Pancho Diacuteaz Bryan Bularz
Bian Latorre Renato Borras Juan Carlos Hernaacutendez y Paulina Troncoso con
quienes estudieacute aprendiacute y compartimos muy buenos momentos durante los
primeros antildeos del doctorado
A Pablo Sarricolea mi compantildeero de vida Gracias por tu constante e
incondicional apoyo Sin ti durante este proceso todo habriacutea sido cuesta arriba
A mis padres Arturo y Malena gracias por su carintildeo apoyo y compromiso
Papaacute aunque no esteacutes a mi lado siempre te recuerdo con mucho carintildeo A mi
madre que ha sido un constante apoyo sobre todo en el cuidado de mis hijos xavi
y panchito
A mis hermanos Rodrigo y David por estar presentes durante toda esta
etapa Siempre con carintildeo y hermandad
A Rodrigo David Matiacuteas Jany Paola y Julio por cuidar a mis hijos con carintildeo
siempre que estuve ausente por el doctorado
8
ABREVIATURAS
N Nitroacutegeno (Nitrogen)
DIN Nitroacutegeno Inorgaacutenico Disuelto (Dissolved Inorganic Nitrogen)
C Carbono (Carbon)
TOC Carbono Inorgaacutenico Total (Total Organic Carbon)
TIC Carbono Inorgaacutenico Total (Total inorganic Carbon)
TC Carbono Total (Total Carbon)
TS Azufre Total (Total Sulfur)
LUCC Cambios de UsoCobertura del Suelo (Land Use and Cover Change)
OM Materia Orgaacutenica (Organic Matter)
POM Particulate Organic Matter (materia orgaacutenica particulada)
CE Common Era
BCE Before Common Era
Cal BP Calibrado en antildeos radiocarbono antes de 1950
ie id est (esto es)
e g Exempli gratia (por ejemplo)
9
RESUMEN
El ldquoAntropocenordquo se caracteriza por pertubaciones de origen humano que
conllevan impactos globales Por ejemplo los cambios de usocobertura del suelo
(LUCC) son conocidos por perturbar el ciclo del N a partir de la Revolucioacuten Industrial
pero especialmente desde la Gran Aceleracioacuten (1950 CE) en adelante Sin
embargo existen incertezas asociadas a la magnitud del impacto y su efecto
acoplado con los cambios climaacuteticos actuales (ie megasequiacutea disminucioacuten de las
precipitaciones) y acoplamientos con otros ciclos biogeoquiacutemicos (eg ciclo del
Carbono) Este impacto ha cambiado la disponibilidad de N en los ecosistemas
terrestres y acuaacuteticos y sus enlaces Los sedimentos lacustres contienen
informacioacuten de las condiciones paleoambientales del lago y su cuenca en el
momento en que se depositaron y el anaacutelisis de los isoacutetopos estables de N (δ15N)
en los sedimentos permiten reconstruir los cambios en la disponibilidad de N a
traveacutes del tiempo En este trabajo usamos una aproximacioacuten multiproxy que incluye
10
anaacutelisis sedimentoloacutegicos geoquiacutemicos e isotoacutepicos (δ15N δ13C) de los sedimentos
lacustres columna de agua y suelo-vegetacioacuten de la cuenca y la reconstruccioacuten de
los LUCC entre 1975 y 2016 a partir de imaacutegenes satelitales El objetivo de esta
tesis fue evaluar el rol de LUCC como principal control del ciclo del N en el sistema
cuenca-lago durante las uacuteltimas centurias en Chile central Nuestros principales
resultados muestran que valores maacutes positivos de δ15N en los sedimentos lacustres
estaacuten relacionados con grandes aportes de N de la cuenca que a su vez son
mayores cuanto mayor la superficie agriacutecola yo los pastizales mientras que tanto
las plantaciones forestales como los bosques favorecen la retencioacuten de nutrientes
en la cuenca (lo que se ve reflejado en un δ15N maacutes negativo) Esta tesis plantea
un cambio de estado en la dinaacutemica del N asociado a la introduccioacuten y expansioacuten
de las plantaciones forestales o bosques los que retienen el Nitroacutegeno en las
cuencas mientras que el clima juega un rol secundario
11
ABSTRACT
The Anthropocene is characterized by human disturbances at the global
scale For example changes in land use are known to disturb the N cycle since the
industrial revolution but especially since the Great Acceleration (1950 CE) onwards
This impact has changed N availability in both terrestrial and aquatic ecosystems
However there are some important uncertainties associated with the extent of this
impact and how it is coupled to ongoing climate change (ie megadroughts rainfall
variability and shifting stationarity) or to other nutrient cycles (e g carbon cycle)
Lake sediments contain paleoenvironmental information regarding the conditions of
the watershed and associated lakes and which the respective sediments are
deposited Nitrogen stable isotope analysis (δ15N) on lake sediments allows us to
reconstruct the changes in N availability through time Here we used a multiproxy
approach that uses sedimentological geochemical and isotopic analyses on
lacustrine sediments water column and soilvegetation from the watershed as well
12
as land use and cover change (LUCC) from 1975 to 2016 obtained from satellite
images The goal of this thesis is to evaluate the role of LUCC as the main driver for
N cycling in a coastal watershed system of central Chile over the last centuries Our
main results show that more positive δ15N values in lake sediments are related to
higher N contributions from the watershed which in turn increase with increased
agricultural andor pasture cover whereas either forest plantations or native forests
can favor nutrient retention in the watershed (δ15N more negative) This thesis
proposes that N dynamics are mainly driven by the introduction and expansion of
forest or tree plantations that retain nitrogen in the watershed whereas climate plays
a secondary role
13
INTRODUCCIOacuteN
El N es un elemento esencial para la vida y limita la productividad en
ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al 1997) Las actividades
humanas han tenido un profundo impacto sobre el ciclo del N global principalmente
a partir la revolucioacuten industrial (IPCC 2013 2007) Las concentraciones de N se
han incrementado por la fijacioacuten industrial de N atmosfeacuterico (viacutea el proceso Haber-
Bosch Erisman et al 2008) por el uso de fertilizantes sobre la base de N para
mejorar el rendimiento de los cultivos la produccioacuten ganadera y ademaacutes de los
cambios en el uso del suelo (abreviado en ingleacutes- LUCC) (Robertson y Vitousek
2009 Galloway et al 2008 Battye et al 2017 Xu et al 2019) Estas actividades
contribuyen significativamente al incremento de la disponibilidad bioloacutegica de N
cuyas consecuencias para los ecosistemas incluye la perdida de diversidad
modificacioacuten de la disponibilidad de nutrientes y acidificacioacuten de los suelos entre
otras Sin embargo se desconoce la cantidad destino y el alcance que ha tenido
14
el N movilizado entre los ecosistemas generado por la influencia de las actividades
humanas (Vitousek et al 1997)
La entrada natural de N en los ecosistemas terrestres y acuaacuteticos es viacutea
fijacioacuten atmosfeacuterica la que es llevada a cabo por microorganismos muchos de ellos
en relaciones simbioacuteticas (eg Rhizobium en leguminosas) y las algas (Vitousek et
al 1997 Battye et al 2017) Las principales salidas estaacuten relacionadas con la
desnitrificacioacuten volatilizacioacuten del amoniaco y por escurrimiento superficial y
subterraacuteneo (Galloway et al 2008 Diacuteaz et al 2016) En el caso de los sistemas
lacustres ademaacutes estaacuten la resuspencioacuten de N del fondo del lago los aportes desde
la cuenca (entradas de N) La depositacioacuten de MO en el fondo del lago que marca
la salida de N de la columna de agua Estas relaciones de intercambio de N tienen
un control geograacutefico (eg latitud exposicioacuten relieve etc) climaacutetico y biotico
(McLauchlan et al 2013) La alteracioacuten provocada por los LUCC no solo genera
las peacuterdidas de nutrientes del suelo por erosioacuten y escurrimiento superficial sino que
tambieacuten modifica la disponibilidad y transferencia de N entre los ecosistemas
terrestres y acuaacuteticos (Elbert et al 2012 McLauchlan et al 2013) Por ejemplo el
reemplazo de la vegetacioacuten nativa por la agricultura yo plantaciones forestales
altera el ciclo del N debido al despeje de los terrenos viacutea quema de la vegetacioacuten
pero tambieacuten debido al reemplazo de la vegetacioacuten preexistente por un
monocultivo (eg trigo pino) y el uso ineficiente de fertilizantes para mejorar el
rendimiento agriacutecola Estas actividades favorecen un incremento de los aportes de
N (y otros nutrientes) viacutea sistemas de drenajes a lagos los que actuacutean como
sumideros El incremento del N derivado de las actividades humanas tanto en los
ecosistemas terrestres como acuaacuteticos genera incertezas relacionados con la
trayectoria y la magnitud de la disponibilidad de N en ambos ecosistemas (Batye et
15
al 2017 Mclauchlan et al 2017) Por lo tanto se requiere de una liacutenea base de
largo plazo para saber coacutemo estas perturbaciones impactan la disponibilidad de N
en las uacuteltimas deacutecadassiglos en los ecosistemas lacustres y por lo tanto el alcance
real que los LUCC han tenido en el ciclo del N
Los ecosistemas mediterraacuteneos y el ciclo del N
Los ecosistemas mediterraacuteneos son especialmente sensibles a los LUCC
pues estos se encuentran sometidos a un estreacutes hiacutedrico durante largas temporadas
estivales y las precipitaciones se concentran en eventos puntuales y a veces con
altos montos pluviomeacutetricos Las precipitaciones tienen un alto potencial de arrastre
de sedimentos y nutrientes los que actuacutean como fertilizantes al llegar a los
ecosistemas lacustres A su vez un incremento en la disponibilidad de N puede
generar un desbalance con otros nutrientes (eg Fosforo) con efectos sobre la
productividad y diversidad de los lagos (Pentildeuelas et al 2012 Sardans et al 2012
McLauchlan 2013) En este sentido en Chile central se observa una disminucioacuten
de las precipitaciones especialmente fuerte en las uacuteltimas deacutecadas por lo que se ha
denoacuteminado como Megasequiacutea (Garreaud et al 2017) y el efecto de las
precipitaciones esporaacutedicas pueden generar un arrastre de nutrientes que no ha
sido evaluado
Los ecosistemas mediterraacuteneos concentran el 20 de la biodiversidad global
(Myers et al 2000) pero existe una escasez de conocimiento respecto a los
efectos del incremento de N en los cuerpos de agua como consecuencia de las
actividades humanas en el pasado histoacuterico (Ochoa-Hueso et al 2011) y coacutemo la
disponibilidad de N ha variado en el tiempo Los LUCC han aumentado el flujo de
N en las cuencas hidrograacuteficas viacutea escurrimiento superficial y lixiviacioacuten
16
favoreciendo la peacuterdida de sedimentos y nutrientes del suelo en el mundo entero
(Elser et al 2011 McLauchlan et al 2013 Xu et al 2017 y 2019) Ello ha
contribuido a la acidificacioacuten y eutrofizacioacuten generalizada de riacuteos y lagos
(McLauchlan et al 2013 Schindler et al 2008)
El ecosistema mediterraacuteneo de Chile central es una regioacuten ampliamente
intervenida desde al menos la colonizacioacuten espantildeola (1600-1800 CE) Los LUCC
han tenido efectos negativos en la disponibilidad de agua especialmente
observados con el desarrollo de las actividades minera (Prieto et al 2019) Aunque
se sabe de los impactos en los ecosistemas de bosque en teacuterminos de cobertura
debidos al desarrollo silvo-agropecuarias (Armesto et al 2010) se desconoce el
impacto de estas actividades sobre el ciclo del N en los cuerpos de agua o en queacute
momento cobran fuerza suficiente para alterar el flujo de N hacia los lagos de Chile
Central Por lo tanto en esta tesis nos centramos en entender coacutemo los LUCC han
afectado la disponibilidad de N en los lagos costeros de Laguna Matanzas y Lago
Vichuqueacuten desde la llegada de los espantildeoles a Chile hasta el presente
Los lagos como sensores ambientales
Los sedimentos lacustres son buenos sensores de cambios en los aportes
de nutrientes contaminacioacuten y eutroficacioacuten de los cuerpos de agua ya que son
capaces de preservar sentildeales quiacutemicas y fiacutesicas al momento de su formacioacuten y
ademaacutes con alta resolucioacuten y a largo plazo (Leng et al 2006) Por lo tanto
constituyen una herramienta uacutetil para reconstruir el flujo de N entre ecosistemas
terrestres y acuaacuteticos en el pasado sus consecuencias (eg incremento de la
productividad lacustre) y el rol del clima y los LUCC en el pasado (McLauchlan et
al 2013 von Gunten et al 2009) Por ejemplo el cultivo de maiacutez por parte de los
17
nativos iroqueses en Canadaacute provocoacute la eutrofizacioacuten del Lago Crawford (43degN)
durante los siglos XII y XV Entre las consecuencias registradas se documentoacute un
claro incremento de la productividad primaria y cambios en la estructura comunitaria
de diatomeas foacutesiles (incremento de Stephanodiscus complex y disminucioacuten de
Cyclotella bodanica en Ekdahl et al 2007) Otro ejemplo demuestra como las
actividades agriacutecolas en la cuenca del riacuteo Mississippi modificaron los patrones de
sedimentacioacuten y aporte de nutrientes al lago Pepin EE UU (44degN) a partir del
asentamiento europeo en 1840 CE y principalmente desde 1940 CE (Engstrom et
al 2009) Para Chile von Gunten et al (2009) a partir de indicadores
limnogeoloacutegicos evidencioacute la contaminacioacuten y eutroficacioacuten de cinco lagos cercanos
a la ciudad de Santiago (33ordmS) como consecuencia de la deposicioacuten atmosfeacuterica
de metales pesados (especialmente Cu) quema de combustible foacutesil y flujo de
nutrientes durante los uacuteltimos 200 antildeos
Caracteriacutesticas limnoloacutegicas de los lagos
Los procesos limnoloacutegicos afectan la distribucioacuten y abundancia de los
organismos en los lagos Estaacuten influenciados por forzamientos externos por
ejemplo la precipitacioacuten o la penetracioacuten de la luz y la morfometriacutea del lago En este
sentido el nivel del lago dependeraacute del balance entre las entradas y salidas de agua
(precipitaciones escurrimiento superficial y subterraacuteneo y la evaporacioacuten) y la forma
de la cuenca (profundidad pendiente aacuterea del espejo de agua)
En funcioacuten de la profundidad de penetracioacuten de la luz es posible diferenciar
dos aacutereas que determinan la distribucioacuten de los organismos la zona foacutetica (donde
penetra la luz y permite la fotosiacutentesis) y la zona afoacutetica (subyacente a la zona
foacutetica) Al depender de la radiacioacuten la zona foacutetica variacutea estacionalmente Ademaacutes
18
puede variar por el grado de turbidez la morfometriacutea del lago y la produccioacuten de
materia orgaacutenica en la columna de agua
Otro factor que influye en la productividad es el reacutegimen de mezcla de la
columna de agua pues favorece la oxigenacioacuten y el reciclado de nutrientes La
mezcla ocurre en condiciones de gran inestabilidad usualmente relacionado con el
reacutegimen de viento Por el contrario un lago estratificado resulta de grandes
diferencias en temperatura o salinidad entre la superficie (epilimnion) y el fondo del
lago (hipolimnion) que separa las masas de agua superficial y de fondo por una
termoclina (si la estratificacioacuten estaacute relacionada con diferencias de temperatura de
las masas de agua) o haloclina (diferencias de salinidad) En funcioacuten del reacutegimen
de mezcla los lagos se pueden clasificar en (Lewis 1983)
1 Amiacutecticos no hay mezcla vertical de la columna de agua
2 Monomiacutecticos friacuteos y caacutelidos se mezcla una vez al antildeo
3 Dimiacutecticos la columna de agua se mezcla dos veces al antildeo
4 Oligomiacutecticos Lagos estratificados la mayor parte del tiempo pero se mezclan a
intervalos irregulares mayores a 1 antildeo
5 Polimiacutecticos friacuteos y caacutelidos la columna de agua se mezcla varias veces al antildeo
El ciclo del N en lagos
Al estar presente en aacutecidos nucleicos aminoaacutecidos y proteiacutenas el N es un
nutriente esencial de los organismos Por lo tanto su disponibilidad en la columna
de agua es clave para la produccioacuten composicioacuten y acumulacioacuten de la MO a traveacutes
del tiempo (Olson 1998 Talbot 2002) Aunque el principal reservorio de N estaacute en
19
la atmosfera (N2) solo unos pocos organismos (eg cianobacterias) pueden fijarlo
directamente Asiacute el Nitroacutegeno Inorgaacutenico Disuelto (DIN) representa la principal
fuente de N disponible para la produccioacuten primaria en los ecosistemas acuaacuteticos
(Talbot et al 2002) El DIN estaacute compuesto por nitrato (NO3-) nitrito (N02
-) y amonio
(NH4+) y los cambios en su disponibilidad condicionan el tipo de produccioacuten primaria
(eg fijadores vs asimiladores de N ver Scott y Grant 2013 Scott 2008)
La Figura 1 resume los principales componentes en lagos del ciclo del N y
sus interacciones La principal entrada natural de N es viacutea fijacioacuten de N atmosfeacuterico
y es llevado a cabo principalmente por bacterias y algas verde-azules capaces de
romper el triple enlace entre los dos aacutetomos de N a un alto costo energeacutetico (Torres
et al 2012) El N fijado es reducido a NH3 Tras la muerte de los organismos el N
es liberado de sus tejidos por hongos y bacterias implicados en la descomposicioacuten
de la MO Una parte se deposita en el fondo del lago y la otra queda disponible para
ser asimilada por el fitoplancton como amonio mediante el proceso de
amonificacioacuten (Talbot 2002) La amonificacioacuten resulta de la reduccioacuten bacteriana
del amoniaco y se da preferentemente en condiciones anoacutexicas La volatizacioacuten del
amoniaco es un proceso por medio este se incorpora a la atmoacutesfera Este proceso
se da preferentemente en lagos aacutecidos (pH gt 85) El nitrato es otra forma de N
bioloacutegicamente disponible para el fitoplancton En los lagos el NO3 del DIN estaacute
compuesto por los aportes desde la cuenca hidrograacutefica en la que se circunscriben
por la resuspensioacuten desde el fondo del lago favorecido por procesos de mezcla
(descrito en seccioacuten anterior) y por los nitratos de la columna de agua Estos uacuteltimos
son el resultado de la nitrificacioacuten proceso realizado por bacterias aeroacutebicas
mediante el cual el amonio se oxida para formar nitrito y nitrato El ciclo se completa
20
con la desnitrificacioacuten que es la reduccioacuten bacteriana de nitrato al N atmosfeacuterico
Este proceso se da preferentemente en condiciones anoacutexicas
Figura 1 Materia Orgaacutenica sedimentaria y su relacioacuten con el ciclo del N y las
variaciones de δ15N (elaborado a partir de Talbot et al 2002) En la figura se
representan los factores clave en la acumulacioacuten de la MO sedimentaria y su
relacioacuten con el ciclo del N El δ15N de la MO es resultado de los aportes de MO
desde la cuenca MO de macroacutefitas y de la productividad bioloacutegica La productividad
en ecosistemas lacustres estaacute condicionada por DIN y la fijacioacuten de N atmosfeacuterico
El δ15N del pool del DIN disponible se va enriqueciendo por la asimilacioacuten
preferencial de 14N por parte del fitoplancton Ademaacutes el DIN remanente se va
enriqueciendo por accioacuten microbiana (amonificacioacuten nitrificacioacuten desnitrificacioacuten)
Reconstruyendo el ciclo del N a partir de variaciones en δ15N
La sentildeal isotoacutepica de N (δ15N) en los sedimentos lacustres puede ser usada
para reconstruir los cambios pasados del ciclo N la transferencia de N entre
ecosistemas terrestres y acuaacuteticos y el estado troacutefico del lago (Bruland y Mackenzie
2015 McLauchlan et al 2013 Woodward et al 2012 von Gunten et al 2009
Leng et al 2006 Meyers y Teranes 2001) La Figura 1 resume los principales
procesos y fuentes que afectan la sentildeal isotoacutepica δ15N (total o ldquobulkrdquo) en la MO de
21
los sedimentos lacustres Entre los que destacan las fuentes de MO (aloacutectona vs
autoacutectona) la bioproductividad lacustre y las entradas de N (ie fijacioacuten bioloacutegica
de N2) (Torres et al 2012) Estos procesos conducen a un fraccionamiento
isotoacutepico cineacutetico que favorece la disociacioacuten entre sus isotopos ldquopesadosrdquo (15N) y
ldquoligerosrdquo (14N) Asiacute por ejemplo los valores de δ15N de la MO se enriquecen en 15N
en la medida que los sedimentos lacustres estaacuten sometidos a perdidas de 14N viacutea
desnitrificacioacuten (Bruland y Mackenzie 2015 Diaz et al 2016 Talbot et al 2002)
Por otra parte el fitoplancton en la medida que aumenta su biomasa (eg
durante la estacioacuten caacutelida) puede llegar a asimilar el DIN hasta agotarse En este
caso el N remanente se va empobreciendo en 14N si no hay reposicioacuten de N (eg
aporte de NO3 de la cuenca) y la MO queda enriquecida en 15N Esta disminucioacuten
induce a una limitacioacuten de fitoplancton por N y los organismos diztroacuteficos pueden
verse estimulados por lo que se incrementa la entrada de N viacutea fijacioacuten de N2 (Scott
y Grantsz 2013) como resultado la MO oscila en valores maacutes bajos (en torno a 0permil)
La cantidad de MO que se deposita en el fondo del lago depende del
predominio de contribuciones provenientes desde la cuenca (aloacutectona) versus las
producidas dentro del mismo lago (autoacutectona) (Torres et al 2012) Aunque en
general los lagos reciben permanentemente aportes de sedimentos y MO desde
su cuenca los lagos pequentildeos tienden a reflejar maacutes los procesos que ocurren
solamente en su cuenca directa y reflejan los valores isotoacutepicos de la misma (Gu et
al 2006) Woodward et al (2012) realizaron un estudio en 50 lagos en Irlanda que
les permitioacute correlacionar diferentes patrones de LUCC (bosques de coniacuteferas
agricultura ganaderiacutea entre otros) con los valores de δ15N (y δ13C) en los
sedimentos superficiales de los lagos Encontraron que los valores de δ15N son maacutes
negativos en cuencas poco impactadas y maacutes positivos en aquellas con alto
22
impacto humano (eg usos de suelo agriacutecola y ganadero) McLauchlan et al (2007)
encontraron valores maacutes positivos de δ15N en los sedimentos del Mirror Lake New
Hampshire (29ordm N) a partir de la desforestacioacuten de su cuenca en 1790 CE (inicio
del asentamiento europeo en el lago) para el desarrollo de la actividad agriacutecola
Estos valores se volvieron maacutes negativos hacia valores similares al pre-
asentamiento europeo despueacutes del cese de la agricultura en la cuenca y la
recuperacioacuten del bosque a partir de 1929 CE
El anaacutelisis de δ15N en los sedimentos lacustres es una herramienta casi sin
explorar en Chile central Proponemos reconstruir la dinaacutemica de transferencia de
N cuenca-lago como consecuencia de los LUCC desde la colonizacioacuten espantildeola en
los lagos costeros de Laguna Matanzas (34ordmS) y Lago Vichuqueacuten (36ordmS) Como son
muacuteltiples los procesos que pueden afectar la sentildeal isotoacutepica de N (medido como
δ15N) en los sedimentos lacustres existen muchos problemas para su
interpretacioacuten paleoambiental (Leng et al 2006) Por ello en esta tesis utilizamos
un enfoque multiproxy que consiste en integrar el anaacutelisis limnoloacutegico y geoquiacutemico
de los sedimentos lacustres con los valores isotoacutepicos modernos de la columna de
agua (mediante monitoreo del POM) la relacioacuten vegetacioacuten-suelo de la cuenca y la
reconstruccioacuten de los principales usos de suelo de las cuencas desde 1975 CE
mediante el uso de imaacutegenes satelitales Considerando estas muacuteltiples liacuteneas de
evidencia nos planteamos utilizar las variaciones del δ15N para reconstruir los
cambios en la disponibilidad de N en los lagos costeros de Chile central y establecer
coacutemo y desde cuaacutendo las actividades humanas en la cuenca son lo suficientemente
importantes como para modificar el ciclo del N en los lagos desde la colonizacioacuten
espantildeola (siglo XVII)
23
Como hipoacutetesis planteamos que la dinaacutemica de transferencia de sedimentos
y nutrientes entre la cuenca y el lago tiene controles geoloacutegicos climaacuteticos y
bioloacutegicos que interactuacutean en el tiempo registraacutendose en la acumulacioacuten de MO de
los sedimentos lacustres (Fig1) Bajo este marco los LUCC modifican esta
dinaacutemica alterando la transferencia de N a los lagos ya que las actividades agriacutecolas
y ganaderas tienden a aportar maacutes N (aumentando δ15N) que la actividad forestal
(la que disminuye δ15N)
En el primer capiacutetulo titulado ldquoA combined approach to establishing the timing
and magnitude of anthropogenic nutrient alteration in a mediterranean coastal lake-
watershed systemrdquo se evaluoacute el rol de los LUCC en el control de la descarga de N
y sedimentos a Laguna Matanzas en un sistema cuenca-lago desde el siglo XVII
Para ello se realizoacute un anaacutelisis de la dinaacutemica sedimentaria lacustre mediante el
anaacutelisis de muacuteltiples proxys sedimentoloacutegicos (ie minerales y textura)
geoquiacutemicos (eg geoquiacutemica orgaacutenica e isotoacutepica) y bioloacutegicos (ie diatomeas) de
Laguna Matanzas que cubren los uacuteltimos 200 antildeos Ademaacutes se realizoacute una
reconstruccioacuten de los usos de suelo entre 1975 y 2016 a partir de imaacutegenes de
sateacutelites y se colectaron muestras de suelo de las principales coberturas de la
cuenca a los cuales se midioacute el δ15N
Entre los principales resultados obtenidos se destaca la influencia de la
ganaderiacutea y la denitrificacioacuten lo que explica los altos valores de δ15N evidenciados
por el registro lacustre durante la colonizacioacuten espantildeola e inicios de la repuacuteblica A
partir de la Gran Aceleracioacuten (1950 CE) la desforestacioacuten y el reemplazo de la
ganaderiacutea por plantaciones forestales tienen un correlato en el registro
sedimentario con valores de δ15N maacutes bajos Estos resultados demuestran que los
LUCC son el factor de primer orden para explicar los cambios observados en
24
nuestro registro de δ15N y a su vez implica un rol secundario del clima como posible
control de la disponibilidad de N en Laguna Matanzas Este capiacutetulo fue sometido
a la revista Scientific Reports y estaacute en revisioacuten desde el 26 de marzo de 2019 En
la elaboracioacuten del mauscrito han colaborado Claudio Latorre Blas Valero-Garceacutes
Matiacuteas Frugone Pablo Sarricolea Santiago Giralt Manuel Contreras-Lopez
Ricardo Prego y Patricia Bernardez
El segundo capiacutetulo se titula ldquoStable isotopes track land use and cover
changes in a mediterranean lake in central Chile over the last 600 yearsrdquo Aquiacute
evaluamos la dinaacutemica del N actual en el sistema cuenca-lago considerando los
valores isotoacutepicos modernos tanto de la vegetacioacuten como del suelo ademaacutes de los
cambios estacionales del POM de la columna de agua Esta informacioacuten se utiliza
como un anaacutelogo moderno para conocer el rol de los LUCC en la disponibilidad de
N en los uacuteltimos 600 antildeos Para ello se colectaron muestras filtradas de la columna
de agua a 2 5 y 20 m dos veces por antildeo entre noviembre de 2015 y julio de 2018
y se obtuvo el δsup1⁵N del POM Ademaacutes se colectoacute muestras de vegetacioacuten y suelo
de la cuenca diferenciando entre especies nativas plantaciones forestales y
vegetacioacuten herbaacutecea Esta informacioacuten se analizoacute en conjunto con la reconstruccioacuten
de los LUCC entre 1975 y 2014 y el anaacutelisis limnoloacutegico del lago Ello nos permitioacute
evaluar coacutemo el δ15N de los sedimentos lacustres refleja los valores δ15N de la
cuenca en el Lago Vichuqueacuten y el control que ejerce el clima en la sentildeal isotoacutepica
de POM Este capiacutetulo seraacute sometido a la revista Science of total environmet
proximamente En la elaboracioacuten del mauscrito han colaborado Claudio Latorre
Blas Valero-Garceacutes Matiacuteas Frugone Pablo Sarricolea Leonardo Villacis y M Laura
Carrevedo
25
Entre los principales resultados encontramos que el δ15N en los sedimentos
lacustres es maacutes positivo con presencia de agricultura en la cuenca y maacutes negativo
cuando esta es reemplazada por cobertura arboacuterea bien sea plantaciones
forestales bien sea bosque nativo A su vez durante la estacioacuten seca y caacutelida la
mayor entrada al lago es viacutea fijacioacuten de N atmosfeacuterico (bajos valores de δ15NPOM)
Durante la estacioacuten huacutemeda en cambio prevalecen los aportes de la cuenca con
altos valores de δ15NPOM Ello estariacutea evidenciando un cambio estacional en la
composicioacuten de especies en verano ligado a especies fijadoras de N y en invierno
las algas y microorganismos que consumen el DIN de la columna de agua
Referencias
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea AM Marquet PA 2010 From the Holocene to the Anthropocene A historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the
next carbon Earth rsquo s Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409 httpsdoiorg102134jeq20090005
Diacuteaz FP Frugone M Gutieacuterrez RA Latorre C 2016 Nitrogen cycling in an
extreme hyperarid environment inferred from δ15 N analyses of plants soils and herbivore diet Sci Rep 6 1ndash11 httpsdoiorg101038srep22226
Ekdahl EJ Teranes JL Wittkop CA Stoermer EF Reavie ED Smol JP
2007 Diatom assemblage response to Iroquoian and Euro-Canadian eutrophication of Crawford Lake Ontario Canada J Paleolimnol 37 233ndash246 httpsdoiorg101007s10933-006-9016-7
Elbert W Weber B Burrows S Steinkamp J Buumldel B Andreae MO
Poumlschl U 2012 Contribution of cryptogamic covers to the global cycles of carbon and nitrogen Nat Geosci 5 459ndash462
26
httpsdoiorg101038ngeo1486 Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506
httpsdoiorg101126science1215567 ARTICLE Engstrom DR Almendinger JE Wolin JA 2009 Historical changes in
sediment and phosphorus loading to the upper Mississippi River Mass-balance reconstructions from the sediments of Lake Pepin J Paleolimnol 41 563ndash588 httpsdoiorg101007s10933-008-9292-5
Erisman J W Sutton M A Galloway J Klimont Z amp Winiwarter W (2008)
How a century of ammonia synthesis changed the world Nature Geoscience 1(10) 636
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892
httpsdoiorg101126science1136674 Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J Christie
D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592 Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
IPCC 2013 Climate Change 2013 The Physical Science BasisContribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change Cambridge University Press Cambridge United Kingdom and New York NY USA
httpsdoiorg101017CBO9781107415324 Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007 Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32iumliquestfrac12S 3050iumliquestfrac12miumliquestfrac12asl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470
27
httpsdoiorg101073pnas0701779104 McLauchlan KK Gerhart LM Battles JJ Craine JM Elmore AJ Higuera
PE Mack MC McNeil BE Nelson DM Pederson N Perakis SS 2017 Centennial-scale reductions in nitrogen availability in temperate forests of the United States Sci Rep 7 1ndash7 httpsdoiorg101038s41598-017-08170-z
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Myers N Mittermeler RA Mittermeler CG Da Fonseca GAB Kent J
2000 Biodiversity hotspots for conservation priorities Nature httpsdoiorg10103835002501
Pentildeuelas J Poulter B Sardans J Ciais P Van Der Velde M Bopp L
Boucher O Godderis Y Hinsinger P Llusia J Nardin E Vicca S Obersteiner M Janssens IA 2013 Human-induced nitrogen-phosphorus imbalances alter natural and managed ecosystems across the globe Nat Commun 4 httpsdoiorg101038ncomms3934
Prieto M Salazar D Valenzuela MJ 2019 The dispossession of the San
Pedro de Inacaliri river Political Ecology extractivism and archaeology Extr Ind Soc httpsdoiorg101016jexis201902004
Robertson GP Vitousek P 2010 Nitrogen in Agriculture Balancing the Cost of
an Essential Resource Ssrn 34 97ndash125 httpsdoiorg101146annurevenviron032108105046
Sardans J Rivas-Ubach A Pentildeuelas J 2012 The CNP stoichiometry of
organisms and ecosystems in a changing world A review and perspectives Perspect Plant Ecol Evol Syst 14 33ndash47 httpsdoiorg101016jppees201108002
Schindler DW Howarth RW Vitousek PM Tilman DG Schlesinger WH
Matson PA Likens GE Aber JD 2007 Human Alteration of the Global Nitrogen Cycle Sources and Consequences Ecol Appl 7 737ndash750 httpsdoiorg1018901051-0761(1997)007[0737haotgn]20co2
Scott E and EM Grantzs 2013 N2 fixation exceeds internal nitrogen loading as
a phytoplankton nutrient source in perpetually nitrogen-limited reservoirs Freshwater Science 2013 32(3)849ndash861 10189912-1901
28
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Talbot M R (2002) Nitrogen isotopes in palaeolimnology In Tracking
environmental change using lake sediments (pp 401-439) Springer Dordrecht
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable
isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K
Yeager KM 2016 Different responses of sedimentary δ15N to climatic changes and anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
29
CAPIacuteTULO 1 A COMBINED APPROACH TO ESTABLISHING THE TIMING
AND MAGNITUDE OF ANTHROPOGENIC NUTRIENT ALTERATION IN A
MEDITERRANEAN COASTAL LAKE- WATERSHED SYSTEM
30
A combined approach to establishing the timing and magnitude of anthropogenic
nutrient alteration in a mediterranean coastal lake- watershed system
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugoneabc Pablo
Sarricolead Santiago Giralte Manuel Contreras-Lopezf Ricardo Prego g Patricia
Bernaacuterdez g Blas Valero-Garceacutesch
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Institute of Earth Sciences Jaume Almera (ICTJA-CSIC) CLluis Solegrave Sabaris sn Barcelona E-
08028 Spain
f Facultad de Ingenieriacutea y Centro de Estudios Avanzados Universidad de Playa Ancha Traslavintildea
450 Vintildea del Mar Chile
g Instituto de Investigaciones Marinas (CSIC) CEduardo Cabello 6 36208 Vigo Spain
h Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding author
E-mail address
clatorrebiopuccl magdalenafuentealbagmailcom
Abstract
Since the industrial revolution and especially during the Great Acceleration (1950
CE) human activities have profoundly altered the global nutrient cycle through land
use and cover changes (LUCC) However the timing and intensity of recent N
variability together with the extent of its impact in terrestrial and aquatic ecosystems
and coupled effects of regional LUCC and climate are not well understood Here
we used a multiproxy approach (sedimentological geochemical and isotopic
31
analyses historical records climate data and satellite images) to evaluate the role
of LUCC as the main control for N cycling in a coastal watershed system of central
Chile during the last few centuries The largest changes in N dynamics occurred in
the mid-1970s associated with the replacement of native forests and grasslands for
livestock grazing by plantations of introduced Monterey pine (Pinus radiata) and
eucalyptus (Eucalyptus globulus) Lake productivity increased and is evidenced by
an increase trend in δ15N values Our study shows that anthropogenic land
usecover changes are key in controlling nutrient supply and N availability in
Mediterranean watershed ndash lake systems and that large-scale forestry
developments during the mid-1970s likely caused the largest changes in central
Chile
Keywords Anthropocene Organic geochemistry watershedndashlake system Stable
Isotope Analyses Land usecover change Nitrogen cycle Mediterranean
ecosystems central Chile
1 INTRODUCTION
Human activities have become the most important driver of the nutrient cycles in
terrestrial and aquatic ecosystems since the industrial revolution (Gruber and
Galloway 2008 Galloway et al 2008 Holtgrieve et al 2011 Fowler et al 2013
Goyette et al 2016) Among these N is a common nutrient that limits productivity
in terrestrial and aquatic ecosystems (Vitousek and Howarth 1991 McLauchlan et
al 2013) With the advent of the Haber-Bosch industrial N fixation process in the
early 20th century total N fluxes have surpassed previous planetary boundaries
32
(Galloway et al 2008 Elser 2011) reaching unprecedented values (ie tipping
points) in the Earth system especially during what is now termed the Great
Acceleration (which began in the 1950s) which intensified in the 1970s (Howarth
2004 Steffen et al 2015) Yet dramatic changes in LUCC have occurred in the last
few centuries mainly linked to forestry and agro-pastoral activities (Poraj-Goacuterska et
al 2017 Ge et al 2019 Cheng et al 2019) These have had large impacts on N
(and Carbon -C-) availability in terrestrial and aquatic ecosystems but the synergic
effect with climate change and global N dynamics has not been established
(Vitousek et al 1997 Mclauchlan et al 2007 2013 Pongratz et al 2010
Woodward et al 2012 Mclauchlan et al 2017)
The onset of the Anthropocene poses significant challenges in mediterranean
regions that have a strong seasonality of hydrological regimes and an annual water
deficit (Stocker et al 2013) Mediterranean climates occur in all continents
(California central Chile Australia South Africa circum-Mediterranean regions)
providing a unique opportunity to investigate global change processes during the
Anthropocene in similar climate settings but with variable geographic and cultural
contexts The effects of global change in mediterranean watersheds have been
analyzed from different perspectives hydrology (Giorgi 2006 Arnell and Gosling
2013 Prudhomme et al 2014) vegetation dynamics (Lenihan et al 2003 Vicente-
Serrano et al 2013 Matesanz and Valladares 2014) sediment dynamics (Garciacutea-
Ruiz et al 2013 Garciacutea-Ruiz 2010 Syvitski et al 2005) changes in
biogeochemical cycles (Thomas et al 2013 Fowler et al 2013 Frank et al 2015)
carbon storage (Muntildeoz-Rojas et al 2015) and biodiversity (Hooper et al 2012) A
recent review showed an extraordinarily high variability of erosion rates in
mediterranean watersheds positive relationships with slope and annual
33
precipitation and the paramount effect of land use (Garciacutea-Ruiz et al 2015)
However the temporal context and effect of LUCC on nutrient supply to
mediterranean lakes has not been analyzed in much detail
Major LUCC in central Chile occurred during the Spanish Colonial period
(17th-18th centuries) (Armesto 1994 Gurnell et al 2001 Figueroa et al 2004
Martiacuten-Foreacutes et al 2012 Contreras-Loacutepez et al 2014) with the onset of
industrialization and mostly during the mid to late 20th century (von Gunten et al
2009 Gayo et al 2012) Recent copper pollution caused by 20th century mining
and industrial smelters has been documented in cores throughout the Andes
(Laguna el Ocho and Laguna Ensuentildeo von Gunten et al 2009) and also from our
surveys in the central valley (Batuco wetland) coastal range (Cordillera de Name)
and along the coast itself (Bucalemito and Colejuda) (Valero-Garceacutes et al 2010
unpublished data)
Paleolimnological studies have shown how these systems respond to
climate LUCC and anthropogenic impacts during the last millennia (Jenny et al
2002 2003 Villa Martiacutenez et al 2002 Frugone-Aacutelvarez et al 2017 Contreras et
al 2018) Furthermore changes in sediment and nutrient cycles have also been
identified in associated terrestrial ecosystems dating as far back as the Spanish
Conquest and related to fire clearance and wood extraction practices of the native
forests (Armesto 1994 Contreras-Loacutepez et al 2014) Nevertheless pollen and
limnological evidence argue for a more recent timing of the largest anthropogenic
impacts in central Chile For example paleo records show that during the mid-20th
century increased soil erosion followed replacement of native forest by Pinus
radiata and Eucalyptus globulus plantations at Laguna Matanzas Aculeo and
34
Vichuqueacuten lakes (Jenny et al 2002 Villa-Martiacutenez et al 2002 2003 Frugone-
Aacutelvarez et al 2017)
Lakes are a central component of the global carbon cycle Lakes act as a
sink of the carbon cycle both by mineralizing terrestrially derived organic matter and
by storing substantial amounts of organic carbon (OC) in their sediments (Anderson
et al 2009) Paleolimnological studies have shown a large increase in OC burial
rates during the last century (Heathcote et al 2015) however the rates and
controls on OC burial by lakes remain uncertain as do the possible effects of future
global change and the coupled effect with the N cycle LUCC intensification of
agriculture and associated nutrient loading together with atmospheric N-deposition
are expected to enhance OC sequestration by lakes Climate change has been
mainly responsible for the increased algal productivity since the end of the 19th
century and during the late 20th century in lakes from both the northern (Ruumlhland et
al2015) and southern hemispheres (Michelutti et al 2015 Carrevedo et al 2015)
but many studies suggest a complex interaction of global warming and
anthropogenic influences and it remains to be proven if climate is indeed the only
factor controlling these transitions (Catalaacuten et al 2013) Alternative causes for
recent N (Galloway et al 2008) increases in high altitude lakes such as catchment
mediated processes cannot be ruled out (Camarero and Catalan 2012 Fritz and
Anderson 2013) Few lake-watershed systems have robust enough chronologies of
recent changes to compare variations in C and N with regional and local processes
and even fewer of these are from the southern hemisphere (McLauchlan et al
2007 Holtgrieve et al 2011)
In this paper we present a multiproxy lake-watershed study including N and
C stable isotope analyses on a series of short cores from Laguna Matanzas in
35
central Chile focused in the last 200 years We complemented our record with land
use surveys satellite and aerial photograph studies Our major objectives are 1) to
reconstruct the dynamics among climate human activities and changes in the N
cycle over the last two centuries 2) to evaluate how human activities have altered
the N cycle during the Great Acceleration (since the mid-20th century)
2 STUDY SITE
Laguna Matanzas (33ordm45rsquoS 71ordm 40acuteW 7 m asl Fig 1) is a coastal lake located
in central Chile near to a large populated area (Santiago gt6106 inhabitants) The
lake has a surface area of 15 km2 with a max depth of 3 m and a watershed of 30
km2 The lake basin is emplaced over Pleistocene-Holocene aeolian and alluvial fan
deposits (SERNAGEOMIN 2003) A recent phase of dune activity occurred from the
mid to late Holocene which mostly sealed off the basin from the ocean (Villa-
Martinez 2002) Climate in Laguna Matanzas is characterized by cool-wet winters
and hot-dry summers with annual precipitation of ~510 mm and a mean annual
temperature of 12ordmC Central Chile is in the transition zone between the southern
hemisphere mid-latitude westerlies belt and the South Pacific Anticyclone (or SPA)
(Garreaud et al 2009) In winter precipitation is modulated by the north-west
displacement of the SPA the northward shift of the westerlies wind belt and an
increased frequency of storm fronts stemming off the Southern Hemisphere
Westerly Winds (SWW) (Frugone-Aacutelvarez et al 2017) Austral summers are
typically dry and warm as a strong SPA blocks the northward migration of storm
tracks stemming off the SWW
36
Historic land cover changes started after the Spanish conquest with a Jesuit
settlement in 1627 CE near El Convento village and the development of a livestock
ranch that included the Matanzas watershed After the Jesuits were expelled from
South America in 1778 CE the farm was bought by Pedro Balmaceda and had more
than 40000 head of cattle around 1800 CE (Contreras-Lopez et al 2014) The first
Pinus radiata and Eucalyptus globulus trees were planted during the second half of
the 19th century and were mostly used for dune stabilization (Albert 1900 Gibson
1972) However the main plantation phase occurred 60 years ago (Villa-Martinez
2002) as a response to the application of Chilean Forestry Laws promulgated in
1931 and 1974 and associated state subsidies
Major land cover changes occurred recently from 1975 to 2008 as shrublands
were replaced by more intensive land uses practices such as farmland and tree
plantations (Schulz et al 2010) Laguna Matanzas is part of the Reserva Nacional
Humedal El Yali protected area (Ramsar Ndeg 878) Despite this protected status the
lake and its watershed have been heavily affected by intense agricultural and
farming activities during the last decades The main inlet ldquoLas Rosasrdquo has been
diverted for crop irrigation causing a significant loss of water input to the lake
Consequently the flooded area of the lake has greatly decreased in the last couple
of decades (Fig 1b) Exotic tree species cover a large surface area of the
watershed Recently other activities such as farms for intensive chicken production
have been emplaced in the watershed
37
Figure 1 Laguna Matanzas a) Digital Elevation Model showing the watershed and
the surface hydrologic connection with Colejuda and Cabildo lakes b) Climograph
depicting the warm dry season in austral summer c) Annual precipitation from
1965-2015 (arrow shows start of the most recent ldquomega-droughtrdquo- see Garreaud et
al 2017) d) A decline of lake area occurs between 2007 and 2018 Lake surface
area decreased first along the western sector (in 2007) followed by more inland
areas (in 2018)
38
3 RESULTS
31 Age Model
The age model for the Matanzas sequence was developed using Bacon software
to establish the deposition rates and age uncertainty (Blaauw and Christen 2011)
It is based on 210Pb and 137Cs dates and two 14C dates (Table 2) According to this
age model the lake sequence spans the last 1000 years (Fig 2) A major breccia
layer (unit 3b) was deposited during the early 18th century which agrees with
historic documents indicating that a tsunami impacted Laguna Matanzas and its
watershed in 1730 CE (Contreras-Loacutepez et al 2014) Here we focus in the last 200
years were the most important changes occurred in terms of LUCC (after the
sedimentary hiatus caused by the tsunami) The Spanish colonial period (17th -18th
century) brought new forms of territorial management along with an intensification
of watershed use which remained relatively unchanged until the 1900s
39
Figure 2 a) Age-depth model obtained for the Laguna de Matanzas sedimentary
sequence A hiatus occurs at 80 cm depth (unit 3b) The section of core used for our
analysis is highlighted in a red rectangle b) Close up of the age model used for
analysis of recent anthropogenic influences on the N cycle c) Information regarding
the 14C dates used to construct age model
Lab code Sample ID
Depth (cm) Material Fraction of modern C
Radiocarbon age
Pmc Error BP Error
D-AMS 021579
MAT11-6A 104-105 Bulk Sediment
8843 041 988 37
D-AMS 001132
MAT11-6A 1345-1355
Bulk Sediment
8482 024 1268 21
POZ-57285
MAT13-12 DIC Water column 10454 035 Modern
Table 2 Laguna Matanzas radiocarbon dates
32 The sediment sequence
Laguna Matanzas sediments consist of massive to banded mud with some silt
intercalations They are composed of silicate minerals (plagioclase quartz and clay
minerals) with relatively high TOC content (Fig 3) Pyrite is a common mineral
indicating dominant anoxic conditions in the lake sediments whereas aragonite
occurs only in the uppermost section Mineralogical analyses visual descriptions
texture and geochemical composition were used to characterize five main facies
(Fig 3) F1 (organic-rich mud) represents baseline sedimentation in a shallow well-
mixed brackish highly productive lake and F1acute (dark orange) is a less organic facies
than F1 (more details see table in the supplementary material) F2 (massive to
banded silty mud) indicates periods of higher clastic input into the lake but finer
(mostly clay minerals) likely from suspension deposition associated with flooding
40
events Aragonite (up to 15 ) occurs in both facies but only in samples from the
uppermost 30 cm and it is interpreted as endogenic linked to higher MgFe waters
and elevated biologic productivity
Figure 3 Sedimentary facies and units mineralogy grain size elemental geochemical
and isotopic composition of Laguna Matanzas core Values highlighted in gray indicate
that these are above average
The banded to laminated fining upward silty clay layers (F3) reflect
deposition by high energy turbidity currents The presence of aragonite suggests
that littoral sediments were incorporated by these currents Non-graded laminated
coarse silt layers (F4) do not have aragonite indicating a dominant watershed
41
sediment source Both facies are interpreted as more energetic flood deposits but
with different sediment sources A unique breccia layer with coarse silt matrix and
cm-long soft clasts (F5) is interpreted as a high-energy event (ie a tsunami)
capable of eroding the littoral zone and depositing coarse clastic material in the
distal zone of the lake Similar coarse breccia layers have been found at several
coastal sites in Chile and interpreted as tsunami-related deposits (Vargas et al
2005 Le Roux et al 2008)
33 Sedimentary units
Three main units and six subunits have been defined (Fig 3) based on
sedimentary facies and sediment composition We use ZrTi as an indicator of the
mineral fraction transported from the watershed (Marzecoacuteva et al 2011) as higher
ZrTi (F3 and F4) is commonly associated with coarser sediments (Cuven et al
2010) A high correlation among Br BrTi and TOC (r = 046-087 p value=0)
supports the use of BrTi as an indicator of lake productivity (Marzecoacuteva et al 2011
Frugone-Aacutelvarez et al 2017) The MnFe ratio is indicative of lake bottom
oxygenation (Naecher et al 2013) as under reducing conditions Mn mobilizes more
than Fe leading to a decreased MnFe ratio (Marzecoacuteva et al 2011) SrTi indicates
periods of increased aragonite formation as Sr is preferentially included in the
aragonite mineral structure (Veizer et al 1971) (See supplementary material)
The basal unit (3c 129 to 99 cm) is relatively organic-rich (TOC mean=26
BioSi mean=5) and composed by F1 (without aragonite) with some coarser F4
flood layers (ZrTi mean=025) Unit 3b (98 to 80 cm) is interpreted as a tsunami or
storm surge deposit (breccia F5 grading into massive to banded silt F2) Unit 3a
(79 to 73 cm) is characterized by relatively low productivity (TOC=2 BrTi=002
42
BioSi=4) under anoxic conditions (MnFe=001) Unit 2a (72 to 31cm) has
relatively less organic content and more intercalated clastic facies F3 and F4 The
top of this unit (43-30 cm) has elevated TS values The Subunit 1b (30-20 cm)
shows increasing TOC BioSi and BrTi values (TOC mean = 29 BioSi mean =
54 BrTi mean=004) and the upper subunit 1a (19-0 cm) has the highest TOC
(mean = 64) and BioSi (mean = 56) high BrTi mean = 010 and the presence
of aragonite More frequent anoxic conditions (MnFe lower than 001) during units
3 and 2 shifted towards more oxic episodes (MnFe = 003) in unit 1 (Fig 3)
34 Isotopic signatures
Figure 4 shows the isotopic signature from soil samples of the major land
usescover present in the Laguna Matanzas used as an end member in comparison
with the lacustrine sedimentary units δ15N from cropland samples exhibit the
highest values whereas grassland and soil samples from lake shore areas have
intermediate values (Fig 4) Tree plantations and native forests have similarly low
δ15N values (+11 permil SD=24) All samples (except those from the lake shore)
exhibit low δ13C values (from ndash285permil to ndash298permil) CNmolar from agriculture land
lakeshore area and non-vegetation areas samples display the lowest values (about
18) CNmolar from tree plantations and native forest have the highest values (383
and 267 respectively)
43
Figure 4 C-N stable isotope plot showing a comparison of lake sediments grouped
by sedimentary units (MAT11-6A) with the soil end members of present-day (lake
shore and land usecover) from Laguna Matanzas
The δ15N values from sediment samples (MAT11-6A) range from ndash15 and
+53permil (mean= +35 SD=05) δ13C values range from ndash266permil to ndash202permil (mean=
ndash240 SD=14) In unit 3c δ15N and δ13C show relatively high values (mean=
+41permil and ndash233permil respectively) δ15N and δ13C from unit 3b and 3a fluctuate at
slightly lower values than in 3c (mean δ15N from 3a= +38permil and 3b mean= +39permil
mean δ13C from 3a= ndash242permil and 3b mean= ndash244permil) In unit 2a δ15N values are
relatively high (mean= +38permil) but show a slightly decreasing trend (from +52 to
+24permil at 66 cm and 36 cm respectively) δ13C also decreases (mean= ndash242permil)
reaching minimum values at 45 cm (ndash266permil) with a sharp increase towards the top
of this unit with maximum values ca ndash210permil Unit 1 exhibits the lowest δ15N values
(1a mean= +11permil 1b mean= +28permil) with negative values in the uppermost
44
sediments (ndash04permil at 14 cm) δ13C values show a decreasing trend over most of
subunit 1b and increase only near the very top of this unit
35 Recent land use changes in the Laguna Matanzas watershed
Major LUCC from 1975 (unit 1b) to 2016 CE in the Laguna Matanzasacutes
watershed is summarized in Figure 5 The watershed has a surface area of 30 km2
of which native forest (36) and grassland areas (44) represented 80 of the
total surface in 1975 The area occupied by agriculture was only 02 and tree
plantations were absent Isolated burned areas (33) were located mostly in the
northern part of the watershed By 1989 tree plantations surface area had increased
to 5 burned areas to 17 and agricultural fields to 9 (a 45-fold increase) and
native forest and grassland sectors decreased to 23 and 27 respectively By
2016 agricultural land and tree plantations have increased to 17 of the total area
whereas native forests decreased to 21
45
Figure 5 Land uses and cover changes from 1975 to 2016 in Laguna Matanzas
watershed from natural cover and areas for livestock grazing (grassland) to the
expansion of agriculture and forest plantation
4 DISCUSSION
41 N and C dynamics in Laguna Matanzas
Small lakes with relatively large watersheds such as Laguna Matanzas would
be expected to have relatively high contributions of allochthonous C to the sediment
OM (Gu et al 2006) Terrestrial C3 plants (δ13C =ndash26 to ndash28permil Meyers and Terranes
2001 Ku et al 2007) are dominant in the Laguna Matanzas watershed Likewise
our soil samples ranged across similar although slightly more negative values
46
(δ13C = ndash30 to ndash28permil Fig 4) to those proposed by Meyers and Terranes (2001)
and are used here as terrestrial end members oil samples were taken from the lake
shore (δ13C = ndash22permil) and POM from surface water (δ13C = ndash24permil) were more
positive than the terrestrial end member and are used as lacustrine end members
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from terrestrial vegetation and more positive δ13C values have increased
aquatic OM (Gu et al 2006 Torres et al 2012) Phytoplankton preferentially uptake
12C leaving the DIC pool enriched in 13C (Gu et al 2006) especially when there are
no important external sources of C (eg decreased C input from the watershed)
Therefore during events of elevated primary productivity the phytoplankton uptakes
12C until its depletion and are then obligated to use the heavier isotope resulting in
an increase in δ13C Changes in lake productivity thus greatly affect the C isotope
signal (Torres et al 2012) with high productivity leading to elevated δ13C values
(Torres et al 2012 Gu et al 2006)
In a similar fashion the N isotope signatures in Laguna Matanzas reflect a
combination of factors including different N sources (autochthonousallochthonous)
and lake processes such as productivity isotope fractionation in the water column
and sediment denitrification Elevated δ15N values from a POM sample (+22permil) and
average values from the lake shore (mean=+34permil SD=028) are used as aquatic
end members whereas terrestrial samples have values from +10 +24 (tree species)
to +77permil +35 (agriculture) which we use as terrestrial end members (Fig 4)
Autochthonous OM in aquatic ecosystems typically displays low δ15N values
when the OM comes from N-fixing species Atmospheric fixation of N2 by
cyanobacteria ends in OM with δ15N values close to 0permil (Leng et al 2006)
Phytoplankton preferentially uptake 14N from Dissolved Inorganic Nitrogen (DIN) in
47
the water column and derived OM typically have δ15N values lower than DIN values
When productivity increases the remaining DIN becomes depleted in 14N which in
turn increases the δ15N values of phytoplankton over time especially if the N not
replenished (Torres et al 2012) Thus high POM δ15N values from Laguna
Matanzas reflect elevated phytoplankton productivity with a 14N- depleted DIN In
addition N-watershed inputs also contribute to high δ15N values Heavily impacted
watersheds by human activities are often reflected in isotope values due to land use
changes and associated modified N fluxes For example the input of N runoff
derived from the use of inorganic fertilizers leads to the presence of elevated δ15N
(between ndash4 to +4permil) values in the water bodies (Elliott and Brush 2006 Diebel and
Vander Zanden 2009) Widory et al (2004) reported a direct relationship between
elevated δ15N values and increased nitrate concentration from manure in the
groundwater of Brittany (France) Terranes and Bernasconi (2000) found a good
correlation between augmented nutrient loading and a progressive increase in δ15N
values of sedimentary OM related to agricultural land use
Post-depositional diagenetic processes can further affect C and N isotope
signatures Several studies have shown a decrease in δ13C values of OM in anoxic
environments particularly during the first years of burial related to the selective
preservation of OM depleted in 13C (Hollander and Smith 2001 Lehmann et al
2002 Gӓlman et al 2009 Torres et al 2012) Diagenetic processes can also lead
to post-burial N isotope enrichment of the sediments Indeed 14N is consumed more
rapidly than 15N by denitrifying bacteria which intensifies under anoxic conditions
(Diebel and Vander Zanden 2009) Thus the remaining OM pool will be enriched
in 15N leaving to elevated δ15N values (Granger et al 2008 Menzel et al 2013)
48
In summary the relatively high δ15N values in sediments of Laguna Matanzas
reflect N input from an agriculturegrassland watershed with positive synergetic
effects from increased lake productivity enrichment of DIN in the water column and
most likely denitrification The increase of algal productivity associated with
increased N terrestrial input andor recycling of lake nutrients (and lesser extent
fixing atmospheric N) and denitrification under anoxic conditions can all increase
δ15N values (Fig 3) In addition elevated lake productivity without C replenishing
(eg by terrestrial C input) produces shifts towards positive δ13C values whereas C
input from the watershed generates more negative δ13C values
42 Recent evolution of the Laguna Matanzas watershed
Sedimentological compositional and geochemical indicators show three
depositional phases in the lake evolution under the human influence in the Laguna
Matanzas over the last two hundred years Although the record is longer (around
1000 years) we analyzed the last two centuries (unit 2 and 1) to provide a recent
historical context for the large changes detected during the 20th century
The first phase lasted from the beginning of the 19th century until ca 1940
(unit 2a Fig 3 4) and was characterized by moderate productivity with elevated
sediment input from the watershed as indicated by our geochemical proxies (BrTi
= 004 AlTi gt 01 Fig 6) Higher lake levels and dominant anoxic bottom conditions
(MnFe = 001 Fig3) can be related to increased precipitation (mean= 557 mmyr)
and lower temperatures (summer annual temperature lt19ordmC) During the Spanish
colonial period the Laguna Matanzas watershed was used as a livestock farm
(Jesuits 1627-1780 unit 3 followed by the Pedro Balmaceda era 1780-1810- unit
2a) (Contreras-Loacutepez et al 2014) The main buildings and farm facilities at the El
49
Convento village During this period livestock grazing and lumber extraction for
mining would have involved extensive deforestation and loss of native vegetation
(eg Armesto et al 1994 2010) However the Matanzas pollen record does not
show any significant regional deforestation during this period (Villa Martiacutenez 2002)
suggesting that the impact may have been highly localized
Lake productivity sediment input and elevated precipitation (Fig 6) all
suggest that N availability was related to this increased input from the watershed
The N from cow manure and soil particles would have led to higher δ15N values
(Elliot and Brush 2016) In addition anoxic lake bottom conditions would lead to
even further enrichment of buried sediment N The δ13C values lend further support
to our interpretation of increased sediment input -and N- from the watershed
Decreased δ13C values reach their lowest values in the entire record (ndash270permil) at
ca 1910 CE (Fig 4 6)
During most of the 19th century human activities in Laguna Matanzas were
similar to those during the Spanish Colonial period However the appearance of
Pinus radiata and Eucalyptus globulus pollen ca 1890 CE (Gibson 1972) the dune
stabilization-afforestation program which began ca 1900 CE (Albert 1900) and the
application of the Chilean forestry law of 1931 (DFL nordm265) contributed to an
increased capacity of the surrounding vegetation to retain nutrients and sediments
The law subsidized forest plantations in areas devoid of vegetation and prohibited
the cutting of forest on slopes greater than 45ordm These land use changes were coeval
with decreased sediment inputs (AlTi trend) from the watershed slightly increased
lake productivity (BrTi from 001 to 003) and a decrease in annual precipitation
(Fig 6) N isotope values become more negative during this period although they
remained high (from +49permil to +37permil) whereas the δ13C trend towards more
50
positive values reflects changes in the N source from watershed to in-lake dynamics
(e g increased endogenic productivity)
The second phase started after 1940 and is clearly marked by an abrupt
change in the general trend of δ15N that oscillates around +41permil (SD = 04permil) during
the first phase to +24permil (SD = 14permil) during the second These δ15N trends reflect
the lowest watershed nutrient and sediment inputs (based on the AlTi record)
decreased precipitation (mean = 318 mm year) and a slight increase in lake
productivity (increased BrTI) Depositional dynamics in the lake likely crossed a
threshold as human activity intensified throughout the watershed and lake levels
decreased
During the Great Acceleration δ15N values shifted towards higher values to
ca 3permil with an increase in δ13C values that are not reflected either in lake
productivity or lake level As the sediment input from the watershed increased and
precipitation remained as low as the previous decade δ15N values during this period
are likely related to watershed clearance which would have increased both nutrient
and sediment input into the lake
The δ13C trend to more positive values reaching the peaks in the 1960s (ndash
212permil) at the same time as the peaks in temperature (196 ordmC mean annual) with a
downward trend in precipitation A shift in OM origin from macrophytes and
watershed input influences to increased lake productivity could explain this trend
(Fig 4 1b)
In the 1970s the Laguna Matanzasacute watershed was mostly covered by native
forest (36) and grassland areas were intended for livestock grazing (44 Fig 5)
Both soils have δ15NSoil lt 5permil (Fig 4) Farming fields occupied a very small area and
tree plantations were almost nonexistent The decreasing trend in δ15N values seen
51
in our record is interrupted by several large peaks that occurred between ca 1975
and ca 1989 when the native forest and grassland areas fell by 23 and 27
respectively largely due to fires affecting 17 of the forests Agriculture fields
increased by 4 and forest plantations increased by 9 (unit 1a) Concomitantly
sediment ndash and likely N - inputs from the watershed decreased (as indicated by the
trend in AlTi) the precipitation was still relatively low (Fig 6) These changes are
likely related to the increase of vegetation cover especially of tree plantations (which
have more negative δ15N values) The small increase in productivity in the lake could
have been favored by increased temperature (von Gunten et al 2009) After 1989
the increase in agriculture land (17 in 2016) correlates with increasing δ15N δ13C
and TOC trends in spite of declining rainfall The increase of forest plantations was
mostly in response to the implementation of the Law Decree of Forestry
Development (DL 701 of 1974) that subsidized forest plantation After 1989 the
increase in agricultural land (17 in 2016) is synchronous with increasing δ15N
δ13C TOC and MnFe trends despite the decline in rainfall and overall lower lake
levels as more water is used for irrigation
The third phase started c 1990 CE (unit 1a) when OM accumulation rates
increase and δ13C δ15N decreased reaching their lowest values in the sequence
around 2000 CE Afterward during the 21st century δ13C and δ15N values again
began to increase The onset of unit 1 is marked by increased lake productivity and
decreased sediment input (AlTi lt 02) synchronous with intensive farming replacing
forestry and extensive agriculture (Fig 5 6)
A change in the general trend of δ15N values which decreased until 1990
(unit 1a) as the native forest and grassland area fell to 23 and 27 respectively
is most likely due to deforestation and fires Agriculture surface increased to 4 and
52
forest plantations increased by 9 (Fig 5) Concomitantly sediment ndash and likely N
ndash inputs from the watershed decreased probably related to the low precipitation (Fig
1b) and the increase of vegetation cover in the watershed in particularly by tree
plantations (with more negative δ15N Fig 4)
At present agriculture and tree plantations occupy around 34 of the
watershed surface whereas native forests and grassland cover 21 and 25
respectively Lake productivity as indicated by BrTi (Fig6) is higher and generates
OM with lower δ15N and higher δ13C values (about +2permil and ndash20permil ca 2000 CE
respectively) due to in-lake processes (ie biological N fixation and nutrient
recycling) and driven by changes in the arboreal cover which diminishes nutrient
flux into the lake (Fig6)
53
Figure 6 Anthropogenic and climatic forcing and lake dynamics response
(productivity sediment input N and C cycles) at Matanzas Lake over the last two
54
centuries Mean annual precipitation reconstructed and temperatures (von Gunten
et al 2009) Vertical gray bars indicate mega-droughts
5 CONCLUSIONS
Human activities have been the main factor controlling the N and C cycle in
the Laguna Matanzas during the last two centuries The N isotope signature in the
lake sediments reflects changes in the watershed fluxes to the lake but also in-lake
processes such as productivity and post-depositional changes Denitrification could
have been a dominant process during periods of increased anoxic conditions which
were more frequent prior to Great Acceleration (1950 CE) Higher δ15N and lower
δ13C values are associated with increased nutrient input from the watershed due to
increased livestock grazing and agriculture pressure (phase 1 Fig 7a) whereas
lower isotope values occurred during periods of increased forest plantations (phase
3 Fig 7c) During periods of increased lake productivity - such as in the last few
decades - δ15N values increased significantly
The most important change in C and N dynamics in the lake occurred after the
1950s (Phase 2 and 3) as tree plantations dominated the watershed Recent
changes in N dynamics can be explained by the higher nutrient contribution
associated with intensive agriculture (i e fertilizers) since the 1990s Although the
replacement of livestock activities with forestry and farming seems to have reduced
nutrient and soil export from the watershed to the lake the inefficient use of fertilizer
(by agriculture) can be the ultimate responsible for lake productivity increase during
the last decades
55
Figure 7 Schematic diagrams illustrating the main factors controlling the
isotope N signal in sediment OM of Laguna Matanzas N input from watershed
depends on human activities and land cover type Agriculture practices and cattle
(grassland development) contribute more N to the lake than native forest and
plantations Periods of higher productivity tend to deplete the dissolved inorganic N
in 14N resulting in higher δ15N (OM) The denitrification processes are more effective
in anoxic conditions associated with higher lake levels
6 METHODS
Short sediment cores were recovered from Laguna Matanzas using an Uwitec
gravity piston corer in 2011 and 2013 (Fig 1a) Sediment cores (MAT11-6A 129 cm
MAT13-2A 149 cm MAT13-3A 33 cm MAT13-4A 99 cm) were split
photographed sub-sampled and stored at the Pyrenean Institute of Ecology (IPE-
CSIC Spain) Core MAT11-6A was obtained from the central sector of the lake and
56
was selected for detailed multiproxy analyses (including elemental geochemistry C
and N isotope analyses XRF and 14C dating)
The isotope analyses (δ13C and δ15N) were performed at the Laboratory of
Biogeochemistry and Applied Stable Isotopes (LABASI-PUC Chile) using a Delta
V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via a
Conflo IV interface Isotope results are expressed in standard delta notation (δ) in
per mil (permil) relative to the standards Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Sediment samples
for δ13Corg were pre-treated with 50 ml of HCl and 50 ml of deionized water and
dried at 60ordmC for 4 hr to remove carbonates (Harris et al 2011)
Total Carbon (TC) Total Organic Carbon (TOC) Total Inorganic Carbon (TIC)
and Total Sulphur (TS) were measured every cm with a LECO SC 144 DR at IPE-
CSIC XRF measurements were carried out every 4 mm in MAT11-1A-1G core using
an AVAATECH X-ray Fluorescence II core scanner at the University of Barcelona
(Spain) Results are expressed as element intensities in counts per second (cps)
Tube voltage was operated at 30 kV and 10 kV to obtain the abundances of 15
elements (Al Si S Cl K Ca Ti V Mn Fe Br Rb Sr Y Zr) with an average at
least of 1600 cps (less for Br=1000)
Biogenic silica content mineralogy and grain size were measured every 4
cm Biogenic silica was measured following Mortlock and Froelich (1989) and
Bernaacuterdez et al (2005) using an Auto Analyzer Technicon AAII for dissolved silicate
analysis Mineralogy was analyzed with a Siemens D-500 X-ray diffractometer (Cu
kα 40 kV 30 mA graphite monochromator) at the ICTJA-CSIC (Spain) Grain size
analyses were performed in a Beckmann Coulter LS 13 320 Particle Size Analyzer
57
at the IPE-CSIC The samples were classified according to textural classes as
follows clay (lt2 μm) silt (20-2 microm) and sand (gt2 microm) fractions
The age-depth model for the Laguna Matanzas sedimentary sequence was
constructed using 210Pb and 137Cs dating techniques (MAT13-4A) as well as two 14C
AMS dates on bulk sediment samples (MAT11-6A fig 2) We dated the dissolved
inorganic carbon (DIC) in the water column and no significant reservoir effect is
present in the modern-day water column (10454 + 035 pcmc Table 2) An age-
depth model was obtained with the Bacon R package to estimate the deposition
rates and associated age uncertainties along the core (Blaauw and Christen 2011)
To estimate LUCC of Laguna Matanzasacutes watershed we use satellite images
Landsat MSS for 1975 Landsat TM for 1989 and Landsat OLI for 2016 all taken in
summer or autumn (Table 1) We performed supervised classification of land uses
(maximum likelihood algorithm) for each year (1975 1989 and 2016) and the results
were mapped using software ArcGIS 102 in 2017
Satellite Images Acquisition Date
Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat OLI 20160404 30 m
Table 1 Landsat imagery
Surface water samples were filtered for obtained particulate organic matter In
addition soil samples from the main land usecover present in the Laguna Matanzas
watershed were collected Elemental C N and their corresponding isotopes from
POM and soil were obtained at the LABASI and used here as end members
Daily precipitation at Santo Domingo (33ordm39`S 71ordm3 6`W)- the nearest weather
station to Laguna Matanzasndash was compiled using the redPrec R package (Fig1 d
Serrano-Notivoli et al 2017 Sarricolea et al 2019) To extend our precipitation
58
reconstruction back to 1824 we correlated this dataset with that available for
Santiago The Santiago data was compiled from data published in the Anales of
Universidad of Chile (1850) for the years 1824 to 1849 Almeyda (1940) for the years
1849 to 1864 and the Quinta Normal series from 1866 to the present (Direccioacuten
Meteoroloacutegica de Chile) We generated a linear regression model between the
presentday Santo Domingo station and the compiled Santiago data with a Pearson
coefficient of 087 and p-valuelt 001
Acknowledgments This research was funded by grants CONICYT AFB170008
to the Institute of Ecology and Biodiversity (IEB) and FONDECYT 1150763 (to CL)
Doctoral grant Becas Chile 21150224 MEDLANT (Spanish Ministry of Economy
and Competitiveness grant CGL2016-76215-R) Additional funding was provided
by the Laboratorio Internacional de Cambio Global (LINCGlobal PUC-CSIC) We
thank R Lopez E Royo and M Gallegos for help with sample analyses We thank
the Laboratory of Biogeochemistry and Applied Stable Isotopes (LABASI) of the
Department of Ecology (PUC) for sample analyses
References
Albert F 1900 Las dunas o sean arenas volantes voladeros arenas muertas invasion de las arenas playas I meacutedanos del centro de Chile comprendiendo el litoral desde el liacutemite norte de la provincia de Aconcagua hasta el liacutemite sur de la de Arauco Memorias cientiacuteficas I Lit
Anderson NJ W D Fritz SC 2009 Holocene carbon burial by lakes in SW
Greenland Glob Chang Biol httpsdoiorg101111j1365-2486200901942x
Armesto J Villagraacuten C Donoso C 1994 La historia del bosque templado
chileno Ambient y Desarro 66 66ndash72 httpsdoiorg101007BF00385244
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A
59
historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Arnell NW Gosling SN 2013 The impacts of climate change on river flow
regimes at the global scale J Hydrol 486 351ndash364 httpsdoiorg101016JJHYDROL201302010
Bernaacuterdez P Prego R Franceacutes G Gonzaacutelez-Aacutelvarez R 2005 Opal content in
the Riacutea de Vigo and Galician continental shelf Biogenic silica in the muddy fraction as an accurate paleoproductivity proxy Cont Shelf Res httpsdoiorg101016jcsr200412009
Blaauw M Christen JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474 httpsdoiorg10121411-BA618
Brush GS 2009 Historical land use nitrogen and coastal eutrophication A
paleoecological perspective Estuaries and Coasts 32 18ndash28 httpsdoiorg101007s12237-008-9106-z
Camarero L Catalan J 2012 Atmospheric phosphorus deposition may cause
lakes to revert from phosphorus limitation back to nitrogen limitation Nat Commun httpsdoiorg101038ncomms2125
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego
R Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and environmental change from a high Andean lake Laguna del Maule central Chile (36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Catalan J Pla-Rabeacutes S Wolfe AP Smol JP Ruumlhland KM Anderson NJ
Kopaacuteček J Stuchliacutek E Schmidt R Koinig KA Camarero L Flower RJ Heiri O Kamenik C Korhola A Leavitt PR Psenner R Renberg I 2013 Global change revealed by palaeolimnological records from remote lakes A review J Paleolimnol httpsdoiorg101007s10933-013-9681-2
Chen Y Y Huang W Wang W H Juang J Y Hong J S Kato T amp
Luyssaert S (2019) Reconstructing Taiwanrsquos land cover changes between 1904 and 2015 from historical maps and satellite images Scientific Reports 9(1) 3643 httpsdoiorg101038s41598-019-40063-1
Contreras S Werne J P Araneda A Urrutia R amp Conejero C A (2018)
Organic matter geochemical signatures (TOC TN CN ratio δ 13 C and δ 15 N) of surface sediment from lakes distributed along a climatological gradient on the western side of the southern Andes Science of The Total Environment 630 878-888 httpsdoiorg101016jscitotenv201802225
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia UC
60
de V 2014 ELEMENTOS DE LA HISTORIA NATURAL DEL An Mus Hist Natulas Vaplaraiso 27 51ndash67
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-010-9453-1
Diebel M Vander Zanden MJ 201209 Nitrogen stable isotopes in streams
stable isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19 1127ndash1134 httpsdoiorg10189008-03271
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506 Espizua LE Pitte P 2009 The Little Ice Age glacier advance in the Central
Andes (35degS) Argentina Palaeogeogr Palaeoclimatol Palaeoecol 281 345ndash350 httpsdoiorg101016JPALAEO200810032
Figueroa JA Castro S A Marquet P A Jaksic FM 2004 Exotic plant
invasions to the mediterranean region of Chile causes history and impacts Invasioacuten de plantas exoacuteticas en la regioacuten mediterraacutenea de Chile causas historia e impactos Rev Chil Hist Nat 465ndash483 httpsdoiorg104067S0716-078X2004000300006
Fowler D Coyle M Skiba U Sutton MA Cape JN Reis S Sheppard
LJ Jenkins A Grizzetti B Galloway N Vitousek P Leach A Bouwman AF Butterbach-bahl K Dentener F Stevenson D Amann M Voss M 2013 The global nitrogen cycle in the twenty- first century Philisophical Trans R Soc B httpsdoiorghttpdxdoiorg101098rstb20130164
Frank D Reichstein M Bahn M Thonicke K Frank D Mahecha MD
Smith P van der Velde M Vicca S Babst F Beer C Buchmann N Canadell JG Ciais P Cramer W Ibrom A Miglietta F Poulter B Rammig A Seneviratne SI Walz A Wattenbach M Zavala MA Zscheischler J 2015 Effects of climate extremes on the terrestrial carbon cycle Concepts processes and potential future impacts Glob Chang Biol httpsdoiorg101111gcb12916
Fritz SC Anderson NJ 2013 The relative influences of climate and catchment
processes on Holocene lake development in glaciated regions J Paleolimnol httpsdoiorg101007s10933-013-9684-z
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
61
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS) implications for past sea level and environmental variability J Quat Sci 32 830ndash844 httpsdoiorg101002jqs2936
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892 httpsdoiorg101126science1136674
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924 httpsdoiorg104319lo20095430917
Garciacutea-Ruiz JM 2010 The effects of land uses on soil erosion in Spain A
review Catena httpsdoiorg101016jcatena201001001
Garciacutea-Ruiz JM Loacutepez-Moreno JI Lasanta T Vicente-Serrano SM
Gonzaacutelez-Sampeacuteriz P Valero-Garceacutes BL Sanjuaacuten Y Begueriacutea S Nadal-Romero E Lana-Renault N Goacutemez-Villar A 2015 Los efectos geoecoloacutegicos del cambio global en el Pirineo Central espantildeol una revisioacuten a distintas escalas espaciales y temporales Pirineos httpsdoiorg103989Pirineos2015170005
Garciacutea-Ruiz JM Nadal-Romero E Lana-Renault N Begueriacutea S 2013
Erosion in Mediterranean landscapes Changes and future challenges Geomorphology 198 20ndash36 httpsdoiorg101016jgeomorph201305023
Garreaud RD Vuille M Compagnucci R Marengo J 2009 Present-day
South American climate Palaeogeogr Palaeoclimatol Palaeoecol httpsdoiorg101016jpalaeo200710032
Gayo EM Latorre C Jordan TE Nester PL Estay SA Ojeda KF
Santoro CM 2012 Late Quaternary hydrological and ecological changes in the hyperarid core of the northern Atacama Desert (~21degS) Earth-Science Rev httpsdoiorg101016jearscirev201204003
Ge Y Zhang K amp Yang X (2019) A 110-year pollen record of land use and land
cover changes in an anthropogenic watershed landscape eastern China Understanding past human-environment interactions Science of The Total Environment 650 2906-2918 httpsdoiorg101016jscitotenv201810058
Goyette J Bennett EM Howarth RW Maranger R 2016 Global
Biogeochemical Cycles 1000ndash1014 httpsdoiorg1010022016GB005384Received
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and
oxygen isotope fractionation during dissimilatory nitrate reduction by
62
denitrifying bacteria 53 2533ndash2545 httpsdoiorg10230740058342
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
Gurnell AM Petts GE Hannah DM Smith BPG Edwards PJ Kollmann
J Ward J V Tockner K 2001 Effects of historical land use on sediment yield from a lacustrine watershed in central Chile Earth Surf Process Landforms 26 63ndash76 httpsdoiorg1010021096-9837(200101)261lt63AID-ESP157gt30CO2-J
Heathcote A J et al Large increases in carbon burial in northern lakes during the
Anthropocene Nat Commun 610016 doi 101038ncomms10016 (2015) Hollander DJ Smith MA 2001 Microbially mediated carbon cycling as a
control on the δ13C of sedimentary carbon in eutrophic Lake Mendota (USA) New models for interpreting isotopic excursions in the sedimentary record Geochim Cosmochim Acta httpsdoiorg101016S0016-7037(00)00506-8
Holtgrieve GW Schindler DE Hobbs WO Leavitt PR Ward EJ Bunting
L Chen G Finney BP Gregory-Eaves I Holmgren S Lisac MJ Lisi PJ Nydick K Rogers LA Saros JE Selbie DT Shapley MD Walsh PB Wolfe AP 2011 A coherent signature of anthropogenic nitrogen deposition to remote watersheds of the Northern Hemisphere Science (80) 334 1545ndash1548 httpsdoiorg101126science1212267
Hooper DU Adair EC Cardinale BJ Byrnes JEK Hungate BA Matulich
KL Gonzalez A Duffy JE Gamfeldt L Connor MI 2012 A global synthesis reveals biodiversity loss as a major driver of ecosystem change Nature httpsdoiorg101038nature11118
Horvatinčić N Sironić A Barešić J Sondi I Krajcar Bronić I Borković D
2016 Mineralogical organic and isotopic composition as palaeoenvironmental records in the lake sediments of two lakes the Plitvice Lakes Croatia Quat Int httpsdoiorg101016jquaint201701022
Howarth RW 2004 Human acceleration of the nitrogen cycle Drivers
consequences and steps toward solutions Water Sci Technol httpsdoiorg1010382Fscientificamerican0490-56
Jenny B Valero-Garceacutes BL Urrutia R Kelts K Veit H Appleby PG Geyh
M 2002 Moisture changes and fluctuations of the Westerlies in Mediterranean Central Chile during the last 2000 years The Laguna Aculeo record (33deg50primeS) Quat Int 87 3ndash18 httpsdoiorg101016S1040-
63
6182(01)00058-1
Jenny B Wilhelm D Valero-Garceacutes BL 2003 The Southern Westerlies in
Central Chile Holocene precipitation estimates based on a water balance model for Laguna Aculeo (33deg50primeS) Clim Dyn 20 269ndash280 httpsdoiorg101007s00382-002-0267-3
Leng M J Lamb A L Heaton T H Marshall J D Wolfe B B Jones M D
amp Arrowsmith C 2006 Isotopes in lake sediments In Isotopes in palaeoenvironmental research (pp 147-184) Springer Dordrecht
Le Roux JP Nielsen SN Kemnitz H Henriquez Aacute 2008 A Pliocene mega-
tsunami deposit and associated features in the Ranquil Formation southern Chile Sediment Geol 203 164ndash180 httpsdoiorg101016jsedgeo200712002
Lehmann M Bernasconi S Barbieri a McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66 3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Lenihan JM Drapek R Bachelet D Neilson RP 2003 Climate change
effects on vegetation distribution carbon and fire in California Ecol Appl httpsdoiorg101890025295
McLauchlan K K Gerhart L M Battles J J Craine J M Elmore A J
Higuera P E amp Perakis S S (2017) Centennial-scale reductions in nitrogen availability in temperate forests of the United States Scientific reports 7(1) 7856 httpsdoi101038s41598-017-08170-z
Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32ordmS 3050 masl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
Martiacuten-Foreacutes I Casado MA Castro I Ovalle C Pozo A Del Acosta-Gallo
B Saacutenchez-Jardoacuten L Miguel JM De 2012 Flora of the mediterranean basin in the chilean ESPINALES Evidence of colonisation Pastos 42 137ndash160
Marzecovaacute A Mikomaumlgi a Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54 httpsdoiorg103176eco2011105
Matesanz S Valladares F 2014 Ecological and evolutionary responses of
Mediterranean plants to global change Environ Exp Bot httpsdoiorg101016jenvexpbot201309004
64
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Menzel P Gaye B Wiesner MG Prasad S Stebich M Das BK Anoop A
Riedel N Basavaiah N 2013 Influence of bottom water anoxia on nitrogen isotopic ratios and amino acid contributions of recent sediments from small eutrophic Lonar Lake central India Limnol Oceanogr 58 1061ndash1074 httpsdoiorg104319lo20135831061
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Michelutti N Wolfe AP Cooke CA Hobbs WO Vuille M Smol JP 2015
Climate change forces new ecological states in tropical Andean lakes PLoS One httpsdoiorg101371journalpone0115338
Muntildeoz-Rojas M De la Rosa D Zavala LM Jordaacuten A Anaya-Romero M
2011 Changes in land cover and vegetation carbon stocks in Andalusia Southern Spain (1956-2007) Sci Total Environ httpsdoiorg101016jscitotenv201104009
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Poraj-Goacuterska A I Żarczyński M J Ahrens A Enters D Weisbrodt D amp
Tylmann W (2017) Impact of historical land use changes on lacustrine sedimentation recorded in varved sediments of Lake Jaczno northeastern Poland Catena 153 182-193 httpdxdoiorg101016jcatena201702007
Pongratz J Reick CH Raddatz T Claussen M 2010 Biogeophysical versus
biogeochemical climate response to historical anthropogenic land cover change Geophys Res Lett 37 1ndash5 httpsdoiorg1010292010GL043010
Prudhomme C Giuntoli I Robinson EL Clark DB Arnell NW Dankers R
Fekete BM Franssen W Gerten D Gosling SN Hagemann S Hannah DM Kim H Masaki Y Satoh Y Stacke T Wada Y Wisser D 2014 Hydrological droughts in the 21st century hotspots and uncertainties from a global multimodel ensemble experiment Proc Natl Acad Sci httpsdoiorg101073pnas1222473110
65
Ruumlhland KM Paterson AM Smol JP 2015 Lake diatom responses to
warming reviewing the evidence J Paleolimnol httpsdoiorg101007s10933-
015-9837-3
Sarricolea P Meseguer-Ruiz Oacute Serrano-Notivoli R Soto M V amp Martin-Vide
J (2019) Trends of daily precipitation concentration in Central-Southern Chile Atmospheric research 215 85-98 httpsdoiorg101016jatmosres201809005
Sernageomin 2003 Mapa geologico de chile version digital Publ Geol Digit Serrano-Notivoli R de Luis M Begueriumlia S 2017 An R package for daily
precipitation climate series reconstruction Environ Model Softw httpsdoiorg101016jenvsoft201611005
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Stine S 1994 Extreme and persistent drought in California and Patagonia during
mediaeval time Nature 369 546ndash549 httpsdoiorg101038369546a0
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL
Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Syvitski JPM Voumlroumlsmarty CJ Kettner AJ Green P 2005 Impact of humans
on the flux of terrestrial sediment to the global coastal ocean Science 308(5720) 376-380 httpsdoiorg101126science1109454
Thomas RQ Zaehle S Templer PH Goodale CL 2013 Global patterns of
nitrogen limitation Confronting two global biogeochemical models with observations Glob Chang Biol httpsdoiorg101111gcb12281
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Urbina Carrasco MX Gorigoitiacutea Abbott N Cisternas Vega M 2016 Aportes a
la historia siacutesmica de Chile el caso del gran terremoto de 1730 Anuario de Estudios Americanos httpsdoiorg103989aeamer2016211
Vargas G Ortlieb L Chapron E Valdes J Marquardt C 2005 Paleoseismic
inferences from a high-resolution marine sedimentary record in northern Chile
66
(23degS) Tectonophysics httpsdoiorg101016jtecto200412031 399(1-4) 381-398httpsdoiorg101016jtecto200412031
Valero-Garceacutes BL Latorre C Morelliacuten M Corella P Maldonado A Frugone M Moreno A 2010 Recent Climate variability and human impact from lacustrine cores in central Chile 2nd International LOTRED-South America Symposium Reconstructing Climate Variations in South America and the Antarctic Peninsula over the last 2000 years Valdivia p 76 (httpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-yearshttpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-years
Vicente-Serrano SM Gouveia C Camarero JJ Begueria S Trigo R
Lopez-Moreno JI Azorin-Molina C Pasho E Lorenzo-Lacruz J Revuelto J Moran-Tejeda E Sanchez-Lorenzo A 2013 Response of vegetation to drought time-scales across global land biomes Proc Natl Acad Sci httpsdoiorg101073pnas1207068110
Villa Martiacutenez R P 2002 Historia del clima y la vegetacioacuten de Chile Central
durante el Holoceno Una reconstruccioacuten basada en el anaacutelisis de polen sedimentos microalgas y carboacuten PhD
Villa-Martiacutenez R Villagraacuten C Jenny B 2003 Pollen evidence for late -
Holocene climatic variability at laguna de Aculeo Central Chile (lat 34o S) The Holocene 3 361ndash367
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Human alteration of the global nitrogen cycle sources and consequences Ecological applications 7(3) 737-750 httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Rein B Urrutia R amp Appleby P (2009) A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 The Holocene 19(6) 873-881 httpsdoiorg1011770959683609336573
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Widory D Kloppmann W Chery L Bonnin J Rochdi H Guinamant JL
2004 Nitrate in groundwater An isotopic multi-tracer approach J Contam Hydrol 72 165ndash188 httpsdoiorg101016jjconhyd200310010
67
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
68
Supplementary material
Facie Name Description Depositional Environment
F1 Organic-rich
mud
Massive to banded black
organic - rich (TOC up to 14 )
mud with aragonite in dm - thick
layers Slightly banded intervals
contain less OM (TOClt4) and
aragonite than massive
intervals High MnFe (oxic
bottom conditions) High CaTi
BrTi and BioSi (up to 5)
Distal low energy environment
high productivity well oxygenated
and brackish waters and relative
low lake level
F2 Massive to
banded silty clay
to fine silt
cm-thick layers mostly
composed by silicates
(plagioclase quartz cristobalite
up to 65 TOC mean=23)
Some layers have relatively high
pyrite content (up to 25) No
carbonates CaTi BrTi and
BioSi (mean=48) are lower
than F1 higher ZrTi (coarser
grain size)
Deposition during periods of
higher sediment input from the
watershed
69
F3 Banded to
laminated light
brown silty clay
cm-thick layers mostly
composed of clay minerals
quartz and plagioclase (up to
42) low organic matter
(TOC mean=13) low pyrite
and BioSi content
(mean=46) and some
aragonite
Flooding events reworking
coastal deposits
F4 Laminated
coarse silts
Thin massive layers (lt2mm)
dominated by silicates Low
TOC (mean=214 ) BrTi
(mean=002) MnFe (lt02)
TIC (lt034) BioSi
(mean=46) and TS values
(lt064) and high ZrTi
Rapid flooding events
transporting material mostly
from within the watershed
F5 Breccia with
coarse silt
matrix
A 17 cm thick (80-97 cm
depth) layer composed by
irregular mm to cm-long ldquosoft-
clastsrdquo of silty sediment
fragments in a coarse silt
matrix Low CaTi BrTi and
MnFe ratios and BioSi
Rapid high energy flood
events
70
(mean=43) and high ZrTi
(gt018)
Table Sedimentological and compositional characteristics of Laguna Matanzas
facies
71
CAPIacuteTULO 2 STABLE ISOTOPES TRACK LAND USE AND COVER
CHANGES IN A MEDITERRANEAN LAKE IN CENTRAL CHILE OVER THE
LAST 600 YEARS
72
Stable isotopes track land use and cover changes in a mediterranean lake in
central Chile over the last 600 years
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugone-Aacutelvarezabc Pablo
Sarricolead Leonardo Villaciacutese M Laura Carrevedob Blas Valero-Garceacutescg
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Departamento de Ecologiacutea Universidad de ChileLas Palmeras 3425 Nuntildeoa Santiago Chile
f Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding authors E-mail addresses clatorrebiopuccl magdalenafuentealbagmailcom
Key words Anthropocene lake sediment δsup1⁵N δsup1sup3C Great Acceleration organic
geochemistry watershedndashlake system Stable Isotope Analyses land usecover
change Nitrogen cycle mediterranean ecosystems central Chile
73
Abstract
Nutrient fluxes in many aquatic ecosystems are currently being overridden by
anthropic controls especially since the industrial revolution (mid-1800s) and the
Great Acceleration (mid-1900s) Land use and cover changes (LUCC) alter the
availability and fluxes of nutrients such as nitrogen that are transferred via runoff
and groundwater into lakes By altering lake productivity and trophic status these
changes are often preserved in the sedimentary record Here we use
geolimnological proxies and stable isotope analyses (δsup1⁵N δsup13C) on lake sediments
to reconstruct the lake-watershed nutrient transfer in a central Chilean lake (Lago
Vichuqueacuten) over the past 600 years We compare our multiproxy record from recent
lake sediments to the soilvegetation relationship across the watershed as well as
land usecover changes from 1975 to 2014 derived from satellite images Our results
show that lake sediment δsup1⁵N values increased with meadow cover but decreased
with tree plantations suggesting increased nitrogen retention when trees dominate
the watershed Our δsup1⁵N record from central Chile can thus be interpreted as a proxy
for nutrient availability over the last 600 years mainly derived from land use changes
coupled with climate drivers Although variable sources of organic matter and in situ
fractionation often hinder straightforward environmental interpretations of stable N
isotope signatures our study shows that lake sediment δsup1⁵N can be a useful tool for
assessing the contribution of past human activities in nutrient and nitrogen cycling
1 Introduction
Nitrogen (N) is a limiting nutrient in terrestrial and aquatic ecosystems (Vitousek
et al 1997) Changes in its availability can drive eutrophication and increase
pollution in these ecosystems (McLauchlan et al 2013) Although recent human
74
impacts on the global N cycle have been significant the consequences of increased
anthropogenic N on these ecosystems are unclear (Gerhart and Mclauchlan 2014
Battye et al 2017) Isotopic signatures are key to understand N dynamics in lakes
nevertheless in situ andor diagenetic fractionation along with multiple sources of
organic matter (OM) often hinder straightforward environmental interpretations from
isotopes Monitoring δ15N and δ13C values as components of the N cycle
specifically those related to the link between terrestrial and aquatic ecosystems can
help differentiate between effects from processes versus sources in stable isotope
values (eg from Particulate Organic Matter -POM- soil and vegetation) and
improve how we interpret variations in δ15N (and δ13C) values at longer temporal
scales
The main processes controlling stable N isotopes in bulk lake OM are source
lake productivity diagenesis and denitrification (Talbot 2001 Leng et al 2006
Fariacuteas et al 2009 Torres et al 2012 Brahney et al 2014) N sources depend on
contributions from the watershed (ie soil and biomass) the transfer of atmospheric
N to the lake (ie atmospheric N2 fixation) and nutrient recycling (Leng et al 2006)
Atmospheric N2 (isotopically defined as δ15N =0) is fixed by cyanobacteria with
minimal fractionation resulting in δ15NOM values that fluctuate from -2 to +2permil (Fogel
and Cifuentes 1993 Talbot 2001 Elliot and Brush 2006) Besides N2 fixation by
cyanobacteria phytoplankton species assimilate dissolved inorganic nitrogen (DIN)
and values typically range from +7 to +10permil (Xu et al 2016 Meyers et al 2003) In
addition seasonal changes in POM occur in the lake water column Gu et al (2006)
sampled the water column of Lake Wauberg (Florida USA) during a hydrologic year
and found a higher development of N fixing species during the summer A major
factor behind this increase are human activities in the watershed which control the
75
inputs of the sediment and nutrients to the lake (Woodward et al 2011) Some
studies have shown higher δ15N values in lake sediments from watersheds that are
highly affected by human activities (Bruland and Mackenzie 2010 Woodward et al
2011) Syntenic fertilizers displayed δ15N values between -4 to +4permil cattle manure
around +8 to +22permil and surface and groundwater OM between 0 to +10permil (Elliott
and Brush 2006 Leng et al 2006) Although relatively low δ15N values from
fertilizers constitute major N input to human-altered watersheds the elevated loss
of 14N via volatilization of ammonia and denitrification leaves the remaining total N
input enriched in 15N (Bruland and Mackenzie 2010)
In addition to the different sources and variations in lake productivity early
diagenesis at the sedimentndashwater interface in the sediment can further alter
sediment OM isotope values (Brahney et al 2014 Galman et al 2009) During
diagenetic loss of N in lacustrine systems kinetic fractionation occurs and the
remaining N is enriched in 15N leading to higher δ15N values (Galman et al 2006
Talbot 2001) Denitrification tends to enrich lake sediments in 15N due to the
assimilation of nitrates by denitrifying bacteria (Granger et al 2008) and is more
prevalent in anoxic environments (Brahney et al 2016 Lehman et al 2002)
Carbon isotopes in lake sediments can also provide useful information about
paleoenvironmental changes OM origin and depositional processes (Meyers et al
2003) Allochthonous organic sources (high CN ratios) produce isotope values
similar to values from catchment vegetation Autochthonous organic matter (low CN
ratio) is influenced by fractionation both in the lake and the watershed leading up to
carbon assimilation by lacustrine biota (Woodward et al 2011) Changes in
productivity greatly affect δ13COM values (Torres et al 2012) as during C uptake
plankton preferentially assimilate 12C leaving the dissolved inorganic carbon (DIC)
76
pool enriched in 13C (Gu et al 2006) with phytoplankton averaging ca 20 permil lower
than the respective δ13CDIC values (Leng et al 2006) Under conditions of low to
moderate primary productivity plankton preferentially uptake the lighter 12C
resulting in low δ13C values displayed by the OM (Torres et al 2012) Conversely
during high primary productivity phytoplankton will uptake 12C until its depletion and
is then forced to assimilate the heavier isotope resulting in an increase in δ13C
values Higher productivity in C-limited lakes due to slow water-atmosphere
exchange of CO2 also results in high δ13C values (Galman et al 2009) In these
cases algae are forced to uptake dissolved bicarbonate with δ13C values between
7 to 10permil higher than those of the atmospheric CO2 dissolved in water (Sun et al
2016 Torres et al 2012 Galman et al 2009)
Stable isotope analyses from lake sediments are thus useful tools to
reconstruct shifts in lake-watershed dynamics caused by changes in limnological
parameters and LUCC Our knowledge of the current processes that can affect
stable isotope signals in a watershed-lake system is limited however as monitoring
studies are scarce Besides in order to use stable isotope signatures to reconstruct
past environmental changes we require a multiproxy approach to understand the
role of the different variables in controlling these values Hence in this study we
carried out a detailed survey of current N dynamics in a coastal central Chilean lake
(Lago Vichuqueacuten) and a multiproxy study of a sediment record spanning the last
600 years The characterization of the recent changes in the watershed since 1970s
is based on satellite images to compare recent changes in the lake and assess how
these are related with climate variability and an ever increasing human footprint
(Lara et al 2012 Gallardo et al 2015 Steffen et al 2015) Our main goal is to
investigate how stable isotope values from lake sediment reflect changes in the lake
77
ndash watershed system during periods of high watershed disruption (eg Spanish
Conquest late XIX century Great Acceleration) and recent climate change (eg
Little Ice Age and current global warming)
2 Study Site
Lago Vichuqueacuten (34ordm49`S 72ordm04W 4 m asl 30 m of depth Fig 1) is a
mesotrophic to eutrophic lake with dominant anoxic bottom conditions that is
stratified year around (Frugone-Aacutelvarez et al 2017) The main inlets are the
Vichuqueacuten and Huintildee rivers and the only outlet is the Llico estuary that drains into
the Pacific Ocean High tides can sporadically shift the flow direction of the Llico
estuary which increases the marine influence in the lake Dune accretion gradually
limited ocean-lake connectivity until the estuary was almost completely closed off
by ca 10 ky BP and the current lake regime formed (Frugone-Aacutelvarez et al 2017)
The area is characterized by a mediterranean climate with cold-wet winters and
hot-dry summers and an annual precipitation of ~650 mm and a mean annual
temperature of 15ordmC During the austral winter months (June - August) precipitation
is modulated by north-west shifts of the South Pacific Anticyclone (SPA) followed by
an increased frequency of storm fronts stemming off the South Westerly Winds
(SWW) A strengthened SPA during austral summers (December - March) which
are typically dry and warm blocks the northward migration of storm tracks stemming
off the SWW (Fig1 Frugone-Aacutelvarez et al 2017)
78
Figure 1 a) The location of the Lago Vichuqueacuten in central Chile b) Map of land
uses and cover in 2016 c) Ombrothermic diagram mediterranean climates are
characterized by cold-wet winters with surplus moisture from June to August and
hot-dry summers d) Lake bathymetry showing location of cores and water sampling
sites used in this study
Although major land cover changes in the area have occurred since 1975 to the
present as the native forests were replaced by tree (Monterey pine and eucalyptus)
plantations the region was settled before the Spanish conquest (Frugone-Alvarez
et al 2018) Historical documents indicate that Vichuqueacuten was occupied by a
Mitimaes colony as part of the Inka empire (Odone 1998) Unlike other Andean
areas during the Inka period the introduction of corn and quinoa in the Vichuqueacuten
watershed do not seem to have intensified land use The Spanish colonial period in
Chile lasted from 1542 CE to the independence in 1810 CE The first historical
document (1550 CE) shows that the areas around Vichuqueacuten were settled by the
Spanish early on as the Vichuqueacuten watershed was given as an ldquoencomiendardquo
system to Juan Cuevas The ldquoencomiendardquo system provided the owners with land
79
and indigenous people to work but also the introduction of wheat wine cattle
grazing and logging of native forests for lumber extraction and increasing land for
agricultural activities (Odone et al 1998 Armesto et al 2010) During the 19th
century (the Republic) the export of wheat to Australia and Canada generated
intensive changes in land cover use The town of Vichuqueacuten became the regional
capital and plans to build a maritime port in the bay of Vichuqueacuten were drawn
However the fall of international markets in 1880 paralyzed these plans During the
20th century the plantation of trees (Pinus radiata and Eucalyptus globulus) in areas
cleared of native forests was subsidized by two national laws DFL nordm265 (1931) and
DFL nordm 701 (1974) both of which provided funds for such plantations During the last
decades the urbanization with summer vacation homes along the shorelines of
Lago Vichuqueacuten has greatly increased Extensive lake eutrophication is currently a
large environmental problem (EULA 2008)
3 Methods
Coring campaigns were organized from 2011 to 2015 (Fig 1d) We recovered
12 short cores and 4 long cores (up to 14 m long) at several sites using a hammer-
modified UWITEC gravity corer Sediment cores (VIC11-2A 170 cm VIC13-2B 170
cm VIC13-2D 1200 cm and VIC15-1A 119 cm) were split photographed sub-
sampled and stored in the Pyrenean Institute of Ecology (IPE-CSIC Spain) Core
VIC13-2B was selected for detailed multiproxy analyses (including elemental
geochemistry C and N isotopes XRF and 14C dating) Stable isotope analyses
(δ13C δ15N) were performed at the Laboratory of Biogeochemistry and Applied
Stable Isotopes (LABASI-PUC Chile) Sediment samples for δ13Corg were pre-
treated with 50 ml of HCl and 50 ml of deionized water and dried at 60ordmC for 4 hr to
remove carbonates (Harris et al 2011) Isotope analyses were conducted using a
80
Delta V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via
a Conflo IV interface Isotope results are expressed in standard delta notation (δ)
and expressed in per mil (permil) relative to the Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Total carbon (TC)
Total Organic Carbon (TOC) Total Inorganic Carbon (TIC) and Total Sulfur (TS)
were measured every 2 cm with a LECO SC 144 DR at the IPE-CSIC
An AVAATECH X-Ray Fluorescence II core scanner with a Rh X-ray tube from
the University of Barcelona was used to obtain XRF logs every 4 mm of resolution
Results are expressed as element intensities in counts per second (cps) Tube
voltage was operated at 30 kV and 10 kV to obtain the abundances of 18 elements
(Pb Al Si S Cl K Ca Ti V Mn Fe Co Zn Br Rb Sr Y Zr) with an average of
at least of 1800 cps (less for Pb=437 and Zn=934 cps) Pb was included due to
similar behavior with Co and Fe Element ratios were calculated to describe changes
in redox conditions (MnFe) endogenic productivity (BrTi) carbonate formation
(SrCa and CaTi) and sediment input (ZrTi) (Barreiro-Lostres et al 2014
Carrevedo et al 2015 Frugone-Aacutelvarez et al 2017 Kylander et al 2011 Moreno
et al 2007a)
Several campaigns were carried out to sample the POM from the water column
two per hydrologic year from November 2015 to August 2018 A liter of water was
recovered in three sites through to the lake two are from the shallower areas (with
samples taken at 2 and 5 m depth at each site) and one in the deeper central portion
(at 2 5 and 20 m depth) of the lake (Fig 1d) The samples were filtered using glass
fiber filters (25 m) and then frozen to obtain the stable carbon and nitrogen isotope
signal of lacustrine POM Additionally soil and vegetation samples from the
following communities native species meadow hydrophytic vegetation and
81
Monterrey pine (Pinus radiata) were taken for stable isotope analyses (see in
supplementary material)
The age model for the complete Lago Vichuqueacuten sedimentary sequence is
based on Frugone-Aacutelvarez et al (2017) with the last 1000 years based on
210Pb137Cs dating techniques in VIC15-1A and 14C AMS dates on bulk sediment
samples (Supplementary Table S1) The 14C measurements of lake water DIC show
a moderate reservoir effect of 180 plusmn 25 14C yr) The revised age-depth model used
here includes three more 14C AMS dates performed with the program Bacon to
establish the deposition rates along the core (Blaauw and Christen 2011 Fig 2)
The age-depth model indicates that average resolution between 0 to 87 cm is lt2
cm per year and from 88 to 170 cm it is lt47 cm per year
82
Figure 2 Chronological age-depth model for the Lago Vichuqueacuten sedimentary
sequence spanning the last 1000 yr (see also Frugone-Alvarez et al 2017)
To estimate land use changes in the watershed we use Landsat MSS images
for 1975 and Landsat TM for 1989 and 2014 all taken in either summer or autumn
(Table 1) We performed supervised classification of land uses (maximum likelihood
83
algorithm) for each year (1975 1989 and 2014) and results were mapped using
ArcGIS 102
Table 1 Images using for LUCC reconstruction
Source of LUCC
Acquisition
Date Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat TM 19991226 30 m
CONAF 2009 30 m
Land cover Chile 2014 30 m
CONAF 2016 30 m
Previous Work on Lago Vichuqueacuten sedimentary sequence
The sediments are organic-poor dark brown to brown laminated silt with some
intercalated thin coarser clastic layers Lacustrine facies have been classified
according to elemental composition (TOC TS TIC and TN) grain size and
sedimentary textures (Frugone-Aacutelvarez et al 2017) Four main types of lacustrine
facies were identified in this short core Facies L1 is a laminated (1cm) black to dark
brown diatom-poor silt with relatively higher TOC percentages (TOC about 185)
TS (about of 18) and TOCTN (about 92) Facies L2 is characterized by a
homogeneous black organic-poor silty clay with low TOC (mean= 13) TS (mean=
13) TOCTN (mean= 88) values Facies L3 is a laminated (1cm) black organic-
poor (TOC mean 14) silts with higher TS (mean= 21) and TOCTN ratios
(mean= 88) Facies L3 is interpreted as ldquobaselinerdquo deposition on the distal areas
84
of Vichuqueacuten lake Mineralogy is dominated by mica (muscovite) clay minerals
(chlorite) and plagioclase (albite) Quartz and calcite occur at the top and pyrite
occurs in the lower part of the sequence Facies T is composed by massive banded
sediments 4 cm thick and coarse grain size It is interpreted as an instantaneous
depositional event (for more detail see Frugone-Aacutelvarez et al 2017) In this work
we identified four subunits based on geochemical and stable isotope signals
4 Results
41 Geochemistry and PCA analysis
High positive correlations exist between Al Si K and Ti (r = 078 ndash 096
supplementary Fig S1) and between Zn Zr Rb and Y (r = 072 - 096) which reflect
the silicate content of the watershed-derived sediments (Marzecoacuteva et al 2011) Zr
is commonly associated with minerals more abundant in coarser deposits Thus the
ZrTi profile could reflect grain size (Cuven et al 2010) showing higher variability
in the upper part of the Lake Vichuqueacuten sequence and in the alternation between
facies L1 and L2 Another group of elements made up of Co Fe and Pb displayed
positive correlations (r = 067ndash 097) and represents the input of heavy metals Br
Cl and Ca show a positive correlation (r = 049 ndash 06 p-value = 00001) BrTi ratio
is interpreted as a productivity indicator due to Br having a strong affinity with humic
and organic compounds (Marzecoacuteva et al 2011 Frugone-Aacutelvarez et al 2017) In
our Lago Vichuqueacuten sequence BrTi values are high from 170 to 132 cm and from
36 cm to the top of core where it shows several sharps peaks (Fig 3) The MnFe
ratio is indicative of lake bottom oxygenation (Naecher et al 2013) as under
reducing conditions Mn tends to become more mobile than Fe leading to a
decreased MnFe (Marzecoacuteva et al 2011) Several sharp peaks in MnFe occurred
85
from 127 to 120 cm and from 57 to 30 cm (MgFegt03) The silicate group and the
Br Cl Ca Mn group are negatively correlated (r= -012 and -066)
Principal Component Analysis (PCA) was undertaken on the XRF
geochemical data to investigate the main factors controlling sediment deposition in
Lago Vichuqueacuten (Fig 3) The four main eigenvectors explain 801 of the variance
(supplementary material Table S2) The principal component (PC1) explain 437
of the total of variance and grouped elements are associated with terrigenous input
to the lake Positive values of the biplot have been attributed to higher heavy metals
deposition (Pb Co and Fe) and negative values to higher silicate input (Al K Ti and
Si) in a high energy catchment environment (Zr) The PC2 explains 198 of the
total of variance and highlights the endogenic productivity in the lake The positive
loading is compound by Br Cl Ca and Sr and to a lesser extent by Y Zn Rb and
Mn The PC2 may explain both the precipitation of carbonates (Ca Sr) and biological
production (Br)
86
Figure 3 Principal Component Analysis of XRF geochemical measurements in
VIC13-2B Lago Vichuqueacuten lake sediments
42 Sedimentary units
Based on geochemical and stable isotope analysis we identified four
lithological subunits in the short core sedimentary sequence Our PCA analyses and
Pearson correlations pointed out which variables were better for characterizing the
subunits (Fig 4) in terms of endogenic productivity heavy metals and terrestrial
input The unit 2c (170-131 cm) is characterized by the occurrence of some clastic
layers (L2 from 160 and 150 cm) high δsup1⁵N values (mean= +59 plusmn 04permil) with
Magnetic Susceptibility (MS) BrTi ZrTi and SrCa values increase towards the top
Also it has relatively low δsup1sup3C values (mean= -265 plusmn 11permil) and CNmolar ratios
(mean=78 plusmn 04) The ZrTi show upwards increasing values reaching the highest
values of the sequence at the top of this unit suggesting a coarsening upward trend
and relatively higher depositional energy The MS trend also indicates higher
erosion in the watershed and enhanced delivery of ferromagnetic minerals likely
from soil particles particularly from 150 to 160 cm (Frugone-Aacutelvarez at al 2017)
The subunit 2b (130-118 cm) is also composed of black silts but it has the
lowest MS values of the whole sequence and its onset is marked by a sharp
decrease in ZrTi This subunit is marked by sharp peaks of MnFe (at 126 and 120
cmgt027) SrCa (at 125 and 120 cmgt033) CaTi (at 124 and 120 cmgt199) TOC
(about 26 at 130 124 122 and 118 cm) and δsup1⁵N (gt+67permil at 126 and 120 cm)
BrTi oscillates around low values (about 01) whereas the δsup1sup3C values range
between -262 and -282permil
87
The unit 2a (58-117 cm) shows increasing and then decreasing MS values
and displayed sharps peaks of δsup1⁵N (gt64 at 102 to 99 87 and 75 cm) and CN
(about 10 at 86 74 66 and 60 cm) SrCa (mean = 04 plusmn 013) BrTi (mean = 008
plusmn 002) MnFe (mean = 002 plusmn 001) and δsup1sup3C (mean = -260 plusmn 07permil) oscillate in
low values The upper part of this unit (72 ndash 57 cm) shows an increase of SrCa
(from 03 to 05)
The upper unit 1a (0 - 57 cm) starts with a fine clastic layer (T at 57 to 54
cm) interpreted as deposition during a high-energy event It is characterized by
lower BrTi (mean = 003 plusmn 002) TOC (mean = 12 plusmn 03) and δsup1sup3C (mean= -
266 plusmn 06permil) higher δsup1⁵N (mean = +55 plusmn 08 permil) This unit is made of alternating
fine (lt 1cm) black and dark brown laminae with carbonate (L1 facies) Recently
deposited sediments show decreasing MS values increasing TOC (mean= 15 plusmn
04 ) sharp peaks of BrTi (gt015 at 33 21 5 and 1 cm) and the lowest δsup1⁵N values
of the entire record (mean= 45 plusmn 06 permil) and δsup1sup3C values (mean= -277 plusmn 09 permil)
Heavy metal content increases abruptly in the upper section of the core (2 - 20 cm)
(peaks of FeTi CoTi and PbTi)
88
Figure 4 Sedimentary units of Lago Vichuqueacuten sedimentary sequence Selected
variables illustrate watershed input (Magnetic Susceptibility ZrTi CoTi and MnFe)
endogenic productivity (SrCa CaTi TS) organic matter source (BrTi TOC
CNmolar and stable isotope records (δ13Corg and δ15Nbulk)
43 Recent seasonal changes of particulate organic matter on water column
The average of δ15NPOM from the three sites across to the lake was +86 plusmn 58
permil oscillating widely from -14 to +193permil (Fig 5) Important seasonal differences
occur at 2 and 5 m depth with more negative values during summer (~53 plusmn 52 permil)
than winter (~122 plusmn 40permil) At 20 m depth δ15NPOM values exhibit small seasonal
ranges (mean = +10permil for winter and +87permil for summer) The δ13CPOM average was
-296 plusmn 33permil with slightly seasonal and water column depth differences However
more positive δ13CPOM values occurred during winter (mean=-290 plusmn 41permil) than in
summer (mean= -301 plusmn 19 permil) Like the δ15NPOM values CNPOM (mean= 83 plusmn 17)
displayed important seasonal and water depth differences Lower CNPOM ratios
89
occurred during summer at 2 and 5 m depth (mean= 95 plusmn 17) and higher and more
constant values during winter (mean = 73 plusmn 09) at the same depth At 20 m CNPOM
shows similar values in both winter (70) and summer (74)
Figure 5 Seasonal POM averages from samples taken of the Lago Vichuqueacuten
water column Samples were taken from 2015 to 2018 at 2 (n=16) 5 (n=14) and 20
(n=8) meters depth
44 Stable isotope values across the Lake Vichuqueacuten watershed
Figure 6 shows modern vegetation soil and sediment isotope values found for
the watershed Huintildee tributary and Vichuqueacuten lake The δsup1⁵NFoliar values from
meadow plantations and macrophytes have similar range values with a mean of
+89 plusmn50 permil (n=18) +74 plusmn 75 permil (n=5) +82 plusmn 36 permil (n=6) respectively Native
vegetation has overall lower δsup1⁵NFoliar values with a mean of 14 plusmn 21 permil (n=28 see
Table 2 in supplementary material) The δsup1sup3CFoliar from terrestrial samples exhibit
similar values across the different plant communities (tree plantation mean=-274 plusmn
13permil meadow mean = -275 plusmn 23 permil native species mean=-272 plusmn 14permil) whereas
macrophytes display slightly more negative values with a mean of -287 plusmn 23permil
Macrophytes substrate displayed the highest δsup1⁵N values with a mean of +60 plusmn
14permil (n=3) Similar and relatively high δsup1⁵N values for soil of plantation (mean= +54
plusmn 13permil n=4) meadow (mean= +49 plusmn 04permil n=8) and surface river sediment
90
(mean= +52 plusmn 17permil n=10) Native forest soils oscillate around slightly more
negative values with a mean of +45 plusmn 12permil (n=4) Slightly more positive soil δsup1sup3C
values occur both underneath native forests and in tree plantations with means of -
284 plusmn 23permil and -298 plusmn 13permil respectively compared to either meadow soils
(mean= -302 plusmn 11permil) aquatic sediments underneath macrophytes (-311 plusmn 10permil)
or from surface river sediments (mean= -312 plusmn 10permil)
Figure 6 Stable isotope content of modern soil river sediment and foliar vegetation
used as end members in the sedimentary sequence of Lago Vichuqueacuten a)
Scatterplot of foliar δsup1⁵N and δsup13C values for vegetation from Lago Vichuqueacuten
watershed (plantation meadow and native species) and macrophytes on Lake
Vichuqueacuten See supplementary material for more detail of vegetation types b)
Scatterplot of soil δsup1⁵N and δsup13C values across the watershed the sediments of the
Huintildee and Vichuqueacuten rivers and the fluvial sediment corresponding to the
macrophyte vegetation
45 Land use and cover change from 1975 to 2014
Major land use changes between 1975 CE and 2016 CE in the Lago
Vichuqueacuten watershed are summarized in Figure 7 The watershed has a surface
area of 535329 km2 of which native vegetation (26) and shrublands (53)
represent 79 of the total surface in 1975 Meadows are confined to the valley and
91
represent 17 of watershed surface Tree plantations initially occupied 1 of the
watershed and were first located along the lake periphery By 1989 the areas of
native forests shrublands and meadows had decreased to 22 31 and 14
respectively whereas tree plantations had expanded to 30 These trends
continued almost invariably until 2016 when shrublands and meadows reached 17
and 5 of the total areas while tree plantations increased to 66 Native forests
had practically disappeared by 1989 and then increased up to 7 of the total area
in 2016 (Fig 6) preferentially occupying the southeastern sector of the watershed
Figure 7 Changes in land use and cover between 1975 and 2016 in the Lago
Vichuqueacuten watershed as measured from satellite images The major change is
represented by the replacement of native forest shrubland and meadows by
plantations of Monterrey pine (Pinus radiata)
Figure 8 shows correlations between lake sediment stable isotope values and
changes in the soil cover from 1975 to 2013 Positive relationships occurred
between sediment δsup1⁵N and δsup13C values from core VIC13-2B (unit 1) and the
92
percentage of native forest (r=089 for δsup1⁵N 082 for δsup13C) shrublands (r = 071 for
δsup1⁵N 057 for δsup13C) and meadow cover (r = 088 for δsup1⁵N 081 for δsup13C) All of these
correlations are significant (p value lt 0001) In contrast significant negative
correlations (p lt0001) occurred between tree plantation cover and lake sediment
stable isotope values (δsup1⁵N r = -082 and δsup13C r = -067) native forest (r = -097)
meadows (r = -086) and shrubland (r =-093)
Figure 8 Correlation plots of land use and cover change versus lake sediment
stable isotope values The δsup1⁵N values are positively correlated with native forests
agricultural fields and meadow cover across the watershed Total Plantation area
increases are negatively correlated with native forest meadow and shrubland total
area Significance levels are indicated by the symbols p-values (0 0001 001
005 01 1) lt=gt symbols ( )
93
5 Discussion
51 Seasonal variability of POM in the water column
The stable isotope values of POM can vary during the annual cycle due to
climate and biologic controls namely temperature and length of the photoperiod
which affect phytoplankton growth rates and isotope fractionation in the water
column (Gu et al 2006 Leng et al 2006) POM from Lago Vichuqueacuten surface
samples (2 and 5 m depth) displayed lower δ13CPOM values during the summer than
in winter During C uptake phytoplankton preferentially utilize 12C leaving the
DICpool enriched in 13C Therefore as temperature increases during the summer
phytoplankton growth generates OM enriched in 12C until this becomes depleted
and then the biomas come to enriched u At the onset of winter the DICpool is now
enriched in 13C and despite an overall decrease in phytoplankton production the
OM produced will also be enriched in 13C In contrast δ13CPOM values at 20 m depth
did not reflect these seasonal differences probably due to water-column
stratification that maintains similar temperatures and biological activity throughout
the year
Lake N availability depends on N sources including inputs from the
watershed and the atmosphere (ie deposition of N compounds and fixation of
atmospheric N2) which varies during the hydrologic year The fixation of atmospheric
N2 is an important natural source of N to the lake occurring mainly during the
summer season associated with higher temperature and light (Gu et al 2006)
Because the δ15NPOM of N2 fixers is close to 0permil with almost no overall isotope
fractionation (Leng et al 2006) when N2 is the major N source δ15NPOM values are
typically low However when DIN concentrations are high or alternatively when little
94
N2 fixation occurs δ15NPOM values will be higher (Gu et al 2006) δ15NPOM values
from summer Lago Vichuqueacuten samples were lower than those from winter with large
differences between the seasons (~5permil Fig 5 and 9) Furthermore δ15NPOM values
were high when monthly average temperature was low and monthly precipitation
was high (Fig 9) This pattern is consistent with the dominance of high N2 fixation
by cyanobacteria associated with increased summer temperatures This correlation
of δ15NPOM values with temperature further suggests a functional group shift i e
from N fixers to phytoplankton that uptake DIN The correlation between wetter
months and higher δ15NPOM values could be caused by increased N input from the
watershed due to increased runoff during the winter season The lack of data of the
δ15NPOM between the 30 mm and the 160 mm of precipitation is due to the
mediterranean-type climate that concentrates precipitations in the winter months
Finally Lago Vichuqueacuten is characterized by pronounced seasonality leading to
higher phytoplankton biomass in summer characterized by low δ15NPOM In winter
low biomass production and increased input from watershed is associated to high
δ15NPOM
95
Figure 9 Seasonal changes can drive δ15NPOM values in Lago Vichuqueacuten Data
correspond to average monthly temperature and total monthly precipitation for the
months when the water samples were taken (years 2015 - 2018) P-valuelt005
52 Stable isotope signatures in the Lake Vichuqueacuten watershed
The natural abundance of 15N14N isotopes of soil and vegetation samples
from the Lago Vichuqueacuten watershed appear to result from a combination of factors
isotope fractionation different N sources for plants and soil microorganisms (eg N2
fixation Nitrates) depth in the soil where N uptake by vegetation occurs and N loss
mechanisms (ie denitrification leaching and ammonia volatilization Hogberg
1997) The lowest δsup1⁵Nfoliar values are associated with native species and are
probably due to the contribution of N-fixing species (eg Sophora) (Fig6 and for
more detail see Table S3 in supplementary material) The number native N-fixers
species present in the Chilean mediterranean vegetation are not well known
however so there could be other N-fixing species in this group Alternatively δsup1⁵Nfoliar
values reflect soil N uptake (Kahmen et al 2008) In environments limited by N
plants have δsup1⁵N values similar to the soil δsup1⁵N values however the denitrification
and volatilization of ammonia can lead to the remain N of soil to come enriched in
15N (Cifuentes et al 1989 Craine et al 2009 Bruland and Mckenzie 2010) Soil N
isotope samples from native species communities tends to display relatively high
δsup1⁵N values respect to foliar samples due to loss of N-soil
The higher foliar and soil δsup1⁵N values obtained from samples of meadows
aquatic macrophytes and tree plantations can be attributed to the presence of
greater amounts of nitrate available for plant uptake (Fig 6) Houmlgberg (1997)
suggests that the availability of different N sources in soils (ie nitrates versus
96
ammonia) with different residence times can also explain these δsup1⁵NFoliar values
Indeed Feigin et al (1974) described differences of up to 20permil between ammonia
and nitrates sources Denitrification and nitrification discriminate much more against
15N than N2 fixation does (Craine et al 2009) leading to soils (and plants after
uptake) enriched in 14N
In general multiple processes that affect the isotopic signal result in similar
δsup1⁵N values between the soil of the watershed and the sediments of the river
However POM isotope fluctuations allow to say that more negative δsup1⁵N values are
associated to lake productivity while more positive δsup1⁵N values are associated with
N input from the watershed
δ13C values in Lago Vichuqueacuten watershed reflect CO2 gas exchange between
C3 plants and algae with the atmosphere During photosynthesis plants
discriminate against the heavier stable isotope of carbon (13C) in favor of the lighter
isotope 12C which results in δ13CFoliar values between -220permil and -320permil (Browman
and Cook 2002 Liu et al 2007) Likewise the δ13CFoliar values from Lago Vichuqueacuten
oscillate around -27permil (Fig 6) Plants transform atmospheric CO2 into soil organic
carbon (C) which in turn reflects this initial discrimination against 13C during C
uptake and post photosynthetic fractionation (Farquhar et al 1982 1989 Badeck
et al 2005 Pausch and Kuzyakov 2017) Slightly more positive δ13CFoliar values
(about 15permil) were measured in comparison with their δ13CSoil values This may be
reflecting the C transference from plants to the soil but also a soil-atmosphere
interchange The preferential assimilation of the light isotopes (12C) during soil
respiration carried by the roots and the microbial biomass that is associated with the
decomposition of litter roots and soil organic matter explain this differential
(Berhardt et al 2006 Blagodatskaya et al 2010 Midwood and Millard 2011)
97
In general the δ13CSoil values from the Lago Vichuqueacuten watershed oscillated
around -290permil and did not vary with our plant classification types Here we use
these values as terrestrial-end members to track changes in source OM (Fig 6)
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from the terrestrial watershed By the other hand more positive δ13C
values most likely reflect an increased aquatic OM component as indicated by POM
isotope fluctuations (Fig 9)
53 Recently land use and cover change and its influences on N inputs to the lake
Tree plantations in the Lago Vichuqueacuten watershed have increased greatly in
the last few decades (from 1 in 1975 to 66 in 2016 Fig 7) replacing previous
native forests (from 26 to 7) meadows (17 to 5) and shrublands (53 to
17) In 1975 tree plantations were confined to the lake perimeter with discrete
patches distributed throughout the watershed The ldquoForest Lawrdquo (DFL 701- passed
in 1974) allocated state funding to afforestation efforts and management of tree
plantations which greatly favored the replacement native forests by introduced trees
This increase is marked by a sharp and steady decrease in lake sediment δ15N and
δ13C values because tree plantations function as a nutrient sink whereas other land
uses such as meadows favor nutrient loss from the watershed (Fig8) Bruland and
Mackenzie (2014) noted a decrease in wetland δ15N values when watershed
forested cover increased and concluded that N inputs to the wetlands are lower from
the forested areas as they generally do not export as much N as agricultural lands
A positive correlation between native vegetation and δ15Ncore values can be
explained by the relatively scarce arboreal cover in the watershed in 1975 when
native forest occupied just 26 of the watershed surface whereas shrublands and
98
meadows occupied more than the 70 of the surface of the watershed with the
concomitant elevated loss of N (Fig 7 and Fig 8)
54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years
Sedimentological compositional and geochemical indicators all show changes in
the N cycle associated to LUCC lake dynamics and climate changes (Fig 10) From
the pre-Columbian indigenous settlement including the Spanish colonial period up
to the start of the Republic (1300 - 1800 CE) the introduction of crops such as
quinoa and wheat but also the clearing of land for extensive agriculture would have
favored the entry of N into the lake Conversely major changes observed during the
last century were characterized by a sharp decrease of N input that were coeval
with the increase of tree plantations
From 1365 until 1550 CE (unit 2c 2b and half of 2a Fig 3 and Fig 10) pre-
Columbian agricultural societies planted mostly crops of corn and quinoa (Ramirez
and Vidal 1985) This period is characterized by large peaks seen in the δsup1⁵N record
(eg 1403 1452 1476 and 1505) with overall high δsup1⁵N values (up to 5permil) indicating
that N input from watershed was elevated and oscillating to the beat of the NT These
positive δsup1⁵N peaks could be due to several causes including a) the clearing of land
for farming b) N loss via denitrification which would be generally augmented in
anoxic environments (Diebel et al 2009) such as Lago Vichuqueacuten (low MnFe
values about 0001 Fig 4) or c) climate especially cold-wet winters or hot-dry
summers can also exert control on the δsup1⁵N record Indeed tree-ring records and
summer temperature reconstructions show overall wetcold conditions during this
period (Christie et al 2009 Von Gunten et al 2009 Fig 10) Increased
precipitation would bring more sediment (and nutrients) from the watershed into the
99
lake and increase lake productivity which is also detected by the geochemical
proxies for productivity (peaks of BrTi SrCa and TOC Fig 4 and Fig 10) (see also
Frugone-Alvarez et al 2017)
Figure 10 Changes in the N availability during the last six centuries in Lago
Vichuqueacuten a) lake sediment δsup1⁵N values displayed more positive values during the
prehistoric period Spanish Colony and the starting 19th century which is associated
with enhanced N input from the watershed by extensive clearing and crop
plantations The inset shows this relationship between sediment δsup1⁵N and
100
percentage of meadow cover over the last 30 years b) Summer temperature
reconstruction from central Chile (von Gunten et al 2009) showing a
correspondence with cold phases and high δsup1⁵N values at Vichuqueacuten except for the
last 100 years c) Palmer Drought Severity Index (PDSI) is an interannual moisture
variability reconstruction for late springndashearly summer during the last six centuries
(Christie et al 2009) Grey shadow indicating higher precipitation periods
From 1550 and 1810 CE (unit 2a) and during the Colonial period several peaks
of δ15N could be further related not only to high precipitation (Fig 10 grey shadow)
but also pulses of enhanced N input from the watershed linked to human land use
In 1550 CE Juan Cuevas was granted lands and indigenous workers under the
encomienda system for agricultural and mining development of the Vichuqueacuten
village (Ramiacuterez and Vidal 1985) Historical documents show that around 1579 CE
the Vichuqueacuten watershed was occupied by indigenous communities dedicated to
wheat plantations and vineyards wood extraction and gold mining (Odone 1998)
The introduction of the Spanish agricultural system implied not just a change in the
types of crops used (from quinoa to vineyards and wheat) but also a clearing of
native species for the continuous increase of agricultural surface and wood
extraction During the Colonial period wheat from Vichuqueacuten was exported to Peru
(Ramiacuterez and Vidal 1985) Armesto et al (2009) indicate that during the XVIII and
XIX centuries the extraction of wood for mining operations was important enough to
cause extensive loss of native forests The independence and instauration of the
Chilean Republic did not change this prevailing system Increases in the
contributions of N to the lake during the second half of the XIX century (peaks in
δsup1⁵N ca 1870 CE) could be due to expanding farm land dedicated for wheat
101
production and increased commercial trade with California and Canada (Ramiacuterez
and Vidal 1985)
In contrast LUCC in the last century are clearly related to the development of
large-scale tree plantations (Fig 7) The δsup1⁵N record reaches the lowest values of
the entire sequence in the last few decades (Fig 10) A marked increase in lake
productivity NT concentration and decreasing sediment input is synchronous (unit
1 Fig 4) with trees replacing meadows shrublands and areas with native forests
(Fig 7 and 8) The implementation of the lsquoForest Lawrsquo in 1974 had a stronger impact
on the landscape and lake ecosystem dynamics than the impacts of ongoing climate
change in the region which is much more recent (Garreaud et al 2018) although
the prevalence of hot dry summers seen over the last decade would also be
associated with low δsup1⁵N values and increase of NT (Fig 10) Heavy metals ratios
(FeTi CoTi and PbTi) show notable increases in lake sediments from 1992 to 2011
CE (Fig 4) Although this could be related to mining in the El Maule region the
closest mines are 60 Km away (Pencahue and Romeral) so local factors related to
shoreline urbanization for the summer homes and an increase in tourist activity
could also be a major factor
6 Conclusions
The N isotope signal in the watershed depends on the rates of exchange
between vegetation and soil cover (Fig 11) Plants preferentially uptake 14N and the
underlying soils become enriched in 15N especially when the terrestrial ecosystem
is N-limited andor significant N loss occurs (ie denitrification andor ammonia
volatilization) LUCC in the Lago Vichuqueacuten watershed have heavily modified the
links between terrestrial and aquatic ecosystems with agriculture practices
102
contributing more N to the lake than tree plantations or native forests In situ lake
processes can also fractionate N isotopes An increase of N-fixing species results
in OM depleted in 15N which results in POM with lower δsup1⁵N values during these
periods During winter phytoplankton is typically enriched in 15N due to the
decreased abundance of N-fixing species and increased N input from the
watershed This in turn generates the highest δsup1⁵NPOM values in Lago Vichuqueacuten
Periods of elevated productivity tend to deplete the dissolved inorganic N of 14N
resulting in even higher δ15N values
Our study illustrates the usefulness of δsup1⁵N as a tool for reconstructing the past
influence of LUCC on N availability in lake ecosystems To constrain the relative
roles of the diverse forcing mechanisms that can alter N cycling in mediterranean
ecosystems all main components of the N cycle should be monitored seasonally
(or monthly) including the measurements of δ15N values in land samples
(vegetation-soil) as well as POM
103
Figure 11 Summary of human and environmental factors controlling the δ15N
values of lake sediments Particulate organic matter(POM) δ15N values in
mediterranean lakes are driven by N input from the watershed that in turn depend
on land use and cover changes (ie forest plantation agriculture) andor seasonal
changes in N sources andor lake ecosystem processes (ie bioproductivity redox
condition denitrification) Arrows indicate the direction of N fluxes (inputoutput from
the N cycle) N cycle processes that deplete lake sediments of 15N are shown in
blue whereas those that enrich sediments in 15N are shown in red
104
Supplementary material
Figure S1 Pearson correlate coefficient between geochemical variables in core
VIC13-2B Positive and large correlations are in blue whereas negative and small
correlations are in red (p valuelt0001)
Figure S2 Principal Component Analysis of geochemical elements from core
VIC13-2B
105
Table S1 Lago Vichuqueacuten radiocarbon samples
RADIOCARBON
LAB CODE
SAMPLE
CODE
DEPTH
(m)
MATERIAL
DATED
14C AGE ERROR
D-AMS 029287
VIC13-2B-
1 043 Bulk 1520 24
D-AMS 029285
VIC13-2B-
2 085 Bulk 1700 22
D-AMS 029286
VIC13-2B-
2 124 Bulk 1100 29
Poz-63883 Chill-2D-1 191 Bulk 945 30
D-AMS 001133
VIC11-2A-
2 201 Bulk 1150 44
Poz-63884
Chill-2D-
1U 299 Bulk 1935 30
Poz-64089
VIC13-2D-
2U 463 Bulk 1845 30
Poz-64090
VIC13-2A-
3U 469 Bulk 1830 35
D-AMS 010068
VIC13-2D-
4U 667 Bulk 2831 25
Poz-63886
VIC13-2D-
4U 719 Bulk 3375 35
106
D-AMS 010069
VIC13-2D-
5U 775 Bulk 3143 27
Poz-64088
VIC13-2D-
5U 807 Bulk 3835 35
D-AMS-010066
VIC13-2D-
7U 1075 Bulk 6174 31
Poz-63885
VIC13-2D-
7U 1197 Bulk 6440 40
Poz-5782 VIC13-15 DIC 180 25
Table S2 Loadings of the trace chemical elements used in the PCA
Elementos PC1 PC2 PC3 PC4
Zr 0922 0025 -0108 -0007
Zn 0913 -0124 -0212 0001
Rb 0898 -0057 -0228 0016
K 0843 0459 0108 0113
Ti 0827 0497 0060 -0029
Al 0806 0467 0080 0107
Si 0803 0474 0133 0136
Y 0784 -0293 -0174 0262
V 0766 0455 0090 -0057
Br 0422 -0716 -0045 0226
Ca 0316 -0429 0577 0489
Sr 0164 -0420 0342 -0182
Cl 0151 -0781 -0397 0162
107
Mn -0121 -0091 0859 0095
S -0174 -0179 -0051 0714
Pb -0349 0414 -0282 0500
Fe -0700 0584 -0023 0280
Co -0704 0564 -0107 0250
Table S3 Stable Isotope values from vegetation of Lago Vichuqueacutenacutes watershed
Taxa Classification δsup1⁵N δsup1sup3C CN
molar
Poaceae Meadow 1216 -2589 3602
Juncacea Meadow 1404 -2450 3855
Cyperaceae Meadow 1031 -2596 1711
Taraxacum
officinale Meadow 836 -2400 2035
Poaceae Meadow 660 -2779 1583
Poaceae Meadow 453 -2813 1401
Poaceae Meadow 966 -2908 4010
Juncus Meadow 1247 -2418 3892
Poaceae Meadow 747 -3177 6992
Poaceae Meadow 942 -2764 3147
Poaceae Meadow 1479 -2634 2895
Poaceae Meadow 1113 -2776 1795
Poaceae Meadow 2215 -2737 7971
Poaceae Meadow 1121 -2944 2934
Poaceae Meadow 638 -3206 1529
108
Macrophytes Macrophytes 886 -3044 2286
Macrophytes Macrophytes 1056 -2720 2673
Macrophytes Macrophytes 769 -3297 1249
Macrophytes Macrophytes 967 -2763 1442
Macrophytes Macrophytes 959 -2670 2105
Macrophytes Macrophytes 334 -2728 1038
Acacia dealbata
Introduced
species 656 -2696 1296
Acacia dealbata
Introduced
species 487 -2941 1782
Acacia dealbata
Introduced
species 220 -2611 3888
Luma apiculata Native species 433 -2542 4135
Luma apiculata Native species 171 -2664 7634
Luma apiculata Native species -001 -2736 6283
Luma apiculata Native species 029 -2764 6425
Azara sp Native species 159 -2868 8408
Azara sp Native species 101 -2606 2885
Baccharis concava Native species 104 -2699 5779
Baccharis concava Native species 265 -2488 4325
Baccharis concava Native species 287 -2562 7802
Baccharis concava Native species 427 -2781 5204
Baccharis linearis Native species 190 -2610 4414
Baccharis linearis Native species 023 -2825 5647
109
Peumus boldus Native species 042 -2969 6327
Peumus boldus Native species 205 -2746 4110
Peumus boldus Native species 183 -2743 6293
Chusquea quila Native species 482 -2801 4275
Poaceae meadow 217 -2629 7214
Lobelia sp Native species 224 -2645 3963
Lobelia sp Native species -091 -2565 4538
Aristotelia chilensis Native species -035 -2785 5247
Aristotelia chilensis Native species -305 -2889 2305
Aristotelia chilensis Native species 093 -2836 5457
Chusquea quila Native species 173 -2754 3534
Chusquea quila Native species 045 -2950 6739
Quillaja saponaria Native species 223 -2838 9385
Scirpus meadow 018 -2820 7115
Sophora sp Native species -184 -2481 2094
Sophora sp Native species -181 -2717 1721
Pinus radiata
Introduced
trees 1581 -2602 3679
Pinus radiata
Introduced
trees 1431 -2784 4852
Pinus radiata
Introduced
trees -091 -2708 9760
Pinus radiata
Introduced
trees 153 -2568 3470
110
Salix sp
Introduced
trees 632 -2878 1921
LITERATURE CITED
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A historical
framework for land cover change in southwestern South America in the past 15000
years Land use policy 27 148ndash160
httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the next
carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014
httpsdoiorg101002eft2235
Blaauw M Christeny JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474
httpsdoiorg10121411-BA618
Bowman DMJS Cook GD 2002 Can stable carbon isotopes (δ13C) in soil
carbon be used to describe the dynamics of Eucalyptus savanna-rainforest
boundaries in the Australian monsoon tropics Austral Ecol
httpsdoiorg101046j1442-9993200201158x
Brahney J Ballantyne AP Turner BL Spaulding SA Otu M Neff JC 2014
Separating the influences of diagenesis productivity and anthropogenic nitrogen
deposition on sedimentary δ15N variations Org Geochem 75 140ndash150
httpsdoiorg101016jorggeochem201407003
111
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409
httpsdoiorg102134jeq20090005
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego R
Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and
environmental change from a high Andean lake Laguna del Maule central Chile
(36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the
Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from
tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Craine JM Elmore AJ Aidar MPM Dawson TE Hobbie EA Kahmen A
Mack MC Kendra K Michelsen A Nardoto GB Pardo LH Pentildeuelas J
Reich B Schuur EAG Stock WD Templer PH Virginia RA Welker JM
Wright IJ 2009 Global patterns of foliar nitrogen isotopes and their relationships
with cliamte mycorrhizal fungi foliar nutrient concentrations and nitrogen
availability New Phytol 183 980ndash992 httpsdoiorg101111j1469-
8137200902917x
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty
Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-
010-9453-1
Diebel M Vander Zanden MJ 2012 Nitrogen stable isotopes in streams stable
isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19
1127ndash1134 httpsdoiorg10189008-03271
112
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in freshwater
wetlands record long-term changes in watershed nitrogen source and land use SO
- Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash
2916
Fariacuteas L Castro-Gonzales M Cornejo M Charpentier J Faundez J
Boontanon N Yoshida N 2009 Denitrification and nitrous oxide cycling within the
upper oxycline of the oxygen minimum zone off the eastern tropical South Pacific
Limnol Oceanogr 54 132ndash144
Farquhar GD Ehleringer JR Hubick KT 1989 Carbon Isotope Discrimination
and Photosynthesis Annu Rev Plant Physiol Plant Mol Biol
httpsdoiorg101146annurevpp40060189002443
Farquhar GD OrsquoLeary MH Berry JA 1982 On the relationship between
carbon isotope discrimination and the intercellular carbon dioxide concentration in
leaves Aust J Plant Physiol httpsdoiorg101071PP9820121
Fogel ML Cifuentes LA 1993 Isotope fractionation during primary production
Org Geochem httpsdoiorg101007978-1-4615-2890-6_3
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A
Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-
resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS)
implications for past sea level and environmental variability J Quat Sci 32 830ndash
844 httpsdoiorg101002jqs2936
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924
httpsdoiorg104319lo20095430917
113
Gerhart LM McLauchlan KK 2014 Reconstructing terrestrial nutrient cycling
using stable nitrogen isotopes in wood Biogeochemistry 120 1ndash21
httpsdoiorg101007s10533-014-9988-8
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and oxygen
isotope fractionation during dissimilatory nitrate reduction by denitrifying bacteria 53
2533ndash2545 httpsdoiorg10230740058342
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater
eutrophic lake Limnol Oceanogr 51 2837ndash2848
httpsdoiorg104319lo20065162837
Harris D Horwaacuteth WR van Kessel C 2001 Acid fumigation of soils to remove
carbonates prior to total organic carbon or CARBON-13 isotopic analysis Soil Sci
Soc Am J 65 1853 httpsdoiorg102136sssaj20011853
Houmlgberg P 1997 Tansley review no 95 natural abundance in soil-plant systems
New Phytol httpsdoiorg101046j1469-8137199700808x
Kylander ME Ampel L Wohlfarth B Veres D 2011 High-resolution X-ray
fluorescence core scanning analysis of Les Echets (France) sedimentary sequence
New insights from chemical proxies J Quat Sci 26 109ndash117
httpsdoiorg101002jqs1438
Lara A Solari ME Prieto MDR Pentildea MP 2012 Reconstruccioacuten de la
cobertura de la vegetacioacuten y uso del suelo hacia 1550 y sus cambios a 2007 en la
ecorregioacuten de los bosques valdivianos lluviosos de Chile (35o - 43o 30acute S) Bosque
(Valdivia) 33 03ndash04 httpsdoiorg104067s0717-92002012000100002
Lehmann M Bernasconi S Barbieri A McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during
114
simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66
3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007
Liu G Li M An L 2007 The Environmental Significance Of Stable Carbon
Isotope Composition Of Modern Plant Leaves In The Northern Tibetan Plateau
China Arctic Antarct Alp Res httpsdoiorg1016571523-0430(07-505)[liu-
g]20co2
Marzecovaacute A Mikomaumlgi A Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54
httpsdoiorg103176eco2011105
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash
1643 httpsdoiorg1011770959683613496289
Meyers PA 2003 Application of organic geochemistry to paleolimnological
reconstruction a summary of examples from the Laurention Great Lakes Org
Geochem 34 261ndash289 httpsdoiorg101016S0146-6380(02)00168-7
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland
Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Odone C 1998 El Pueblo de Indios de Vichuqueacuten siglos XVI y XVII Rev Hist
Indiacutegena 3 19ndash67
Pausch J Kuzyakov Y 2018 Carbon input by roots into the soil Quantification of
rhizodeposition from root to ecosystem scale Glob Chang Biol
httpsdoiorg101111gcb13850
115
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98
httpsdoiorg1011772053019614564785
Sun W Zhang E Jones RT Liu E Shen J 2016 Biogeochemical processes
and response to climate change recorded in the isotopes of lacustrine organic
matter southeastern Qinghai-Tibetan Plateau China Palaeogeogr Palaeoclimatol
Palaeoecol httpsdoiorg101016jpalaeo201604013
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of
different trophic status J Paleolimnol 47 693ndash706
httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl
httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M 2009 High-resolution quantitative climate
reconstruction over the past 1000 years and pollution history derived from lake
sediments in Central Chile Philos Fak PhD 246
Woodward C Chang J Zawadzki A Shulmeister J Haworth R Collecutt S
Jacobsen G 2011 Evidence against early nineteenth century major European
induced environmental impacts by illegal settlers in the New England Tablelands
south eastern Australia Quat Sci Rev 30 3743ndash3747
httpsdoiorg101016jquascirev201110014
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K Yeager
KM 2016 Different responses of sedimentary δ15N to climatic changes and
116
anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau
J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
117
DISCUSION GENERAL
El Nitroacutegeno (N) es un elemento clave que controla la dinaacutemica y
funcionamiento de ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al
1997) Pese a que el N (N2) es muy abundante en la atmoacutesfera (78) es escaso
en los ecosistemas pues la mayoriacutea de la biota no puede acceder a su forma
molecular (Galloway et al 1995 Zaehle et al 2013) En este sentido la entrada
natural de N en los ecosistemas es viacutea fijacioacuten bioloacutegica del N2 por bacterias que lo
convierten en compuestos bioloacutegicamente disponibles (Battye et al 2017) Debido
a que la fijacioacuten es un proceso energeacuteticamente costoso los ecosistemas
comuacutenmente estaacuten limitados por N (Vitousek and Howarth 1991) Los LUCC
contribuyen al incremento del N disponible y son una de las principales causas de
eutroficacioacuten y contaminacioacuten de los cuerpos de agua (Woodward et al 2012)
En Chile central los LUCC principalmente relacionados con las actividades
agriacutecolas y forestales han causado grandes cambios en el ciclo del N debido al
118
reemplazo de la cobertura natural por un monocultivo y al uso de fertilizantes que
modifican los aportes de MO y N a los cuerpos de agua El programa de
estabilizacioacuten de laderas impulsada por el gobierno (Albert 1900) y la ley forestal
de 1974 han fomentado la forestacioacuten con plantaciones de Pinus radiata y
Eucalyptus globulus y en el caso de Lago Vichuqueacuten estas actualmente cubren maacutes
del 60 de su cuenca (Cap2 Fig 5) Ademaacutes en Laguna Matanzas la
sobreexplotacioacuten del agua en conjunto con el cambio climaacutetico actual el que ha
conducido a una de las megasequiacuteas maacutes severas en las uacuteltimas deacutecadas
(Garreaud et al 2017) generoacute una combinacioacuten letal que secoacute el lago en tan solo
10 antildeos (Cap1 Fig1) Las evidencias obtenidas a partir de estos dos lagos
permiten identificar las huellas del Antropoceno en Chile central basadas en el
registro sedimentario lacustre
La transferencia de N desde los ecosistemas terrestres a acuaacuteticos es un
proceso natural con controles bioacuteticos climaacuteticos y geoloacutegicos El enlace
hidroloacutegico entre los ecosistemas terrestres y acuaacuteticos hace que el estado troacutefico
de los lagos esteacute vinculado al de su cuenca (McLauchlan et al 2013) En Chile
central se sabe poco del impacto de los LUCC sobre el ciclo de nutrientes en los
ecosistemas lacustres aunque al menos desde la colonizacioacuten espantildeola se tienen
registros de influencia humana en las cuencas Durante la colonia espantildeola
Laguna Matanzas fue una hacienda ganadera que exportaba productos caacuternicos al
Peruacute (Contreras-Lopez et al 2014) A su vez en el Lago Vichuqueacuten se realizaban
extensas actividades agriacutecolas hasta comienzos del s XX como por ejemplo
cultivos de vid y trigo ademaacutes de extraccioacuten de madera y mineriacutea de oro (Odone
1997) En las uacuteltimas deacutecadas los LUCC han estado vinculados principalmente con
el desarrollo forestal al compaacutes de una estrategia gubernamental de fomentar con
119
incentivos financieros (Ley de Bosques DFL nordm265 de 1931 y DFL 701 de 1974)
esta actividad El incremento de la superficie forestal es especialmente fuerte en
ambos lagos a partir de la deacutecada de 1980 y hasta 2016 (Laguna de Matanzas 0-
17 Lago Vichuqueacuten 1-66) Este aumento habriacutea sido en detrimento del bosque
nativo y los pastizales (Cap 1 Fig 5 y Cap 2 Fig 8) El incremento de la superficie
forestal generariacutea una disminucioacuten de los aportes de sedimentos y nutrientes al lago
y en este sentido un cambio de estado en los flujos de N (e g tipping points) que
a su vez ha quedado registrado en la sentildeal isotoacutepica (δ15N) y la acumulacioacuten de
MO en los sedimentos lacustres
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central
Los sedimentos lacustres son verdaderos ldquosumiderosrdquo que pueden llegar a
registrar lo que ocurre en cuanto a la historia ambiental de su cuenca En esta tesis
se utilizan los isoacutetopos estables de N (15N y 14N) en sedimentos lacustres para
reconstruir los cambios en la disponibilidad de N en el tiempo y establecer la
magnitud de impacto generado por actividades humanas El fraccionamiento
cineacutetico en los lagos es mediado por procesos bioloacutegicos que mediante la
asimilacioacuten de N genera un producto (eg fitoplancton) maacutes ldquoligerordquo (valores maacutes
bajos de 15N) que su remanente (eg DIN) (Fogel y Cifuentes 1993) Sin embargo
en periacuteodos en que los lagos estaacuten limitados por N el fitoplancton discrimina menos
y la MO resultante queda enriquecida en 15N (Fig 1 a y b) Ademaacutes la
desnitrificacioacuten acentuada en lagos de fondo anoacutexicos (eg Laguna Matanzas
entre 1800 y 1940 Cap1 Fig 6) conduce a un enriquecimiento en 15N de los
sedimentos y de la columna de agua Esta informacioacuten queda reflejada en la MO
120
de los sedimentos lacustres lo que permite conocer la disponibilidad del N en el
tiempo a partir de las variaciones de 15N
En esta tesis analizamos la MO de los sedimentos lacustres para reconstruir
la influencia de los LUCC en los aportes de sedimentos y N al lago En teacuterminos de
asimilacioacuten de N se puede distinguir entre dos grupos principales de productores
primarios que componen el POM (Fig1)
1 Los que asimilan el DIN compuesto por amonio nitratos y nitritos Aquiacute el
δ15N resultante fluctuaraacute en torno a valores maacutes positivos en la medida que
la productividad se incrementa o no hay reposicioacuten del DIN (Fig 1)
2 Los fijadores de N Dado que es un proceso energeacuteticamente costoso en
ambientes que no estaacuten limitados por N muchas veces son excluiacutedas
competitivamente por el resto del fitoplancton Si el DIN queda agotado por
el incremento de la productividad (o bien si el ambiente estaacute limitado ya sea
por N o por otros nutrientes) los fijadores de N pueden florecer y la MO que
se genera a partir de ello tendraacute un δ15N con valores en torno a 0permil
De este modo la MO en los sedimentos lacustres dependeraacute de la
composicioacuten de productores primarios (ie fijadores vs asimiladores del DIN)
ademaacutes de los aportes desde la cuenca tanto de MO como del N de la cuenca que
pasa a formar parte del DIN (eg fertilizantes estieacutercol) (Fig 1)
La MO de los lagos estudiados en esta tesis ha sido analizada a partir de
variables geoquiacutemicas isotoacutepicas frotis y estaacute compuesta principalmente por
diatomeas y MO amorfa A partir de los datos obtenidos en esta tesis los valores
de δ15N aumentan cuando hay una mayor cobertura agriacutecola o pastizales A su vez
tiende a disminuir cuando aumenta la cobertura arboacuterea independiente si esta es
por plantaciones forestales o por bosque nativo
121
Por otra parte la dinaacutemica del lago tambieacuten imprime caracteriacutesticas
especiales en el POM observaacutendose variaciones estacionales en los valores
δ15NPOM Durante la estacioacuten caacutelida y seca se observan valores maacutes bajos que
durante la estacioacuten friacutea y huacutemeda posiblemente relacionado con el incremento de
la contribucioacuten de especies fijadoras de N (Lago Vichuqueacuten Fig 9 Cap 2) durante
el verano En la estacioacuten friacutea y huacutemeda el DIN seriacutea la principal fuente de N Las
mayores entradas de MO y N terrestre debidos a un incremento del lavado de la
cuenca como consecuencia de las precipitaciones y la mineralizacioacuten de la MO
podriacutean estimular la produccioacuten de MO lacustre por crecimiento del fitoplancton
Como consecuencia se observan tendencias decrecientes de los valores de
δ15N en la MO sedimentaria cuando la productividad en el lago estaacute relacionada
con especies fijadoras de N mientras que el δ15N muestra aumentos cuando la
productividad del lago estaacute asociada principalmente al consumo del DIN pero
tambieacuten cuando predominan procesos de perdida de N (eg desnitrificacioacuten Fig
1)
Hemos identificado tres fases en la dinaacutemica del δ15N para ambos lagos
Entre 1800 y 1940 CE la cuenca de laguna de Matanzas estaba dominada por
actividades agriacutecolas y ganaderas (δ15Nsuelo gt 4permil Cap 1 Fig 4 y 6) Las entradas
de sedimentos y MO desde la cuenca son altas (δ13C lt -209permil ZrTi ~029 AlTi
~013) y coincide con un clima maacutes huacutemedo y friacuteo (Von Gunten et al 2009
Christiansen et al 2011) El lago teniacutea un nivel de agua maacutes profundo dominado
por condiciones anoacutexicas (MnFe ~0013) que contribuiriacutean a mayores valores de
δ15N en la columna de agua y por ende de los sedimentos acumulados en el fondo
debido a la peacuterdida diferencial de 14N viacutea desnitrificacioacuten (Fig 7 Cap1) La baja
122
produccioacuten de MO lacustre (BrTi ~002 TOC~17) coincide con un bajo contenido
de NT (~478 μg) y valores maacutes positivos de δ15N (gt 4permil)
La actividad agriacutecola tambieacuten dominaba en la cuenca del Lago Vichuqueacuten
durante este periacuteodo (δ15Nsuelo gt4permil Cap2 Fig 6) El aporte sedimentario desde la
cuenca es alto (AlTi ~029 ZrTi ~032) y fue favorecido por mayores
precipitaciones a las actuales (Christiansen et al 2011) El Lago Vichuqueacuten es un
lago estratificado anoacutexico (MnFe ~002) y profundo (~30 m) por lo que la
desnitrificacioacuten juega un rol importante en el enriquecimiento de la MO
sedimentaria La baja produccioacuten de MO (BrTi ~007 TOC ~12) de origen
lacustre (δ13C ~ -268permil) coincide con un bajo contenido de NT (~309μg +6) y
valores maacutes positivos de δ15N (56permil +03)
Durante esta fase en ambos lagos los aportes de N de la cuenca parecen
ser los principales responsables en las fluctuaciones del δ15N Esta sentildeal podriacutea
estar amplificada por bajas temperaturas (que afectan la productividad lacustre) y
altas precipitaciones que favorecen el lavado de la cuenca y aumentan el aporte de
sedimentos y MO desde la cuenca predominantemente agriacutecola
Entre 1940 y 1980 CE en Laguna Matanzas se observa un incremento en
la acumulacioacuten de MO algal (TOC ~2 +1 δ13C ~-230+09permil) El ambiente
deposicional es ligeramente menos turbulento que el anterior (AlTi ~011+001
ZrTi ~03+001) y el lago estaacute en una fase de transicioacuten hacia condiciones maacutes
oacutexicas (MnFe ~002+0004) y maacutes productivas (BrTi ~004+0007) A su vez δ15N
tiende hacia valores maacutes negativos (δ15N ~30permil+03) mientras que el NT aumenta
oscilando en antifase con el δ15N
En Lago Vichuqueacuten en cambio se observa un ligero incremento en la
acumulacioacuten de MO (TOC ~13 +02) de origen terrestre (δ13C ~-27permil +04) La
123
productividad es similar al periacuteodo anterior (BrTi ~007+003) y el ambiente
deposicional es menos turbulento (AlTi ~02 +009 Zrti ~033 +007) El δ15N y el
NT oscilan en torno a valores ligeramente maacutes bajos (δ15N ~51permil +04 NT ~292μg
+6) Este periacuteodo se caracteriza por ser maacutes caacutelido que el anterior lo que
posibilitariacutea un incremento en la productividad lacustre en Laguna Matanzas pero
que no es observada en el Lago Vichuqueacuten
Entre 1980 y la actualidad en Laguna Matanzas habriacutea un incremento de la
acumulacioacuten de MO (TOC ~59 + 3) asociada a un incremento en la productividad
del lago (BrTi ~009+005) y a los aportes de MO terrestres (δ13C ~-24 + 2permil) El
lago se vuelve maacutes oacutexico (Mn ~0003 +0006) y las entradas de sedimento
disminuyen (AlTi~ 009 +002) El δ15N tiende a valores maacutes negativos (δ15N ~13permil
+1) y el NT aumenta de 20 a 55 μg (NT ~346 + 9 μg) tomando valores sin
precedentes en todo el registro En los sedimentos lacustres del Lago Vichuqueacuten
tambieacuten se observa un incremento de la acumulacioacuten de MO (TOC ~17 + 05)
asociada a un incremento en la productividad del lago (BrTi ~01 + 003) y las
entradas de sedimento desde la cuenca disminuyen (AlTi ~005 + 005) El δ15N
(~43 + 05permil) tiende a valores maacutes negativos y el NT incrementa de 20 a 55 μg (NT
~346 + 9 μg) oscilando en antifase
Durante esta fase en ambos lagos se observa un aumento en la
acumulacioacuten de MO sedimentaria un incremento en el NT y valores maacutes negativos
de δ15N que coincide con el incremento de la superficie forestal de las cuencas
(Laguna Matanzas gt17 Lago Vichuqueacuten gt60)
124
Figura 2 Comparacioacuten de los cambios del ciclo del N entre Laguna Matanzas y
Lago Vichuqueacuten con los procesos biogeoquiacutemicos contrastantes en rojo En L
Matanzas las condiciones oacutexicas podriacutean estar favoreciendo la nitrificacioacuten del
amonio En Vichuqueacuten las condiciones anoacutexicas favorecen un aumento en δ15N de
la MO en los sedimentos por desnitrificacioacuten y mineralizacioacuten
Los ambientes mediterraacuteneos en el que los lagos del presente estudio se
encuentran insertos estaacuten caracterizados por una estacioacuten seca prolongada y las
precipitaciones ocurren en eventos puntuales alcanzando altos montos
pluviomeacutetricos en poco tiempo Las lluvias favorecen un lavado de la cuenca la
perdida de N y otros nutrientes Por lo que es esperable que los sedimentos del
lago reflejen mejor los valores isotoacutepicos de los suelos de la cuenca durante los
periacuteodos con altos montos pluviomeacutetricos En el Lago Vichuqueacuten por ejemplo el
125
POM superficial (2 y 5 m de profundidad) despliega valores isotoacutepicos maacutes
positivos en invierno presumiblemente como resultado de mayores aportes de MO
y N de su cuenca (Fig 1b) En ambos lagos los valores maacutes positivos en los
sedimentos lacustres ocurren durante periacuteodos con mayores montos pluviomeacutetricos
(Cap1 Fig 6 y Cap 2 Fig 12)
Ademaacutes del efecto de la precipitacioacuten en el aporte de nutrientes al lago en
esta tesis hemos encontrado que los mayores aportes de N desde la cuenca se dan
cuando la cobertura vegetal del suelo es menor En Laguna Matanzas el desarrollo
de la actividad ganadera desde la llegada de los espantildeoles a la cuenca habriacutea
incrementado los aportes de N al lago Los valores de δ15N en los sedimentos
lacustres depositados durante este periacuteodo son los maacutes positivos de todo el registro
(~4permil) A su vez en Lago Vichuqueacuten los valores isotoacutepicos maacutes positivos se
registran entre 1300 y 1900 CE que coinciden con el mayor desarrollo de
actividades agriacutecola y de extraccioacuten de maderas de la cuenca (Fig 10 Cap 2)
Los recientes LUCC (a partir de 1975) en las cuencas de Laguna Matanzas
y Lago Vichuqueacuten estaacuten caracterizados por un incremento de la superficie forestal
y agriacutecola en reemplazo de la cobertura de pastizal (Laguna Matanzas) y bosque
nativo (Lago Vichuqueacuten) Los valores de δ15N en los testigos sedimentarios fueron
maacutes positivos cuanto mayor la superficie agriacutecola de la cuenca y maacutes negativos
cuanto mayor la superficie forestal Aunque por los alcances de esta tesis no
podemos ser concluyentes respecto del origen del NT (aloacutectonoautoacutectono) parece
ser que esta maacutes ligado a los aportes de la cuenca pese a la disminucioacuten del aporte
sedimentario observado en ambos lagos Las plantaciones forestales a diferencia
del bosque nativo son fertilizadas con una mezcla de N P y K (Toral et al 2005)
Por lo que pese a la disminucioacuten del aporte sedimentario la concentracioacuten de
126
nutrientes en el aporte al lago puede verse incrementada por el tipo de manejo
forestal con respecto al bosque nativo
Los resultados del primer capiacutetulo demuestran que 1) las plantaciones
forestales yo bosque nativo retiene maacutes sedimentos y MO mientras que el uso de
suelo agriacutecola y los pastizales destinados al desarrollo de la actividad ganadera lo
libera 2) Al parecer la desnitrificacioacuten es el proceso natural maacutes importante de
perdida de N en lagos costeros y favorece el enriquecimiento de 15N tanto en la
columna de agua (ie 15N-DIN) como en los sedimentos una vez enterrados y 3) la
desnitrificacioacuten depende de los cambios en las condiciones REDOX en el lago La
oxigenacioacuten en lagos poco profundos -como Laguna Matanzas- es altamente
fluctuante a escalas decadal-centenal depende de la profundidad de la laacutemina de
agua y de la variabilidad de la precipitacioacuten En este sentido en Laguna Matanzas
habriacutea condiciones anoacutexicas en ambientes cuando el lago alcanza niveles maacutes
altos Estas condiciones se observan entre 1820 y 1940 CE y son coetaacuteneas con
episodios maacutes huacutemedos y friacuteos (von Gunten et al 2009 Christie et al 2009) pero
tambieacuten con una fuerte actividad ganadera en la cuenca
Las fuentes variables de N y su fluctuacioacuten en el tiempo hacen necesario
contar con un anaacutelogo moderno que permite usar al δ15N de los sedimentos
lacustres como un indicador indirecto de los cambios en la disponibilidad de N en
el tiempo Por ello en el segundo capiacutetulo integramos muestras de suelo-
vegetacioacuten y un monitoreo de POM de la columna de agua en verano e invierno La
composicioacuten isotoacutepica de POM estariacutea reflejando cambios de la fuente de N (fijacioacuten
vs consumo del DIN) que variacutea durante el antildeo hidroloacutegico Durante el verano la
mayor temperatura y horas-luz favoreceriacutean las entradas de N al lago viacutea fijacioacuten
bioloacutegica de N2 atmosfeacuterico resultando en un incremento de la MO con un valor
127
isotoacutepico bajo (POM verano~5permil Fig 1b) El N2 (δ15N = 0permil) es fijado praacutecticamente
sin fraccionamiento isotoacutepico generando un δ15N de la OM cercano a 0permil (Leng et
al 2006) En invierno las precipitaciones pueden estar favoreciendo un incremento
en la entrada de nutrientes al lago El DIN de la columna de agua llega a contener
valores de δ15N maacutes altos reflejando estos mayores aportes de la cuenca y el POM
del Lago Vichuqueacuten queda enriquecido en 15N (δ15N ~11permil Cap 2 Fig 9) Estas
variaciones estacionales pueden estar evidenciando un cambio en la composicioacuten
de especies de POM desde especies fijadoras a especies que consumen el N de la
columna de agua Sin embargo para corroborar esta informacioacuten seriacutea deseable
contar con el anaacutelisis de la composicioacuten de especies de las muestras de agua
extraidas
Entre los resultados maacutes importantes estaacuten los valores isotoacutepicos del suelo
y la biomasa representativa de la cuenca que incluye un listado de las especies
nativas de la zona que hasta ahora no se contaba (Cap 2 Tabla en material
suplementario) Las especies colectadas despliegan valores de δ15Nfoliar maacutes
positivos (plantaciones forestales y pastizal ~89permil) que el suelo (35permil) salvo por
las especies nativas las que oscilan en torno a valores cercanos al atmosfeacuterico
(~14permil) La razoacuten isotoacutepica 15N14N refleja la dinaacutemica de intercambio de N entre la
vegetacioacuten y el suelo (Koba et al 2002) La discriminacioacuten es positiva en la mayoriacutea
de los sistemas bioloacutegicos por lo que las plantas tienen valores de δ15N maacutes bajos
que el suelo (Evans et al 2001) como sucede con las especies nativas en Lago
Vichuqueacuten En consecuencia las plantas quedan empobrecidas en 15N y conducen
a un enriquecimiento en 15N del suelo sobretodo si no hay reposicioacuten de N por otras
viacuteas (eg fijacioacuten de N) Por lo tanto los valores maacutes negativos de δsup1⁵Nfoliar de las
especies nativas pueden estar relacionados con el consumo preferencial de 14N del
128
suelo donde la discriminacioacuten isotoacutepica de la vegetacioacuten contra el 15N conduce a
valores del δsup1⁵Nsuelo maacutes altos (Houmlgberg 1997) Por el contrario los valores maacutes
positivos de δsup1⁵Nfoliar de las plantaciones forestales y pastos con respecto al suelo
puede deberse por una parte que el suelo no cuenta con mecanismos naturales de
reposicioacuten de N salvo por la descomposicioacuten de la hojarasca que puede ser maacutes
lenta en Pinus radiata que en bosque nativo (Lusk et al 2001) y por otra por el alto
impacto de los aportes de N (y otros nutrientes) derivado de las actividades
humanas (eg uso de fertilizantes) en el suelo
El alcance maacutes significativo de esta tesis se relaciona con un cambio en la
tendencia maacutes negativa de δsup1⁵N y el incremento del NT a partir de 1940 y a partir
de 1970 CE en ambos lagos La causa maacutes probable de esta tendencia es el
reemplazo de la cobertura natural o preexistente a 1940 CE por plantaciones
forestales
En la figura 2 se observa una siacutentesis de los principales procesos que
afectan la sentildeal isotoacutepica de la MO en los sedimentos lacustres de L Matanzas y
L Vichuqueacuten La entrada de N y MO desde la cuenca depende del tipo de cobertura
Las plantaciones forestales y el bosque nativo retienen maacutes nutrientes y sedimentos
en la cuenca mientras que la agricultura los favorece Sin embargo las praacutecticas
de manejo que incluyen la aplicacioacuten de N (y otros nutrientes) pueden aportar maacutes
nutrientes al lago que la cobertra de bosque nativo Cuando las actividades
forestales cubren una mayor superficie de la cuenca (1940 en adelante) δsup1⁵N oscila
en valores maacutes negativos y el NT tiende a incrementarse en los sedimentos
lacustres Cabe destacar que los valores de δsup1⁵N observados en los testigos
sedimentarios a fines del s XX no tienen precedente durante la conquista y colonia
espantildeola o durante el resto del periodo de la Repuacuteblica
129
Figura 2 Modelo de transferencia de N entre los ecosistemas terrestres y
acuaacuteticos El δ15N de los sedimentos lacustres depende de la interaccioacuten entre los
aportes de la cuenca y porcesos ldquoin-lakerdquo Flechas indican la direccioacuten del flujo de
N En azul las salidasentradas de 14N en rojo las salidasentradas de 15N δ15N de
la interaccioacuten plantas-suelo depende del tipo de cobertura de suelo
130
CONCLUSIONES GENERALES
La transferencia de N entre cuencas y lagos es un factor de control del ciclo
del N en los lagos y queda reflejado en la sentildeal isotoacutepica de los sedimentos
lacustres (Leng et al 2006 McLauchlan et al 2007) En Laguna Matanzas el
suelo de las especies nativas y las plantaciones forestales despliegan valores de
δ15N maacutes bajos (~1permil) que los de Lago Vichuqueacuten (~35permil) Coincidentemente los
sedimentos lacustres de Laguna Matanzas oscilan con valores de δ15N maacutes bajos
(-15 a +45permil Fig 1) que los de Lago Vichuqueacuten (+3 a +8permil)
Por otra parte el estudio de la MO de los sedimentos lacustres ha permitido
reconstruir los cambios en los aportes de MO y N desde la cuenca Aunque no es
posible afirmar queacute tipo de N estaacute entrando al lago (eg Nitroacutegeno orgaacutenico e
inorganico) si podemos decir que los cambios en las fluctuaciones de δ15N son
coincentes con el crecimiento de la actividad forestal (1980 - hasta 2016 L
Matanzas 0-17 y L Vichuqueacuten 1-66) El δ15N fluctuacutea hacia valores maacutes
negativos Ademaacutes se observa un incremento del NT en los sedimentos lacustres
cuanto mayor es la superficie forestal
Entre 1800-1940 las cuencas eran fundamentalmente agriacutecolas y
ganaderas (Cap1 Fig 7 y Cap 2 Fig 12) el δ15N de los sedimentos lacustres
oscila con valores maacutes positivos (L Matanzas +40 plusmn 05permil y L Vichuqueacuten 56 plusmn
033permil Fig 1) hay una baja bioproductividad en el lago (BrTi ~002 TOC~17)
lo que coincide con un bajo contenido de NT (~478 μg) Este es un periacuteodo de altas
precipitaciones (Christie et al 2009) que tienden a favorecer el lavado de la cuenca
131
y con ello el aporte de MO sedimentos y nutrientes desde la cuenca tiende a verse
favorecido Aunque las principales actividades humanas en estas cuencas son
diferentes (ganaderiacutea en Laguna Matanzas Contreras-Lopez et al 2014
agricultura en Lago Vichuqueacuten Odone 1997) en ambos casos hubo un reemplazo
de la cobertura natural de bosques nativo que favorecioacute mayores aportes de N y
sedimentos desde la cuenca en un efecto sumado con el aumento de las
precipitaciones
A partir de 1980 CE en ambos lagos se observa una disminucioacuten en los
valores de δ15N y un incremento del NT que no tiene precedentes en todo el registro
y pese a que ambos lagos son limnologicamente muy diferentes En Lago
Vichuqueacuten el δ15N tiende a oscilar en antifase con el NT (Fig 1) En el caso de
Laguna Matanzas se observa una tendencia hacia valores maacutes negativos y a partir
de 1990s a valores maacutes positivos mientras que el NT tiende a incrementarse de
manera sostenida Ello podriacutea estar relacionado con el crecimiento de la actividad
forestal habriacutea una mayor retencioacuten de N y sedimentos en la cuenca ligado al
incremento de la superficie forestal Ademaacutes en el caso de L Matanzas el
incremento de la actividad agriacutecola (sumado al de las plantaciones forestales)
podriacutea expicar el cambio en la tendencia en la deacutecada de los 1990s
En el contexto de Antropoceno esta tesis nos permite identificar un gran
impacto de las actividades humanas en Chile central a partir de la deacutecada de 1940
y una Gran Aceleracoacuten a partir de la deacutecada de 1970s En el registro sedimentario
de los lagos en estudio se observa una gran acumulacioacuten de MO el δ15N oscila
hacia valores maacutes negativos y un gran incremento del NT y el crecimiento de la
actividad forestal parece ser la principal responsable Ademaacutes la sobreexplotacioacuten
132
del agua junto al cambio climaacutetico global ha generado una combinacioacuten letal para
los lagos costeros de Chile central
Figura 1 Comparacioacuten de las fluctuaciones de δ15N durante los uacuteltimos 300
antildeos en Laguna Matanzas y Lago Vichuqueacuten
133
Referencias
Battye W Aneja VP Schlesinger WH 2017 Earthrsquos Future Is nitrogen the next carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia
UC de V 2014 Elementos de la historia natural del An Mus Hist Natulas Vaplaraiso 27 51ndash67
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Evans RD Evans RD 2001 Physiological mechanisms influencing
plant nitrogen isotope composition 6 121ndash126 Fogel ML Cifuentes LA 1993 Isotope fractionation during primary
production Org Geochem 73ndash98 httpsdoiorg101007978-1-4615-2890-6_3 Galloway JN Schlesinger WH Ii HL Schnoor L Tg N 1995
Nitrogen fixation Anthropogenic enhancement-environmental response reactive N Global Biogeochem Cycles 9 235ndash252
Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J
Christie D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Koba K Koyama LA Kohzu A Nadelhoffer KJ 2003 Natural 15 N
Abundance of Plants Coniferous Forest httpsdoiorg101007s10021-002-0132-6
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos
Transactions American Geophysical Union httpsdoiorg1010292007EO070007 McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE
2007 Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
134
Phytologist N Oct N 2006 Tansley Review No 95 15N Natural Abundance in Soil-Plant Systems Peter Hogberg 137 179ndash203
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Turner BL Kasperson RE Matson PA McCarthy JJ Corell RW
Christensen L Eckley N Kasperson JX Luers A Martello ML Polsky C Pulsipher A Schiller A 2003 A framework for vulnerability analysis in sustainability science Proc Natl Acad Sci U S A httpsdoiorg101073pnas1231335100
Vitousek PM Aber JD Howarth RW Likens GE Matson PA
Schindler DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
Vitousek PM Howarth RW 1991 Nitrogen limitation on land and in the
sea How can it occur Biogeochemistry httpsdoiorg101007BF00002772 von Gunten L Grosjean M Rein B Urrutia R Appleby P 2009 A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 Holocene 19 873ndash881 httpsdoiorg1011770959683609336573
Xu R Tian H Pan S Dangal SRS Chen J Chang J Lu Y Maria
Skiba U Tubiello FN Zhang B 2019 Increased nitrogen enrichment and shifted patterns in the worldrsquos grassland 1860-2016 Earth Syst Sci Data httpsdoiorg105194essd-11-175-2019
Zaehle S 2013 Terrestrial nitrogen-carbon cycle interactions at the global
scale Philos Trans R Soc B Biol Sci 368 httpsdoiorg101098rstb20130125
8
ABREVIATURAS
N Nitroacutegeno (Nitrogen)
DIN Nitroacutegeno Inorgaacutenico Disuelto (Dissolved Inorganic Nitrogen)
C Carbono (Carbon)
TOC Carbono Inorgaacutenico Total (Total Organic Carbon)
TIC Carbono Inorgaacutenico Total (Total inorganic Carbon)
TC Carbono Total (Total Carbon)
TS Azufre Total (Total Sulfur)
LUCC Cambios de UsoCobertura del Suelo (Land Use and Cover Change)
OM Materia Orgaacutenica (Organic Matter)
POM Particulate Organic Matter (materia orgaacutenica particulada)
CE Common Era
BCE Before Common Era
Cal BP Calibrado en antildeos radiocarbono antes de 1950
ie id est (esto es)
e g Exempli gratia (por ejemplo)
9
RESUMEN
El ldquoAntropocenordquo se caracteriza por pertubaciones de origen humano que
conllevan impactos globales Por ejemplo los cambios de usocobertura del suelo
(LUCC) son conocidos por perturbar el ciclo del N a partir de la Revolucioacuten Industrial
pero especialmente desde la Gran Aceleracioacuten (1950 CE) en adelante Sin
embargo existen incertezas asociadas a la magnitud del impacto y su efecto
acoplado con los cambios climaacuteticos actuales (ie megasequiacutea disminucioacuten de las
precipitaciones) y acoplamientos con otros ciclos biogeoquiacutemicos (eg ciclo del
Carbono) Este impacto ha cambiado la disponibilidad de N en los ecosistemas
terrestres y acuaacuteticos y sus enlaces Los sedimentos lacustres contienen
informacioacuten de las condiciones paleoambientales del lago y su cuenca en el
momento en que se depositaron y el anaacutelisis de los isoacutetopos estables de N (δ15N)
en los sedimentos permiten reconstruir los cambios en la disponibilidad de N a
traveacutes del tiempo En este trabajo usamos una aproximacioacuten multiproxy que incluye
10
anaacutelisis sedimentoloacutegicos geoquiacutemicos e isotoacutepicos (δ15N δ13C) de los sedimentos
lacustres columna de agua y suelo-vegetacioacuten de la cuenca y la reconstruccioacuten de
los LUCC entre 1975 y 2016 a partir de imaacutegenes satelitales El objetivo de esta
tesis fue evaluar el rol de LUCC como principal control del ciclo del N en el sistema
cuenca-lago durante las uacuteltimas centurias en Chile central Nuestros principales
resultados muestran que valores maacutes positivos de δ15N en los sedimentos lacustres
estaacuten relacionados con grandes aportes de N de la cuenca que a su vez son
mayores cuanto mayor la superficie agriacutecola yo los pastizales mientras que tanto
las plantaciones forestales como los bosques favorecen la retencioacuten de nutrientes
en la cuenca (lo que se ve reflejado en un δ15N maacutes negativo) Esta tesis plantea
un cambio de estado en la dinaacutemica del N asociado a la introduccioacuten y expansioacuten
de las plantaciones forestales o bosques los que retienen el Nitroacutegeno en las
cuencas mientras que el clima juega un rol secundario
11
ABSTRACT
The Anthropocene is characterized by human disturbances at the global
scale For example changes in land use are known to disturb the N cycle since the
industrial revolution but especially since the Great Acceleration (1950 CE) onwards
This impact has changed N availability in both terrestrial and aquatic ecosystems
However there are some important uncertainties associated with the extent of this
impact and how it is coupled to ongoing climate change (ie megadroughts rainfall
variability and shifting stationarity) or to other nutrient cycles (e g carbon cycle)
Lake sediments contain paleoenvironmental information regarding the conditions of
the watershed and associated lakes and which the respective sediments are
deposited Nitrogen stable isotope analysis (δ15N) on lake sediments allows us to
reconstruct the changes in N availability through time Here we used a multiproxy
approach that uses sedimentological geochemical and isotopic analyses on
lacustrine sediments water column and soilvegetation from the watershed as well
12
as land use and cover change (LUCC) from 1975 to 2016 obtained from satellite
images The goal of this thesis is to evaluate the role of LUCC as the main driver for
N cycling in a coastal watershed system of central Chile over the last centuries Our
main results show that more positive δ15N values in lake sediments are related to
higher N contributions from the watershed which in turn increase with increased
agricultural andor pasture cover whereas either forest plantations or native forests
can favor nutrient retention in the watershed (δ15N more negative) This thesis
proposes that N dynamics are mainly driven by the introduction and expansion of
forest or tree plantations that retain nitrogen in the watershed whereas climate plays
a secondary role
13
INTRODUCCIOacuteN
El N es un elemento esencial para la vida y limita la productividad en
ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al 1997) Las actividades
humanas han tenido un profundo impacto sobre el ciclo del N global principalmente
a partir la revolucioacuten industrial (IPCC 2013 2007) Las concentraciones de N se
han incrementado por la fijacioacuten industrial de N atmosfeacuterico (viacutea el proceso Haber-
Bosch Erisman et al 2008) por el uso de fertilizantes sobre la base de N para
mejorar el rendimiento de los cultivos la produccioacuten ganadera y ademaacutes de los
cambios en el uso del suelo (abreviado en ingleacutes- LUCC) (Robertson y Vitousek
2009 Galloway et al 2008 Battye et al 2017 Xu et al 2019) Estas actividades
contribuyen significativamente al incremento de la disponibilidad bioloacutegica de N
cuyas consecuencias para los ecosistemas incluye la perdida de diversidad
modificacioacuten de la disponibilidad de nutrientes y acidificacioacuten de los suelos entre
otras Sin embargo se desconoce la cantidad destino y el alcance que ha tenido
14
el N movilizado entre los ecosistemas generado por la influencia de las actividades
humanas (Vitousek et al 1997)
La entrada natural de N en los ecosistemas terrestres y acuaacuteticos es viacutea
fijacioacuten atmosfeacuterica la que es llevada a cabo por microorganismos muchos de ellos
en relaciones simbioacuteticas (eg Rhizobium en leguminosas) y las algas (Vitousek et
al 1997 Battye et al 2017) Las principales salidas estaacuten relacionadas con la
desnitrificacioacuten volatilizacioacuten del amoniaco y por escurrimiento superficial y
subterraacuteneo (Galloway et al 2008 Diacuteaz et al 2016) En el caso de los sistemas
lacustres ademaacutes estaacuten la resuspencioacuten de N del fondo del lago los aportes desde
la cuenca (entradas de N) La depositacioacuten de MO en el fondo del lago que marca
la salida de N de la columna de agua Estas relaciones de intercambio de N tienen
un control geograacutefico (eg latitud exposicioacuten relieve etc) climaacutetico y biotico
(McLauchlan et al 2013) La alteracioacuten provocada por los LUCC no solo genera
las peacuterdidas de nutrientes del suelo por erosioacuten y escurrimiento superficial sino que
tambieacuten modifica la disponibilidad y transferencia de N entre los ecosistemas
terrestres y acuaacuteticos (Elbert et al 2012 McLauchlan et al 2013) Por ejemplo el
reemplazo de la vegetacioacuten nativa por la agricultura yo plantaciones forestales
altera el ciclo del N debido al despeje de los terrenos viacutea quema de la vegetacioacuten
pero tambieacuten debido al reemplazo de la vegetacioacuten preexistente por un
monocultivo (eg trigo pino) y el uso ineficiente de fertilizantes para mejorar el
rendimiento agriacutecola Estas actividades favorecen un incremento de los aportes de
N (y otros nutrientes) viacutea sistemas de drenajes a lagos los que actuacutean como
sumideros El incremento del N derivado de las actividades humanas tanto en los
ecosistemas terrestres como acuaacuteticos genera incertezas relacionados con la
trayectoria y la magnitud de la disponibilidad de N en ambos ecosistemas (Batye et
15
al 2017 Mclauchlan et al 2017) Por lo tanto se requiere de una liacutenea base de
largo plazo para saber coacutemo estas perturbaciones impactan la disponibilidad de N
en las uacuteltimas deacutecadassiglos en los ecosistemas lacustres y por lo tanto el alcance
real que los LUCC han tenido en el ciclo del N
Los ecosistemas mediterraacuteneos y el ciclo del N
Los ecosistemas mediterraacuteneos son especialmente sensibles a los LUCC
pues estos se encuentran sometidos a un estreacutes hiacutedrico durante largas temporadas
estivales y las precipitaciones se concentran en eventos puntuales y a veces con
altos montos pluviomeacutetricos Las precipitaciones tienen un alto potencial de arrastre
de sedimentos y nutrientes los que actuacutean como fertilizantes al llegar a los
ecosistemas lacustres A su vez un incremento en la disponibilidad de N puede
generar un desbalance con otros nutrientes (eg Fosforo) con efectos sobre la
productividad y diversidad de los lagos (Pentildeuelas et al 2012 Sardans et al 2012
McLauchlan 2013) En este sentido en Chile central se observa una disminucioacuten
de las precipitaciones especialmente fuerte en las uacuteltimas deacutecadas por lo que se ha
denoacuteminado como Megasequiacutea (Garreaud et al 2017) y el efecto de las
precipitaciones esporaacutedicas pueden generar un arrastre de nutrientes que no ha
sido evaluado
Los ecosistemas mediterraacuteneos concentran el 20 de la biodiversidad global
(Myers et al 2000) pero existe una escasez de conocimiento respecto a los
efectos del incremento de N en los cuerpos de agua como consecuencia de las
actividades humanas en el pasado histoacuterico (Ochoa-Hueso et al 2011) y coacutemo la
disponibilidad de N ha variado en el tiempo Los LUCC han aumentado el flujo de
N en las cuencas hidrograacuteficas viacutea escurrimiento superficial y lixiviacioacuten
16
favoreciendo la peacuterdida de sedimentos y nutrientes del suelo en el mundo entero
(Elser et al 2011 McLauchlan et al 2013 Xu et al 2017 y 2019) Ello ha
contribuido a la acidificacioacuten y eutrofizacioacuten generalizada de riacuteos y lagos
(McLauchlan et al 2013 Schindler et al 2008)
El ecosistema mediterraacuteneo de Chile central es una regioacuten ampliamente
intervenida desde al menos la colonizacioacuten espantildeola (1600-1800 CE) Los LUCC
han tenido efectos negativos en la disponibilidad de agua especialmente
observados con el desarrollo de las actividades minera (Prieto et al 2019) Aunque
se sabe de los impactos en los ecosistemas de bosque en teacuterminos de cobertura
debidos al desarrollo silvo-agropecuarias (Armesto et al 2010) se desconoce el
impacto de estas actividades sobre el ciclo del N en los cuerpos de agua o en queacute
momento cobran fuerza suficiente para alterar el flujo de N hacia los lagos de Chile
Central Por lo tanto en esta tesis nos centramos en entender coacutemo los LUCC han
afectado la disponibilidad de N en los lagos costeros de Laguna Matanzas y Lago
Vichuqueacuten desde la llegada de los espantildeoles a Chile hasta el presente
Los lagos como sensores ambientales
Los sedimentos lacustres son buenos sensores de cambios en los aportes
de nutrientes contaminacioacuten y eutroficacioacuten de los cuerpos de agua ya que son
capaces de preservar sentildeales quiacutemicas y fiacutesicas al momento de su formacioacuten y
ademaacutes con alta resolucioacuten y a largo plazo (Leng et al 2006) Por lo tanto
constituyen una herramienta uacutetil para reconstruir el flujo de N entre ecosistemas
terrestres y acuaacuteticos en el pasado sus consecuencias (eg incremento de la
productividad lacustre) y el rol del clima y los LUCC en el pasado (McLauchlan et
al 2013 von Gunten et al 2009) Por ejemplo el cultivo de maiacutez por parte de los
17
nativos iroqueses en Canadaacute provocoacute la eutrofizacioacuten del Lago Crawford (43degN)
durante los siglos XII y XV Entre las consecuencias registradas se documentoacute un
claro incremento de la productividad primaria y cambios en la estructura comunitaria
de diatomeas foacutesiles (incremento de Stephanodiscus complex y disminucioacuten de
Cyclotella bodanica en Ekdahl et al 2007) Otro ejemplo demuestra como las
actividades agriacutecolas en la cuenca del riacuteo Mississippi modificaron los patrones de
sedimentacioacuten y aporte de nutrientes al lago Pepin EE UU (44degN) a partir del
asentamiento europeo en 1840 CE y principalmente desde 1940 CE (Engstrom et
al 2009) Para Chile von Gunten et al (2009) a partir de indicadores
limnogeoloacutegicos evidencioacute la contaminacioacuten y eutroficacioacuten de cinco lagos cercanos
a la ciudad de Santiago (33ordmS) como consecuencia de la deposicioacuten atmosfeacuterica
de metales pesados (especialmente Cu) quema de combustible foacutesil y flujo de
nutrientes durante los uacuteltimos 200 antildeos
Caracteriacutesticas limnoloacutegicas de los lagos
Los procesos limnoloacutegicos afectan la distribucioacuten y abundancia de los
organismos en los lagos Estaacuten influenciados por forzamientos externos por
ejemplo la precipitacioacuten o la penetracioacuten de la luz y la morfometriacutea del lago En este
sentido el nivel del lago dependeraacute del balance entre las entradas y salidas de agua
(precipitaciones escurrimiento superficial y subterraacuteneo y la evaporacioacuten) y la forma
de la cuenca (profundidad pendiente aacuterea del espejo de agua)
En funcioacuten de la profundidad de penetracioacuten de la luz es posible diferenciar
dos aacutereas que determinan la distribucioacuten de los organismos la zona foacutetica (donde
penetra la luz y permite la fotosiacutentesis) y la zona afoacutetica (subyacente a la zona
foacutetica) Al depender de la radiacioacuten la zona foacutetica variacutea estacionalmente Ademaacutes
18
puede variar por el grado de turbidez la morfometriacutea del lago y la produccioacuten de
materia orgaacutenica en la columna de agua
Otro factor que influye en la productividad es el reacutegimen de mezcla de la
columna de agua pues favorece la oxigenacioacuten y el reciclado de nutrientes La
mezcla ocurre en condiciones de gran inestabilidad usualmente relacionado con el
reacutegimen de viento Por el contrario un lago estratificado resulta de grandes
diferencias en temperatura o salinidad entre la superficie (epilimnion) y el fondo del
lago (hipolimnion) que separa las masas de agua superficial y de fondo por una
termoclina (si la estratificacioacuten estaacute relacionada con diferencias de temperatura de
las masas de agua) o haloclina (diferencias de salinidad) En funcioacuten del reacutegimen
de mezcla los lagos se pueden clasificar en (Lewis 1983)
1 Amiacutecticos no hay mezcla vertical de la columna de agua
2 Monomiacutecticos friacuteos y caacutelidos se mezcla una vez al antildeo
3 Dimiacutecticos la columna de agua se mezcla dos veces al antildeo
4 Oligomiacutecticos Lagos estratificados la mayor parte del tiempo pero se mezclan a
intervalos irregulares mayores a 1 antildeo
5 Polimiacutecticos friacuteos y caacutelidos la columna de agua se mezcla varias veces al antildeo
El ciclo del N en lagos
Al estar presente en aacutecidos nucleicos aminoaacutecidos y proteiacutenas el N es un
nutriente esencial de los organismos Por lo tanto su disponibilidad en la columna
de agua es clave para la produccioacuten composicioacuten y acumulacioacuten de la MO a traveacutes
del tiempo (Olson 1998 Talbot 2002) Aunque el principal reservorio de N estaacute en
19
la atmosfera (N2) solo unos pocos organismos (eg cianobacterias) pueden fijarlo
directamente Asiacute el Nitroacutegeno Inorgaacutenico Disuelto (DIN) representa la principal
fuente de N disponible para la produccioacuten primaria en los ecosistemas acuaacuteticos
(Talbot et al 2002) El DIN estaacute compuesto por nitrato (NO3-) nitrito (N02
-) y amonio
(NH4+) y los cambios en su disponibilidad condicionan el tipo de produccioacuten primaria
(eg fijadores vs asimiladores de N ver Scott y Grant 2013 Scott 2008)
La Figura 1 resume los principales componentes en lagos del ciclo del N y
sus interacciones La principal entrada natural de N es viacutea fijacioacuten de N atmosfeacuterico
y es llevado a cabo principalmente por bacterias y algas verde-azules capaces de
romper el triple enlace entre los dos aacutetomos de N a un alto costo energeacutetico (Torres
et al 2012) El N fijado es reducido a NH3 Tras la muerte de los organismos el N
es liberado de sus tejidos por hongos y bacterias implicados en la descomposicioacuten
de la MO Una parte se deposita en el fondo del lago y la otra queda disponible para
ser asimilada por el fitoplancton como amonio mediante el proceso de
amonificacioacuten (Talbot 2002) La amonificacioacuten resulta de la reduccioacuten bacteriana
del amoniaco y se da preferentemente en condiciones anoacutexicas La volatizacioacuten del
amoniaco es un proceso por medio este se incorpora a la atmoacutesfera Este proceso
se da preferentemente en lagos aacutecidos (pH gt 85) El nitrato es otra forma de N
bioloacutegicamente disponible para el fitoplancton En los lagos el NO3 del DIN estaacute
compuesto por los aportes desde la cuenca hidrograacutefica en la que se circunscriben
por la resuspensioacuten desde el fondo del lago favorecido por procesos de mezcla
(descrito en seccioacuten anterior) y por los nitratos de la columna de agua Estos uacuteltimos
son el resultado de la nitrificacioacuten proceso realizado por bacterias aeroacutebicas
mediante el cual el amonio se oxida para formar nitrito y nitrato El ciclo se completa
20
con la desnitrificacioacuten que es la reduccioacuten bacteriana de nitrato al N atmosfeacuterico
Este proceso se da preferentemente en condiciones anoacutexicas
Figura 1 Materia Orgaacutenica sedimentaria y su relacioacuten con el ciclo del N y las
variaciones de δ15N (elaborado a partir de Talbot et al 2002) En la figura se
representan los factores clave en la acumulacioacuten de la MO sedimentaria y su
relacioacuten con el ciclo del N El δ15N de la MO es resultado de los aportes de MO
desde la cuenca MO de macroacutefitas y de la productividad bioloacutegica La productividad
en ecosistemas lacustres estaacute condicionada por DIN y la fijacioacuten de N atmosfeacuterico
El δ15N del pool del DIN disponible se va enriqueciendo por la asimilacioacuten
preferencial de 14N por parte del fitoplancton Ademaacutes el DIN remanente se va
enriqueciendo por accioacuten microbiana (amonificacioacuten nitrificacioacuten desnitrificacioacuten)
Reconstruyendo el ciclo del N a partir de variaciones en δ15N
La sentildeal isotoacutepica de N (δ15N) en los sedimentos lacustres puede ser usada
para reconstruir los cambios pasados del ciclo N la transferencia de N entre
ecosistemas terrestres y acuaacuteticos y el estado troacutefico del lago (Bruland y Mackenzie
2015 McLauchlan et al 2013 Woodward et al 2012 von Gunten et al 2009
Leng et al 2006 Meyers y Teranes 2001) La Figura 1 resume los principales
procesos y fuentes que afectan la sentildeal isotoacutepica δ15N (total o ldquobulkrdquo) en la MO de
21
los sedimentos lacustres Entre los que destacan las fuentes de MO (aloacutectona vs
autoacutectona) la bioproductividad lacustre y las entradas de N (ie fijacioacuten bioloacutegica
de N2) (Torres et al 2012) Estos procesos conducen a un fraccionamiento
isotoacutepico cineacutetico que favorece la disociacioacuten entre sus isotopos ldquopesadosrdquo (15N) y
ldquoligerosrdquo (14N) Asiacute por ejemplo los valores de δ15N de la MO se enriquecen en 15N
en la medida que los sedimentos lacustres estaacuten sometidos a perdidas de 14N viacutea
desnitrificacioacuten (Bruland y Mackenzie 2015 Diaz et al 2016 Talbot et al 2002)
Por otra parte el fitoplancton en la medida que aumenta su biomasa (eg
durante la estacioacuten caacutelida) puede llegar a asimilar el DIN hasta agotarse En este
caso el N remanente se va empobreciendo en 14N si no hay reposicioacuten de N (eg
aporte de NO3 de la cuenca) y la MO queda enriquecida en 15N Esta disminucioacuten
induce a una limitacioacuten de fitoplancton por N y los organismos diztroacuteficos pueden
verse estimulados por lo que se incrementa la entrada de N viacutea fijacioacuten de N2 (Scott
y Grantsz 2013) como resultado la MO oscila en valores maacutes bajos (en torno a 0permil)
La cantidad de MO que se deposita en el fondo del lago depende del
predominio de contribuciones provenientes desde la cuenca (aloacutectona) versus las
producidas dentro del mismo lago (autoacutectona) (Torres et al 2012) Aunque en
general los lagos reciben permanentemente aportes de sedimentos y MO desde
su cuenca los lagos pequentildeos tienden a reflejar maacutes los procesos que ocurren
solamente en su cuenca directa y reflejan los valores isotoacutepicos de la misma (Gu et
al 2006) Woodward et al (2012) realizaron un estudio en 50 lagos en Irlanda que
les permitioacute correlacionar diferentes patrones de LUCC (bosques de coniacuteferas
agricultura ganaderiacutea entre otros) con los valores de δ15N (y δ13C) en los
sedimentos superficiales de los lagos Encontraron que los valores de δ15N son maacutes
negativos en cuencas poco impactadas y maacutes positivos en aquellas con alto
22
impacto humano (eg usos de suelo agriacutecola y ganadero) McLauchlan et al (2007)
encontraron valores maacutes positivos de δ15N en los sedimentos del Mirror Lake New
Hampshire (29ordm N) a partir de la desforestacioacuten de su cuenca en 1790 CE (inicio
del asentamiento europeo en el lago) para el desarrollo de la actividad agriacutecola
Estos valores se volvieron maacutes negativos hacia valores similares al pre-
asentamiento europeo despueacutes del cese de la agricultura en la cuenca y la
recuperacioacuten del bosque a partir de 1929 CE
El anaacutelisis de δ15N en los sedimentos lacustres es una herramienta casi sin
explorar en Chile central Proponemos reconstruir la dinaacutemica de transferencia de
N cuenca-lago como consecuencia de los LUCC desde la colonizacioacuten espantildeola en
los lagos costeros de Laguna Matanzas (34ordmS) y Lago Vichuqueacuten (36ordmS) Como son
muacuteltiples los procesos que pueden afectar la sentildeal isotoacutepica de N (medido como
δ15N) en los sedimentos lacustres existen muchos problemas para su
interpretacioacuten paleoambiental (Leng et al 2006) Por ello en esta tesis utilizamos
un enfoque multiproxy que consiste en integrar el anaacutelisis limnoloacutegico y geoquiacutemico
de los sedimentos lacustres con los valores isotoacutepicos modernos de la columna de
agua (mediante monitoreo del POM) la relacioacuten vegetacioacuten-suelo de la cuenca y la
reconstruccioacuten de los principales usos de suelo de las cuencas desde 1975 CE
mediante el uso de imaacutegenes satelitales Considerando estas muacuteltiples liacuteneas de
evidencia nos planteamos utilizar las variaciones del δ15N para reconstruir los
cambios en la disponibilidad de N en los lagos costeros de Chile central y establecer
coacutemo y desde cuaacutendo las actividades humanas en la cuenca son lo suficientemente
importantes como para modificar el ciclo del N en los lagos desde la colonizacioacuten
espantildeola (siglo XVII)
23
Como hipoacutetesis planteamos que la dinaacutemica de transferencia de sedimentos
y nutrientes entre la cuenca y el lago tiene controles geoloacutegicos climaacuteticos y
bioloacutegicos que interactuacutean en el tiempo registraacutendose en la acumulacioacuten de MO de
los sedimentos lacustres (Fig1) Bajo este marco los LUCC modifican esta
dinaacutemica alterando la transferencia de N a los lagos ya que las actividades agriacutecolas
y ganaderas tienden a aportar maacutes N (aumentando δ15N) que la actividad forestal
(la que disminuye δ15N)
En el primer capiacutetulo titulado ldquoA combined approach to establishing the timing
and magnitude of anthropogenic nutrient alteration in a mediterranean coastal lake-
watershed systemrdquo se evaluoacute el rol de los LUCC en el control de la descarga de N
y sedimentos a Laguna Matanzas en un sistema cuenca-lago desde el siglo XVII
Para ello se realizoacute un anaacutelisis de la dinaacutemica sedimentaria lacustre mediante el
anaacutelisis de muacuteltiples proxys sedimentoloacutegicos (ie minerales y textura)
geoquiacutemicos (eg geoquiacutemica orgaacutenica e isotoacutepica) y bioloacutegicos (ie diatomeas) de
Laguna Matanzas que cubren los uacuteltimos 200 antildeos Ademaacutes se realizoacute una
reconstruccioacuten de los usos de suelo entre 1975 y 2016 a partir de imaacutegenes de
sateacutelites y se colectaron muestras de suelo de las principales coberturas de la
cuenca a los cuales se midioacute el δ15N
Entre los principales resultados obtenidos se destaca la influencia de la
ganaderiacutea y la denitrificacioacuten lo que explica los altos valores de δ15N evidenciados
por el registro lacustre durante la colonizacioacuten espantildeola e inicios de la repuacuteblica A
partir de la Gran Aceleracioacuten (1950 CE) la desforestacioacuten y el reemplazo de la
ganaderiacutea por plantaciones forestales tienen un correlato en el registro
sedimentario con valores de δ15N maacutes bajos Estos resultados demuestran que los
LUCC son el factor de primer orden para explicar los cambios observados en
24
nuestro registro de δ15N y a su vez implica un rol secundario del clima como posible
control de la disponibilidad de N en Laguna Matanzas Este capiacutetulo fue sometido
a la revista Scientific Reports y estaacute en revisioacuten desde el 26 de marzo de 2019 En
la elaboracioacuten del mauscrito han colaborado Claudio Latorre Blas Valero-Garceacutes
Matiacuteas Frugone Pablo Sarricolea Santiago Giralt Manuel Contreras-Lopez
Ricardo Prego y Patricia Bernardez
El segundo capiacutetulo se titula ldquoStable isotopes track land use and cover
changes in a mediterranean lake in central Chile over the last 600 yearsrdquo Aquiacute
evaluamos la dinaacutemica del N actual en el sistema cuenca-lago considerando los
valores isotoacutepicos modernos tanto de la vegetacioacuten como del suelo ademaacutes de los
cambios estacionales del POM de la columna de agua Esta informacioacuten se utiliza
como un anaacutelogo moderno para conocer el rol de los LUCC en la disponibilidad de
N en los uacuteltimos 600 antildeos Para ello se colectaron muestras filtradas de la columna
de agua a 2 5 y 20 m dos veces por antildeo entre noviembre de 2015 y julio de 2018
y se obtuvo el δsup1⁵N del POM Ademaacutes se colectoacute muestras de vegetacioacuten y suelo
de la cuenca diferenciando entre especies nativas plantaciones forestales y
vegetacioacuten herbaacutecea Esta informacioacuten se analizoacute en conjunto con la reconstruccioacuten
de los LUCC entre 1975 y 2014 y el anaacutelisis limnoloacutegico del lago Ello nos permitioacute
evaluar coacutemo el δ15N de los sedimentos lacustres refleja los valores δ15N de la
cuenca en el Lago Vichuqueacuten y el control que ejerce el clima en la sentildeal isotoacutepica
de POM Este capiacutetulo seraacute sometido a la revista Science of total environmet
proximamente En la elaboracioacuten del mauscrito han colaborado Claudio Latorre
Blas Valero-Garceacutes Matiacuteas Frugone Pablo Sarricolea Leonardo Villacis y M Laura
Carrevedo
25
Entre los principales resultados encontramos que el δ15N en los sedimentos
lacustres es maacutes positivo con presencia de agricultura en la cuenca y maacutes negativo
cuando esta es reemplazada por cobertura arboacuterea bien sea plantaciones
forestales bien sea bosque nativo A su vez durante la estacioacuten seca y caacutelida la
mayor entrada al lago es viacutea fijacioacuten de N atmosfeacuterico (bajos valores de δ15NPOM)
Durante la estacioacuten huacutemeda en cambio prevalecen los aportes de la cuenca con
altos valores de δ15NPOM Ello estariacutea evidenciando un cambio estacional en la
composicioacuten de especies en verano ligado a especies fijadoras de N y en invierno
las algas y microorganismos que consumen el DIN de la columna de agua
Referencias
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea AM Marquet PA 2010 From the Holocene to the Anthropocene A historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the
next carbon Earth rsquo s Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409 httpsdoiorg102134jeq20090005
Diacuteaz FP Frugone M Gutieacuterrez RA Latorre C 2016 Nitrogen cycling in an
extreme hyperarid environment inferred from δ15 N analyses of plants soils and herbivore diet Sci Rep 6 1ndash11 httpsdoiorg101038srep22226
Ekdahl EJ Teranes JL Wittkop CA Stoermer EF Reavie ED Smol JP
2007 Diatom assemblage response to Iroquoian and Euro-Canadian eutrophication of Crawford Lake Ontario Canada J Paleolimnol 37 233ndash246 httpsdoiorg101007s10933-006-9016-7
Elbert W Weber B Burrows S Steinkamp J Buumldel B Andreae MO
Poumlschl U 2012 Contribution of cryptogamic covers to the global cycles of carbon and nitrogen Nat Geosci 5 459ndash462
26
httpsdoiorg101038ngeo1486 Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506
httpsdoiorg101126science1215567 ARTICLE Engstrom DR Almendinger JE Wolin JA 2009 Historical changes in
sediment and phosphorus loading to the upper Mississippi River Mass-balance reconstructions from the sediments of Lake Pepin J Paleolimnol 41 563ndash588 httpsdoiorg101007s10933-008-9292-5
Erisman J W Sutton M A Galloway J Klimont Z amp Winiwarter W (2008)
How a century of ammonia synthesis changed the world Nature Geoscience 1(10) 636
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892
httpsdoiorg101126science1136674 Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J Christie
D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592 Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
IPCC 2013 Climate Change 2013 The Physical Science BasisContribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change Cambridge University Press Cambridge United Kingdom and New York NY USA
httpsdoiorg101017CBO9781107415324 Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007 Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32iumliquestfrac12S 3050iumliquestfrac12miumliquestfrac12asl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470
27
httpsdoiorg101073pnas0701779104 McLauchlan KK Gerhart LM Battles JJ Craine JM Elmore AJ Higuera
PE Mack MC McNeil BE Nelson DM Pederson N Perakis SS 2017 Centennial-scale reductions in nitrogen availability in temperate forests of the United States Sci Rep 7 1ndash7 httpsdoiorg101038s41598-017-08170-z
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Myers N Mittermeler RA Mittermeler CG Da Fonseca GAB Kent J
2000 Biodiversity hotspots for conservation priorities Nature httpsdoiorg10103835002501
Pentildeuelas J Poulter B Sardans J Ciais P Van Der Velde M Bopp L
Boucher O Godderis Y Hinsinger P Llusia J Nardin E Vicca S Obersteiner M Janssens IA 2013 Human-induced nitrogen-phosphorus imbalances alter natural and managed ecosystems across the globe Nat Commun 4 httpsdoiorg101038ncomms3934
Prieto M Salazar D Valenzuela MJ 2019 The dispossession of the San
Pedro de Inacaliri river Political Ecology extractivism and archaeology Extr Ind Soc httpsdoiorg101016jexis201902004
Robertson GP Vitousek P 2010 Nitrogen in Agriculture Balancing the Cost of
an Essential Resource Ssrn 34 97ndash125 httpsdoiorg101146annurevenviron032108105046
Sardans J Rivas-Ubach A Pentildeuelas J 2012 The CNP stoichiometry of
organisms and ecosystems in a changing world A review and perspectives Perspect Plant Ecol Evol Syst 14 33ndash47 httpsdoiorg101016jppees201108002
Schindler DW Howarth RW Vitousek PM Tilman DG Schlesinger WH
Matson PA Likens GE Aber JD 2007 Human Alteration of the Global Nitrogen Cycle Sources and Consequences Ecol Appl 7 737ndash750 httpsdoiorg1018901051-0761(1997)007[0737haotgn]20co2
Scott E and EM Grantzs 2013 N2 fixation exceeds internal nitrogen loading as
a phytoplankton nutrient source in perpetually nitrogen-limited reservoirs Freshwater Science 2013 32(3)849ndash861 10189912-1901
28
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Talbot M R (2002) Nitrogen isotopes in palaeolimnology In Tracking
environmental change using lake sediments (pp 401-439) Springer Dordrecht
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable
isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K
Yeager KM 2016 Different responses of sedimentary δ15N to climatic changes and anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
29
CAPIacuteTULO 1 A COMBINED APPROACH TO ESTABLISHING THE TIMING
AND MAGNITUDE OF ANTHROPOGENIC NUTRIENT ALTERATION IN A
MEDITERRANEAN COASTAL LAKE- WATERSHED SYSTEM
30
A combined approach to establishing the timing and magnitude of anthropogenic
nutrient alteration in a mediterranean coastal lake- watershed system
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugoneabc Pablo
Sarricolead Santiago Giralte Manuel Contreras-Lopezf Ricardo Prego g Patricia
Bernaacuterdez g Blas Valero-Garceacutesch
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Institute of Earth Sciences Jaume Almera (ICTJA-CSIC) CLluis Solegrave Sabaris sn Barcelona E-
08028 Spain
f Facultad de Ingenieriacutea y Centro de Estudios Avanzados Universidad de Playa Ancha Traslavintildea
450 Vintildea del Mar Chile
g Instituto de Investigaciones Marinas (CSIC) CEduardo Cabello 6 36208 Vigo Spain
h Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding author
E-mail address
clatorrebiopuccl magdalenafuentealbagmailcom
Abstract
Since the industrial revolution and especially during the Great Acceleration (1950
CE) human activities have profoundly altered the global nutrient cycle through land
use and cover changes (LUCC) However the timing and intensity of recent N
variability together with the extent of its impact in terrestrial and aquatic ecosystems
and coupled effects of regional LUCC and climate are not well understood Here
we used a multiproxy approach (sedimentological geochemical and isotopic
31
analyses historical records climate data and satellite images) to evaluate the role
of LUCC as the main control for N cycling in a coastal watershed system of central
Chile during the last few centuries The largest changes in N dynamics occurred in
the mid-1970s associated with the replacement of native forests and grasslands for
livestock grazing by plantations of introduced Monterey pine (Pinus radiata) and
eucalyptus (Eucalyptus globulus) Lake productivity increased and is evidenced by
an increase trend in δ15N values Our study shows that anthropogenic land
usecover changes are key in controlling nutrient supply and N availability in
Mediterranean watershed ndash lake systems and that large-scale forestry
developments during the mid-1970s likely caused the largest changes in central
Chile
Keywords Anthropocene Organic geochemistry watershedndashlake system Stable
Isotope Analyses Land usecover change Nitrogen cycle Mediterranean
ecosystems central Chile
1 INTRODUCTION
Human activities have become the most important driver of the nutrient cycles in
terrestrial and aquatic ecosystems since the industrial revolution (Gruber and
Galloway 2008 Galloway et al 2008 Holtgrieve et al 2011 Fowler et al 2013
Goyette et al 2016) Among these N is a common nutrient that limits productivity
in terrestrial and aquatic ecosystems (Vitousek and Howarth 1991 McLauchlan et
al 2013) With the advent of the Haber-Bosch industrial N fixation process in the
early 20th century total N fluxes have surpassed previous planetary boundaries
32
(Galloway et al 2008 Elser 2011) reaching unprecedented values (ie tipping
points) in the Earth system especially during what is now termed the Great
Acceleration (which began in the 1950s) which intensified in the 1970s (Howarth
2004 Steffen et al 2015) Yet dramatic changes in LUCC have occurred in the last
few centuries mainly linked to forestry and agro-pastoral activities (Poraj-Goacuterska et
al 2017 Ge et al 2019 Cheng et al 2019) These have had large impacts on N
(and Carbon -C-) availability in terrestrial and aquatic ecosystems but the synergic
effect with climate change and global N dynamics has not been established
(Vitousek et al 1997 Mclauchlan et al 2007 2013 Pongratz et al 2010
Woodward et al 2012 Mclauchlan et al 2017)
The onset of the Anthropocene poses significant challenges in mediterranean
regions that have a strong seasonality of hydrological regimes and an annual water
deficit (Stocker et al 2013) Mediterranean climates occur in all continents
(California central Chile Australia South Africa circum-Mediterranean regions)
providing a unique opportunity to investigate global change processes during the
Anthropocene in similar climate settings but with variable geographic and cultural
contexts The effects of global change in mediterranean watersheds have been
analyzed from different perspectives hydrology (Giorgi 2006 Arnell and Gosling
2013 Prudhomme et al 2014) vegetation dynamics (Lenihan et al 2003 Vicente-
Serrano et al 2013 Matesanz and Valladares 2014) sediment dynamics (Garciacutea-
Ruiz et al 2013 Garciacutea-Ruiz 2010 Syvitski et al 2005) changes in
biogeochemical cycles (Thomas et al 2013 Fowler et al 2013 Frank et al 2015)
carbon storage (Muntildeoz-Rojas et al 2015) and biodiversity (Hooper et al 2012) A
recent review showed an extraordinarily high variability of erosion rates in
mediterranean watersheds positive relationships with slope and annual
33
precipitation and the paramount effect of land use (Garciacutea-Ruiz et al 2015)
However the temporal context and effect of LUCC on nutrient supply to
mediterranean lakes has not been analyzed in much detail
Major LUCC in central Chile occurred during the Spanish Colonial period
(17th-18th centuries) (Armesto 1994 Gurnell et al 2001 Figueroa et al 2004
Martiacuten-Foreacutes et al 2012 Contreras-Loacutepez et al 2014) with the onset of
industrialization and mostly during the mid to late 20th century (von Gunten et al
2009 Gayo et al 2012) Recent copper pollution caused by 20th century mining
and industrial smelters has been documented in cores throughout the Andes
(Laguna el Ocho and Laguna Ensuentildeo von Gunten et al 2009) and also from our
surveys in the central valley (Batuco wetland) coastal range (Cordillera de Name)
and along the coast itself (Bucalemito and Colejuda) (Valero-Garceacutes et al 2010
unpublished data)
Paleolimnological studies have shown how these systems respond to
climate LUCC and anthropogenic impacts during the last millennia (Jenny et al
2002 2003 Villa Martiacutenez et al 2002 Frugone-Aacutelvarez et al 2017 Contreras et
al 2018) Furthermore changes in sediment and nutrient cycles have also been
identified in associated terrestrial ecosystems dating as far back as the Spanish
Conquest and related to fire clearance and wood extraction practices of the native
forests (Armesto 1994 Contreras-Loacutepez et al 2014) Nevertheless pollen and
limnological evidence argue for a more recent timing of the largest anthropogenic
impacts in central Chile For example paleo records show that during the mid-20th
century increased soil erosion followed replacement of native forest by Pinus
radiata and Eucalyptus globulus plantations at Laguna Matanzas Aculeo and
34
Vichuqueacuten lakes (Jenny et al 2002 Villa-Martiacutenez et al 2002 2003 Frugone-
Aacutelvarez et al 2017)
Lakes are a central component of the global carbon cycle Lakes act as a
sink of the carbon cycle both by mineralizing terrestrially derived organic matter and
by storing substantial amounts of organic carbon (OC) in their sediments (Anderson
et al 2009) Paleolimnological studies have shown a large increase in OC burial
rates during the last century (Heathcote et al 2015) however the rates and
controls on OC burial by lakes remain uncertain as do the possible effects of future
global change and the coupled effect with the N cycle LUCC intensification of
agriculture and associated nutrient loading together with atmospheric N-deposition
are expected to enhance OC sequestration by lakes Climate change has been
mainly responsible for the increased algal productivity since the end of the 19th
century and during the late 20th century in lakes from both the northern (Ruumlhland et
al2015) and southern hemispheres (Michelutti et al 2015 Carrevedo et al 2015)
but many studies suggest a complex interaction of global warming and
anthropogenic influences and it remains to be proven if climate is indeed the only
factor controlling these transitions (Catalaacuten et al 2013) Alternative causes for
recent N (Galloway et al 2008) increases in high altitude lakes such as catchment
mediated processes cannot be ruled out (Camarero and Catalan 2012 Fritz and
Anderson 2013) Few lake-watershed systems have robust enough chronologies of
recent changes to compare variations in C and N with regional and local processes
and even fewer of these are from the southern hemisphere (McLauchlan et al
2007 Holtgrieve et al 2011)
In this paper we present a multiproxy lake-watershed study including N and
C stable isotope analyses on a series of short cores from Laguna Matanzas in
35
central Chile focused in the last 200 years We complemented our record with land
use surveys satellite and aerial photograph studies Our major objectives are 1) to
reconstruct the dynamics among climate human activities and changes in the N
cycle over the last two centuries 2) to evaluate how human activities have altered
the N cycle during the Great Acceleration (since the mid-20th century)
2 STUDY SITE
Laguna Matanzas (33ordm45rsquoS 71ordm 40acuteW 7 m asl Fig 1) is a coastal lake located
in central Chile near to a large populated area (Santiago gt6106 inhabitants) The
lake has a surface area of 15 km2 with a max depth of 3 m and a watershed of 30
km2 The lake basin is emplaced over Pleistocene-Holocene aeolian and alluvial fan
deposits (SERNAGEOMIN 2003) A recent phase of dune activity occurred from the
mid to late Holocene which mostly sealed off the basin from the ocean (Villa-
Martinez 2002) Climate in Laguna Matanzas is characterized by cool-wet winters
and hot-dry summers with annual precipitation of ~510 mm and a mean annual
temperature of 12ordmC Central Chile is in the transition zone between the southern
hemisphere mid-latitude westerlies belt and the South Pacific Anticyclone (or SPA)
(Garreaud et al 2009) In winter precipitation is modulated by the north-west
displacement of the SPA the northward shift of the westerlies wind belt and an
increased frequency of storm fronts stemming off the Southern Hemisphere
Westerly Winds (SWW) (Frugone-Aacutelvarez et al 2017) Austral summers are
typically dry and warm as a strong SPA blocks the northward migration of storm
tracks stemming off the SWW
36
Historic land cover changes started after the Spanish conquest with a Jesuit
settlement in 1627 CE near El Convento village and the development of a livestock
ranch that included the Matanzas watershed After the Jesuits were expelled from
South America in 1778 CE the farm was bought by Pedro Balmaceda and had more
than 40000 head of cattle around 1800 CE (Contreras-Lopez et al 2014) The first
Pinus radiata and Eucalyptus globulus trees were planted during the second half of
the 19th century and were mostly used for dune stabilization (Albert 1900 Gibson
1972) However the main plantation phase occurred 60 years ago (Villa-Martinez
2002) as a response to the application of Chilean Forestry Laws promulgated in
1931 and 1974 and associated state subsidies
Major land cover changes occurred recently from 1975 to 2008 as shrublands
were replaced by more intensive land uses practices such as farmland and tree
plantations (Schulz et al 2010) Laguna Matanzas is part of the Reserva Nacional
Humedal El Yali protected area (Ramsar Ndeg 878) Despite this protected status the
lake and its watershed have been heavily affected by intense agricultural and
farming activities during the last decades The main inlet ldquoLas Rosasrdquo has been
diverted for crop irrigation causing a significant loss of water input to the lake
Consequently the flooded area of the lake has greatly decreased in the last couple
of decades (Fig 1b) Exotic tree species cover a large surface area of the
watershed Recently other activities such as farms for intensive chicken production
have been emplaced in the watershed
37
Figure 1 Laguna Matanzas a) Digital Elevation Model showing the watershed and
the surface hydrologic connection with Colejuda and Cabildo lakes b) Climograph
depicting the warm dry season in austral summer c) Annual precipitation from
1965-2015 (arrow shows start of the most recent ldquomega-droughtrdquo- see Garreaud et
al 2017) d) A decline of lake area occurs between 2007 and 2018 Lake surface
area decreased first along the western sector (in 2007) followed by more inland
areas (in 2018)
38
3 RESULTS
31 Age Model
The age model for the Matanzas sequence was developed using Bacon software
to establish the deposition rates and age uncertainty (Blaauw and Christen 2011)
It is based on 210Pb and 137Cs dates and two 14C dates (Table 2) According to this
age model the lake sequence spans the last 1000 years (Fig 2) A major breccia
layer (unit 3b) was deposited during the early 18th century which agrees with
historic documents indicating that a tsunami impacted Laguna Matanzas and its
watershed in 1730 CE (Contreras-Loacutepez et al 2014) Here we focus in the last 200
years were the most important changes occurred in terms of LUCC (after the
sedimentary hiatus caused by the tsunami) The Spanish colonial period (17th -18th
century) brought new forms of territorial management along with an intensification
of watershed use which remained relatively unchanged until the 1900s
39
Figure 2 a) Age-depth model obtained for the Laguna de Matanzas sedimentary
sequence A hiatus occurs at 80 cm depth (unit 3b) The section of core used for our
analysis is highlighted in a red rectangle b) Close up of the age model used for
analysis of recent anthropogenic influences on the N cycle c) Information regarding
the 14C dates used to construct age model
Lab code Sample ID
Depth (cm) Material Fraction of modern C
Radiocarbon age
Pmc Error BP Error
D-AMS 021579
MAT11-6A 104-105 Bulk Sediment
8843 041 988 37
D-AMS 001132
MAT11-6A 1345-1355
Bulk Sediment
8482 024 1268 21
POZ-57285
MAT13-12 DIC Water column 10454 035 Modern
Table 2 Laguna Matanzas radiocarbon dates
32 The sediment sequence
Laguna Matanzas sediments consist of massive to banded mud with some silt
intercalations They are composed of silicate minerals (plagioclase quartz and clay
minerals) with relatively high TOC content (Fig 3) Pyrite is a common mineral
indicating dominant anoxic conditions in the lake sediments whereas aragonite
occurs only in the uppermost section Mineralogical analyses visual descriptions
texture and geochemical composition were used to characterize five main facies
(Fig 3) F1 (organic-rich mud) represents baseline sedimentation in a shallow well-
mixed brackish highly productive lake and F1acute (dark orange) is a less organic facies
than F1 (more details see table in the supplementary material) F2 (massive to
banded silty mud) indicates periods of higher clastic input into the lake but finer
(mostly clay minerals) likely from suspension deposition associated with flooding
40
events Aragonite (up to 15 ) occurs in both facies but only in samples from the
uppermost 30 cm and it is interpreted as endogenic linked to higher MgFe waters
and elevated biologic productivity
Figure 3 Sedimentary facies and units mineralogy grain size elemental geochemical
and isotopic composition of Laguna Matanzas core Values highlighted in gray indicate
that these are above average
The banded to laminated fining upward silty clay layers (F3) reflect
deposition by high energy turbidity currents The presence of aragonite suggests
that littoral sediments were incorporated by these currents Non-graded laminated
coarse silt layers (F4) do not have aragonite indicating a dominant watershed
41
sediment source Both facies are interpreted as more energetic flood deposits but
with different sediment sources A unique breccia layer with coarse silt matrix and
cm-long soft clasts (F5) is interpreted as a high-energy event (ie a tsunami)
capable of eroding the littoral zone and depositing coarse clastic material in the
distal zone of the lake Similar coarse breccia layers have been found at several
coastal sites in Chile and interpreted as tsunami-related deposits (Vargas et al
2005 Le Roux et al 2008)
33 Sedimentary units
Three main units and six subunits have been defined (Fig 3) based on
sedimentary facies and sediment composition We use ZrTi as an indicator of the
mineral fraction transported from the watershed (Marzecoacuteva et al 2011) as higher
ZrTi (F3 and F4) is commonly associated with coarser sediments (Cuven et al
2010) A high correlation among Br BrTi and TOC (r = 046-087 p value=0)
supports the use of BrTi as an indicator of lake productivity (Marzecoacuteva et al 2011
Frugone-Aacutelvarez et al 2017) The MnFe ratio is indicative of lake bottom
oxygenation (Naecher et al 2013) as under reducing conditions Mn mobilizes more
than Fe leading to a decreased MnFe ratio (Marzecoacuteva et al 2011) SrTi indicates
periods of increased aragonite formation as Sr is preferentially included in the
aragonite mineral structure (Veizer et al 1971) (See supplementary material)
The basal unit (3c 129 to 99 cm) is relatively organic-rich (TOC mean=26
BioSi mean=5) and composed by F1 (without aragonite) with some coarser F4
flood layers (ZrTi mean=025) Unit 3b (98 to 80 cm) is interpreted as a tsunami or
storm surge deposit (breccia F5 grading into massive to banded silt F2) Unit 3a
(79 to 73 cm) is characterized by relatively low productivity (TOC=2 BrTi=002
42
BioSi=4) under anoxic conditions (MnFe=001) Unit 2a (72 to 31cm) has
relatively less organic content and more intercalated clastic facies F3 and F4 The
top of this unit (43-30 cm) has elevated TS values The Subunit 1b (30-20 cm)
shows increasing TOC BioSi and BrTi values (TOC mean = 29 BioSi mean =
54 BrTi mean=004) and the upper subunit 1a (19-0 cm) has the highest TOC
(mean = 64) and BioSi (mean = 56) high BrTi mean = 010 and the presence
of aragonite More frequent anoxic conditions (MnFe lower than 001) during units
3 and 2 shifted towards more oxic episodes (MnFe = 003) in unit 1 (Fig 3)
34 Isotopic signatures
Figure 4 shows the isotopic signature from soil samples of the major land
usescover present in the Laguna Matanzas used as an end member in comparison
with the lacustrine sedimentary units δ15N from cropland samples exhibit the
highest values whereas grassland and soil samples from lake shore areas have
intermediate values (Fig 4) Tree plantations and native forests have similarly low
δ15N values (+11 permil SD=24) All samples (except those from the lake shore)
exhibit low δ13C values (from ndash285permil to ndash298permil) CNmolar from agriculture land
lakeshore area and non-vegetation areas samples display the lowest values (about
18) CNmolar from tree plantations and native forest have the highest values (383
and 267 respectively)
43
Figure 4 C-N stable isotope plot showing a comparison of lake sediments grouped
by sedimentary units (MAT11-6A) with the soil end members of present-day (lake
shore and land usecover) from Laguna Matanzas
The δ15N values from sediment samples (MAT11-6A) range from ndash15 and
+53permil (mean= +35 SD=05) δ13C values range from ndash266permil to ndash202permil (mean=
ndash240 SD=14) In unit 3c δ15N and δ13C show relatively high values (mean=
+41permil and ndash233permil respectively) δ15N and δ13C from unit 3b and 3a fluctuate at
slightly lower values than in 3c (mean δ15N from 3a= +38permil and 3b mean= +39permil
mean δ13C from 3a= ndash242permil and 3b mean= ndash244permil) In unit 2a δ15N values are
relatively high (mean= +38permil) but show a slightly decreasing trend (from +52 to
+24permil at 66 cm and 36 cm respectively) δ13C also decreases (mean= ndash242permil)
reaching minimum values at 45 cm (ndash266permil) with a sharp increase towards the top
of this unit with maximum values ca ndash210permil Unit 1 exhibits the lowest δ15N values
(1a mean= +11permil 1b mean= +28permil) with negative values in the uppermost
44
sediments (ndash04permil at 14 cm) δ13C values show a decreasing trend over most of
subunit 1b and increase only near the very top of this unit
35 Recent land use changes in the Laguna Matanzas watershed
Major LUCC from 1975 (unit 1b) to 2016 CE in the Laguna Matanzasacutes
watershed is summarized in Figure 5 The watershed has a surface area of 30 km2
of which native forest (36) and grassland areas (44) represented 80 of the
total surface in 1975 The area occupied by agriculture was only 02 and tree
plantations were absent Isolated burned areas (33) were located mostly in the
northern part of the watershed By 1989 tree plantations surface area had increased
to 5 burned areas to 17 and agricultural fields to 9 (a 45-fold increase) and
native forest and grassland sectors decreased to 23 and 27 respectively By
2016 agricultural land and tree plantations have increased to 17 of the total area
whereas native forests decreased to 21
45
Figure 5 Land uses and cover changes from 1975 to 2016 in Laguna Matanzas
watershed from natural cover and areas for livestock grazing (grassland) to the
expansion of agriculture and forest plantation
4 DISCUSSION
41 N and C dynamics in Laguna Matanzas
Small lakes with relatively large watersheds such as Laguna Matanzas would
be expected to have relatively high contributions of allochthonous C to the sediment
OM (Gu et al 2006) Terrestrial C3 plants (δ13C =ndash26 to ndash28permil Meyers and Terranes
2001 Ku et al 2007) are dominant in the Laguna Matanzas watershed Likewise
our soil samples ranged across similar although slightly more negative values
46
(δ13C = ndash30 to ndash28permil Fig 4) to those proposed by Meyers and Terranes (2001)
and are used here as terrestrial end members oil samples were taken from the lake
shore (δ13C = ndash22permil) and POM from surface water (δ13C = ndash24permil) were more
positive than the terrestrial end member and are used as lacustrine end members
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from terrestrial vegetation and more positive δ13C values have increased
aquatic OM (Gu et al 2006 Torres et al 2012) Phytoplankton preferentially uptake
12C leaving the DIC pool enriched in 13C (Gu et al 2006) especially when there are
no important external sources of C (eg decreased C input from the watershed)
Therefore during events of elevated primary productivity the phytoplankton uptakes
12C until its depletion and are then obligated to use the heavier isotope resulting in
an increase in δ13C Changes in lake productivity thus greatly affect the C isotope
signal (Torres et al 2012) with high productivity leading to elevated δ13C values
(Torres et al 2012 Gu et al 2006)
In a similar fashion the N isotope signatures in Laguna Matanzas reflect a
combination of factors including different N sources (autochthonousallochthonous)
and lake processes such as productivity isotope fractionation in the water column
and sediment denitrification Elevated δ15N values from a POM sample (+22permil) and
average values from the lake shore (mean=+34permil SD=028) are used as aquatic
end members whereas terrestrial samples have values from +10 +24 (tree species)
to +77permil +35 (agriculture) which we use as terrestrial end members (Fig 4)
Autochthonous OM in aquatic ecosystems typically displays low δ15N values
when the OM comes from N-fixing species Atmospheric fixation of N2 by
cyanobacteria ends in OM with δ15N values close to 0permil (Leng et al 2006)
Phytoplankton preferentially uptake 14N from Dissolved Inorganic Nitrogen (DIN) in
47
the water column and derived OM typically have δ15N values lower than DIN values
When productivity increases the remaining DIN becomes depleted in 14N which in
turn increases the δ15N values of phytoplankton over time especially if the N not
replenished (Torres et al 2012) Thus high POM δ15N values from Laguna
Matanzas reflect elevated phytoplankton productivity with a 14N- depleted DIN In
addition N-watershed inputs also contribute to high δ15N values Heavily impacted
watersheds by human activities are often reflected in isotope values due to land use
changes and associated modified N fluxes For example the input of N runoff
derived from the use of inorganic fertilizers leads to the presence of elevated δ15N
(between ndash4 to +4permil) values in the water bodies (Elliott and Brush 2006 Diebel and
Vander Zanden 2009) Widory et al (2004) reported a direct relationship between
elevated δ15N values and increased nitrate concentration from manure in the
groundwater of Brittany (France) Terranes and Bernasconi (2000) found a good
correlation between augmented nutrient loading and a progressive increase in δ15N
values of sedimentary OM related to agricultural land use
Post-depositional diagenetic processes can further affect C and N isotope
signatures Several studies have shown a decrease in δ13C values of OM in anoxic
environments particularly during the first years of burial related to the selective
preservation of OM depleted in 13C (Hollander and Smith 2001 Lehmann et al
2002 Gӓlman et al 2009 Torres et al 2012) Diagenetic processes can also lead
to post-burial N isotope enrichment of the sediments Indeed 14N is consumed more
rapidly than 15N by denitrifying bacteria which intensifies under anoxic conditions
(Diebel and Vander Zanden 2009) Thus the remaining OM pool will be enriched
in 15N leaving to elevated δ15N values (Granger et al 2008 Menzel et al 2013)
48
In summary the relatively high δ15N values in sediments of Laguna Matanzas
reflect N input from an agriculturegrassland watershed with positive synergetic
effects from increased lake productivity enrichment of DIN in the water column and
most likely denitrification The increase of algal productivity associated with
increased N terrestrial input andor recycling of lake nutrients (and lesser extent
fixing atmospheric N) and denitrification under anoxic conditions can all increase
δ15N values (Fig 3) In addition elevated lake productivity without C replenishing
(eg by terrestrial C input) produces shifts towards positive δ13C values whereas C
input from the watershed generates more negative δ13C values
42 Recent evolution of the Laguna Matanzas watershed
Sedimentological compositional and geochemical indicators show three
depositional phases in the lake evolution under the human influence in the Laguna
Matanzas over the last two hundred years Although the record is longer (around
1000 years) we analyzed the last two centuries (unit 2 and 1) to provide a recent
historical context for the large changes detected during the 20th century
The first phase lasted from the beginning of the 19th century until ca 1940
(unit 2a Fig 3 4) and was characterized by moderate productivity with elevated
sediment input from the watershed as indicated by our geochemical proxies (BrTi
= 004 AlTi gt 01 Fig 6) Higher lake levels and dominant anoxic bottom conditions
(MnFe = 001 Fig3) can be related to increased precipitation (mean= 557 mmyr)
and lower temperatures (summer annual temperature lt19ordmC) During the Spanish
colonial period the Laguna Matanzas watershed was used as a livestock farm
(Jesuits 1627-1780 unit 3 followed by the Pedro Balmaceda era 1780-1810- unit
2a) (Contreras-Loacutepez et al 2014) The main buildings and farm facilities at the El
49
Convento village During this period livestock grazing and lumber extraction for
mining would have involved extensive deforestation and loss of native vegetation
(eg Armesto et al 1994 2010) However the Matanzas pollen record does not
show any significant regional deforestation during this period (Villa Martiacutenez 2002)
suggesting that the impact may have been highly localized
Lake productivity sediment input and elevated precipitation (Fig 6) all
suggest that N availability was related to this increased input from the watershed
The N from cow manure and soil particles would have led to higher δ15N values
(Elliot and Brush 2016) In addition anoxic lake bottom conditions would lead to
even further enrichment of buried sediment N The δ13C values lend further support
to our interpretation of increased sediment input -and N- from the watershed
Decreased δ13C values reach their lowest values in the entire record (ndash270permil) at
ca 1910 CE (Fig 4 6)
During most of the 19th century human activities in Laguna Matanzas were
similar to those during the Spanish Colonial period However the appearance of
Pinus radiata and Eucalyptus globulus pollen ca 1890 CE (Gibson 1972) the dune
stabilization-afforestation program which began ca 1900 CE (Albert 1900) and the
application of the Chilean forestry law of 1931 (DFL nordm265) contributed to an
increased capacity of the surrounding vegetation to retain nutrients and sediments
The law subsidized forest plantations in areas devoid of vegetation and prohibited
the cutting of forest on slopes greater than 45ordm These land use changes were coeval
with decreased sediment inputs (AlTi trend) from the watershed slightly increased
lake productivity (BrTi from 001 to 003) and a decrease in annual precipitation
(Fig 6) N isotope values become more negative during this period although they
remained high (from +49permil to +37permil) whereas the δ13C trend towards more
50
positive values reflects changes in the N source from watershed to in-lake dynamics
(e g increased endogenic productivity)
The second phase started after 1940 and is clearly marked by an abrupt
change in the general trend of δ15N that oscillates around +41permil (SD = 04permil) during
the first phase to +24permil (SD = 14permil) during the second These δ15N trends reflect
the lowest watershed nutrient and sediment inputs (based on the AlTi record)
decreased precipitation (mean = 318 mm year) and a slight increase in lake
productivity (increased BrTI) Depositional dynamics in the lake likely crossed a
threshold as human activity intensified throughout the watershed and lake levels
decreased
During the Great Acceleration δ15N values shifted towards higher values to
ca 3permil with an increase in δ13C values that are not reflected either in lake
productivity or lake level As the sediment input from the watershed increased and
precipitation remained as low as the previous decade δ15N values during this period
are likely related to watershed clearance which would have increased both nutrient
and sediment input into the lake
The δ13C trend to more positive values reaching the peaks in the 1960s (ndash
212permil) at the same time as the peaks in temperature (196 ordmC mean annual) with a
downward trend in precipitation A shift in OM origin from macrophytes and
watershed input influences to increased lake productivity could explain this trend
(Fig 4 1b)
In the 1970s the Laguna Matanzasacute watershed was mostly covered by native
forest (36) and grassland areas were intended for livestock grazing (44 Fig 5)
Both soils have δ15NSoil lt 5permil (Fig 4) Farming fields occupied a very small area and
tree plantations were almost nonexistent The decreasing trend in δ15N values seen
51
in our record is interrupted by several large peaks that occurred between ca 1975
and ca 1989 when the native forest and grassland areas fell by 23 and 27
respectively largely due to fires affecting 17 of the forests Agriculture fields
increased by 4 and forest plantations increased by 9 (unit 1a) Concomitantly
sediment ndash and likely N - inputs from the watershed decreased (as indicated by the
trend in AlTi) the precipitation was still relatively low (Fig 6) These changes are
likely related to the increase of vegetation cover especially of tree plantations (which
have more negative δ15N values) The small increase in productivity in the lake could
have been favored by increased temperature (von Gunten et al 2009) After 1989
the increase in agriculture land (17 in 2016) correlates with increasing δ15N δ13C
and TOC trends in spite of declining rainfall The increase of forest plantations was
mostly in response to the implementation of the Law Decree of Forestry
Development (DL 701 of 1974) that subsidized forest plantation After 1989 the
increase in agricultural land (17 in 2016) is synchronous with increasing δ15N
δ13C TOC and MnFe trends despite the decline in rainfall and overall lower lake
levels as more water is used for irrigation
The third phase started c 1990 CE (unit 1a) when OM accumulation rates
increase and δ13C δ15N decreased reaching their lowest values in the sequence
around 2000 CE Afterward during the 21st century δ13C and δ15N values again
began to increase The onset of unit 1 is marked by increased lake productivity and
decreased sediment input (AlTi lt 02) synchronous with intensive farming replacing
forestry and extensive agriculture (Fig 5 6)
A change in the general trend of δ15N values which decreased until 1990
(unit 1a) as the native forest and grassland area fell to 23 and 27 respectively
is most likely due to deforestation and fires Agriculture surface increased to 4 and
52
forest plantations increased by 9 (Fig 5) Concomitantly sediment ndash and likely N
ndash inputs from the watershed decreased probably related to the low precipitation (Fig
1b) and the increase of vegetation cover in the watershed in particularly by tree
plantations (with more negative δ15N Fig 4)
At present agriculture and tree plantations occupy around 34 of the
watershed surface whereas native forests and grassland cover 21 and 25
respectively Lake productivity as indicated by BrTi (Fig6) is higher and generates
OM with lower δ15N and higher δ13C values (about +2permil and ndash20permil ca 2000 CE
respectively) due to in-lake processes (ie biological N fixation and nutrient
recycling) and driven by changes in the arboreal cover which diminishes nutrient
flux into the lake (Fig6)
53
Figure 6 Anthropogenic and climatic forcing and lake dynamics response
(productivity sediment input N and C cycles) at Matanzas Lake over the last two
54
centuries Mean annual precipitation reconstructed and temperatures (von Gunten
et al 2009) Vertical gray bars indicate mega-droughts
5 CONCLUSIONS
Human activities have been the main factor controlling the N and C cycle in
the Laguna Matanzas during the last two centuries The N isotope signature in the
lake sediments reflects changes in the watershed fluxes to the lake but also in-lake
processes such as productivity and post-depositional changes Denitrification could
have been a dominant process during periods of increased anoxic conditions which
were more frequent prior to Great Acceleration (1950 CE) Higher δ15N and lower
δ13C values are associated with increased nutrient input from the watershed due to
increased livestock grazing and agriculture pressure (phase 1 Fig 7a) whereas
lower isotope values occurred during periods of increased forest plantations (phase
3 Fig 7c) During periods of increased lake productivity - such as in the last few
decades - δ15N values increased significantly
The most important change in C and N dynamics in the lake occurred after the
1950s (Phase 2 and 3) as tree plantations dominated the watershed Recent
changes in N dynamics can be explained by the higher nutrient contribution
associated with intensive agriculture (i e fertilizers) since the 1990s Although the
replacement of livestock activities with forestry and farming seems to have reduced
nutrient and soil export from the watershed to the lake the inefficient use of fertilizer
(by agriculture) can be the ultimate responsible for lake productivity increase during
the last decades
55
Figure 7 Schematic diagrams illustrating the main factors controlling the
isotope N signal in sediment OM of Laguna Matanzas N input from watershed
depends on human activities and land cover type Agriculture practices and cattle
(grassland development) contribute more N to the lake than native forest and
plantations Periods of higher productivity tend to deplete the dissolved inorganic N
in 14N resulting in higher δ15N (OM) The denitrification processes are more effective
in anoxic conditions associated with higher lake levels
6 METHODS
Short sediment cores were recovered from Laguna Matanzas using an Uwitec
gravity piston corer in 2011 and 2013 (Fig 1a) Sediment cores (MAT11-6A 129 cm
MAT13-2A 149 cm MAT13-3A 33 cm MAT13-4A 99 cm) were split
photographed sub-sampled and stored at the Pyrenean Institute of Ecology (IPE-
CSIC Spain) Core MAT11-6A was obtained from the central sector of the lake and
56
was selected for detailed multiproxy analyses (including elemental geochemistry C
and N isotope analyses XRF and 14C dating)
The isotope analyses (δ13C and δ15N) were performed at the Laboratory of
Biogeochemistry and Applied Stable Isotopes (LABASI-PUC Chile) using a Delta
V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via a
Conflo IV interface Isotope results are expressed in standard delta notation (δ) in
per mil (permil) relative to the standards Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Sediment samples
for δ13Corg were pre-treated with 50 ml of HCl and 50 ml of deionized water and
dried at 60ordmC for 4 hr to remove carbonates (Harris et al 2011)
Total Carbon (TC) Total Organic Carbon (TOC) Total Inorganic Carbon (TIC)
and Total Sulphur (TS) were measured every cm with a LECO SC 144 DR at IPE-
CSIC XRF measurements were carried out every 4 mm in MAT11-1A-1G core using
an AVAATECH X-ray Fluorescence II core scanner at the University of Barcelona
(Spain) Results are expressed as element intensities in counts per second (cps)
Tube voltage was operated at 30 kV and 10 kV to obtain the abundances of 15
elements (Al Si S Cl K Ca Ti V Mn Fe Br Rb Sr Y Zr) with an average at
least of 1600 cps (less for Br=1000)
Biogenic silica content mineralogy and grain size were measured every 4
cm Biogenic silica was measured following Mortlock and Froelich (1989) and
Bernaacuterdez et al (2005) using an Auto Analyzer Technicon AAII for dissolved silicate
analysis Mineralogy was analyzed with a Siemens D-500 X-ray diffractometer (Cu
kα 40 kV 30 mA graphite monochromator) at the ICTJA-CSIC (Spain) Grain size
analyses were performed in a Beckmann Coulter LS 13 320 Particle Size Analyzer
57
at the IPE-CSIC The samples were classified according to textural classes as
follows clay (lt2 μm) silt (20-2 microm) and sand (gt2 microm) fractions
The age-depth model for the Laguna Matanzas sedimentary sequence was
constructed using 210Pb and 137Cs dating techniques (MAT13-4A) as well as two 14C
AMS dates on bulk sediment samples (MAT11-6A fig 2) We dated the dissolved
inorganic carbon (DIC) in the water column and no significant reservoir effect is
present in the modern-day water column (10454 + 035 pcmc Table 2) An age-
depth model was obtained with the Bacon R package to estimate the deposition
rates and associated age uncertainties along the core (Blaauw and Christen 2011)
To estimate LUCC of Laguna Matanzasacutes watershed we use satellite images
Landsat MSS for 1975 Landsat TM for 1989 and Landsat OLI for 2016 all taken in
summer or autumn (Table 1) We performed supervised classification of land uses
(maximum likelihood algorithm) for each year (1975 1989 and 2016) and the results
were mapped using software ArcGIS 102 in 2017
Satellite Images Acquisition Date
Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat OLI 20160404 30 m
Table 1 Landsat imagery
Surface water samples were filtered for obtained particulate organic matter In
addition soil samples from the main land usecover present in the Laguna Matanzas
watershed were collected Elemental C N and their corresponding isotopes from
POM and soil were obtained at the LABASI and used here as end members
Daily precipitation at Santo Domingo (33ordm39`S 71ordm3 6`W)- the nearest weather
station to Laguna Matanzasndash was compiled using the redPrec R package (Fig1 d
Serrano-Notivoli et al 2017 Sarricolea et al 2019) To extend our precipitation
58
reconstruction back to 1824 we correlated this dataset with that available for
Santiago The Santiago data was compiled from data published in the Anales of
Universidad of Chile (1850) for the years 1824 to 1849 Almeyda (1940) for the years
1849 to 1864 and the Quinta Normal series from 1866 to the present (Direccioacuten
Meteoroloacutegica de Chile) We generated a linear regression model between the
presentday Santo Domingo station and the compiled Santiago data with a Pearson
coefficient of 087 and p-valuelt 001
Acknowledgments This research was funded by grants CONICYT AFB170008
to the Institute of Ecology and Biodiversity (IEB) and FONDECYT 1150763 (to CL)
Doctoral grant Becas Chile 21150224 MEDLANT (Spanish Ministry of Economy
and Competitiveness grant CGL2016-76215-R) Additional funding was provided
by the Laboratorio Internacional de Cambio Global (LINCGlobal PUC-CSIC) We
thank R Lopez E Royo and M Gallegos for help with sample analyses We thank
the Laboratory of Biogeochemistry and Applied Stable Isotopes (LABASI) of the
Department of Ecology (PUC) for sample analyses
References
Albert F 1900 Las dunas o sean arenas volantes voladeros arenas muertas invasion de las arenas playas I meacutedanos del centro de Chile comprendiendo el litoral desde el liacutemite norte de la provincia de Aconcagua hasta el liacutemite sur de la de Arauco Memorias cientiacuteficas I Lit
Anderson NJ W D Fritz SC 2009 Holocene carbon burial by lakes in SW
Greenland Glob Chang Biol httpsdoiorg101111j1365-2486200901942x
Armesto J Villagraacuten C Donoso C 1994 La historia del bosque templado
chileno Ambient y Desarro 66 66ndash72 httpsdoiorg101007BF00385244
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A
59
historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Arnell NW Gosling SN 2013 The impacts of climate change on river flow
regimes at the global scale J Hydrol 486 351ndash364 httpsdoiorg101016JJHYDROL201302010
Bernaacuterdez P Prego R Franceacutes G Gonzaacutelez-Aacutelvarez R 2005 Opal content in
the Riacutea de Vigo and Galician continental shelf Biogenic silica in the muddy fraction as an accurate paleoproductivity proxy Cont Shelf Res httpsdoiorg101016jcsr200412009
Blaauw M Christen JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474 httpsdoiorg10121411-BA618
Brush GS 2009 Historical land use nitrogen and coastal eutrophication A
paleoecological perspective Estuaries and Coasts 32 18ndash28 httpsdoiorg101007s12237-008-9106-z
Camarero L Catalan J 2012 Atmospheric phosphorus deposition may cause
lakes to revert from phosphorus limitation back to nitrogen limitation Nat Commun httpsdoiorg101038ncomms2125
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego
R Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and environmental change from a high Andean lake Laguna del Maule central Chile (36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Catalan J Pla-Rabeacutes S Wolfe AP Smol JP Ruumlhland KM Anderson NJ
Kopaacuteček J Stuchliacutek E Schmidt R Koinig KA Camarero L Flower RJ Heiri O Kamenik C Korhola A Leavitt PR Psenner R Renberg I 2013 Global change revealed by palaeolimnological records from remote lakes A review J Paleolimnol httpsdoiorg101007s10933-013-9681-2
Chen Y Y Huang W Wang W H Juang J Y Hong J S Kato T amp
Luyssaert S (2019) Reconstructing Taiwanrsquos land cover changes between 1904 and 2015 from historical maps and satellite images Scientific Reports 9(1) 3643 httpsdoiorg101038s41598-019-40063-1
Contreras S Werne J P Araneda A Urrutia R amp Conejero C A (2018)
Organic matter geochemical signatures (TOC TN CN ratio δ 13 C and δ 15 N) of surface sediment from lakes distributed along a climatological gradient on the western side of the southern Andes Science of The Total Environment 630 878-888 httpsdoiorg101016jscitotenv201802225
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia UC
60
de V 2014 ELEMENTOS DE LA HISTORIA NATURAL DEL An Mus Hist Natulas Vaplaraiso 27 51ndash67
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-010-9453-1
Diebel M Vander Zanden MJ 201209 Nitrogen stable isotopes in streams
stable isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19 1127ndash1134 httpsdoiorg10189008-03271
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506 Espizua LE Pitte P 2009 The Little Ice Age glacier advance in the Central
Andes (35degS) Argentina Palaeogeogr Palaeoclimatol Palaeoecol 281 345ndash350 httpsdoiorg101016JPALAEO200810032
Figueroa JA Castro S A Marquet P A Jaksic FM 2004 Exotic plant
invasions to the mediterranean region of Chile causes history and impacts Invasioacuten de plantas exoacuteticas en la regioacuten mediterraacutenea de Chile causas historia e impactos Rev Chil Hist Nat 465ndash483 httpsdoiorg104067S0716-078X2004000300006
Fowler D Coyle M Skiba U Sutton MA Cape JN Reis S Sheppard
LJ Jenkins A Grizzetti B Galloway N Vitousek P Leach A Bouwman AF Butterbach-bahl K Dentener F Stevenson D Amann M Voss M 2013 The global nitrogen cycle in the twenty- first century Philisophical Trans R Soc B httpsdoiorghttpdxdoiorg101098rstb20130164
Frank D Reichstein M Bahn M Thonicke K Frank D Mahecha MD
Smith P van der Velde M Vicca S Babst F Beer C Buchmann N Canadell JG Ciais P Cramer W Ibrom A Miglietta F Poulter B Rammig A Seneviratne SI Walz A Wattenbach M Zavala MA Zscheischler J 2015 Effects of climate extremes on the terrestrial carbon cycle Concepts processes and potential future impacts Glob Chang Biol httpsdoiorg101111gcb12916
Fritz SC Anderson NJ 2013 The relative influences of climate and catchment
processes on Holocene lake development in glaciated regions J Paleolimnol httpsdoiorg101007s10933-013-9684-z
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
61
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS) implications for past sea level and environmental variability J Quat Sci 32 830ndash844 httpsdoiorg101002jqs2936
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892 httpsdoiorg101126science1136674
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924 httpsdoiorg104319lo20095430917
Garciacutea-Ruiz JM 2010 The effects of land uses on soil erosion in Spain A
review Catena httpsdoiorg101016jcatena201001001
Garciacutea-Ruiz JM Loacutepez-Moreno JI Lasanta T Vicente-Serrano SM
Gonzaacutelez-Sampeacuteriz P Valero-Garceacutes BL Sanjuaacuten Y Begueriacutea S Nadal-Romero E Lana-Renault N Goacutemez-Villar A 2015 Los efectos geoecoloacutegicos del cambio global en el Pirineo Central espantildeol una revisioacuten a distintas escalas espaciales y temporales Pirineos httpsdoiorg103989Pirineos2015170005
Garciacutea-Ruiz JM Nadal-Romero E Lana-Renault N Begueriacutea S 2013
Erosion in Mediterranean landscapes Changes and future challenges Geomorphology 198 20ndash36 httpsdoiorg101016jgeomorph201305023
Garreaud RD Vuille M Compagnucci R Marengo J 2009 Present-day
South American climate Palaeogeogr Palaeoclimatol Palaeoecol httpsdoiorg101016jpalaeo200710032
Gayo EM Latorre C Jordan TE Nester PL Estay SA Ojeda KF
Santoro CM 2012 Late Quaternary hydrological and ecological changes in the hyperarid core of the northern Atacama Desert (~21degS) Earth-Science Rev httpsdoiorg101016jearscirev201204003
Ge Y Zhang K amp Yang X (2019) A 110-year pollen record of land use and land
cover changes in an anthropogenic watershed landscape eastern China Understanding past human-environment interactions Science of The Total Environment 650 2906-2918 httpsdoiorg101016jscitotenv201810058
Goyette J Bennett EM Howarth RW Maranger R 2016 Global
Biogeochemical Cycles 1000ndash1014 httpsdoiorg1010022016GB005384Received
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and
oxygen isotope fractionation during dissimilatory nitrate reduction by
62
denitrifying bacteria 53 2533ndash2545 httpsdoiorg10230740058342
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
Gurnell AM Petts GE Hannah DM Smith BPG Edwards PJ Kollmann
J Ward J V Tockner K 2001 Effects of historical land use on sediment yield from a lacustrine watershed in central Chile Earth Surf Process Landforms 26 63ndash76 httpsdoiorg1010021096-9837(200101)261lt63AID-ESP157gt30CO2-J
Heathcote A J et al Large increases in carbon burial in northern lakes during the
Anthropocene Nat Commun 610016 doi 101038ncomms10016 (2015) Hollander DJ Smith MA 2001 Microbially mediated carbon cycling as a
control on the δ13C of sedimentary carbon in eutrophic Lake Mendota (USA) New models for interpreting isotopic excursions in the sedimentary record Geochim Cosmochim Acta httpsdoiorg101016S0016-7037(00)00506-8
Holtgrieve GW Schindler DE Hobbs WO Leavitt PR Ward EJ Bunting
L Chen G Finney BP Gregory-Eaves I Holmgren S Lisac MJ Lisi PJ Nydick K Rogers LA Saros JE Selbie DT Shapley MD Walsh PB Wolfe AP 2011 A coherent signature of anthropogenic nitrogen deposition to remote watersheds of the Northern Hemisphere Science (80) 334 1545ndash1548 httpsdoiorg101126science1212267
Hooper DU Adair EC Cardinale BJ Byrnes JEK Hungate BA Matulich
KL Gonzalez A Duffy JE Gamfeldt L Connor MI 2012 A global synthesis reveals biodiversity loss as a major driver of ecosystem change Nature httpsdoiorg101038nature11118
Horvatinčić N Sironić A Barešić J Sondi I Krajcar Bronić I Borković D
2016 Mineralogical organic and isotopic composition as palaeoenvironmental records in the lake sediments of two lakes the Plitvice Lakes Croatia Quat Int httpsdoiorg101016jquaint201701022
Howarth RW 2004 Human acceleration of the nitrogen cycle Drivers
consequences and steps toward solutions Water Sci Technol httpsdoiorg1010382Fscientificamerican0490-56
Jenny B Valero-Garceacutes BL Urrutia R Kelts K Veit H Appleby PG Geyh
M 2002 Moisture changes and fluctuations of the Westerlies in Mediterranean Central Chile during the last 2000 years The Laguna Aculeo record (33deg50primeS) Quat Int 87 3ndash18 httpsdoiorg101016S1040-
63
6182(01)00058-1
Jenny B Wilhelm D Valero-Garceacutes BL 2003 The Southern Westerlies in
Central Chile Holocene precipitation estimates based on a water balance model for Laguna Aculeo (33deg50primeS) Clim Dyn 20 269ndash280 httpsdoiorg101007s00382-002-0267-3
Leng M J Lamb A L Heaton T H Marshall J D Wolfe B B Jones M D
amp Arrowsmith C 2006 Isotopes in lake sediments In Isotopes in palaeoenvironmental research (pp 147-184) Springer Dordrecht
Le Roux JP Nielsen SN Kemnitz H Henriquez Aacute 2008 A Pliocene mega-
tsunami deposit and associated features in the Ranquil Formation southern Chile Sediment Geol 203 164ndash180 httpsdoiorg101016jsedgeo200712002
Lehmann M Bernasconi S Barbieri a McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66 3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Lenihan JM Drapek R Bachelet D Neilson RP 2003 Climate change
effects on vegetation distribution carbon and fire in California Ecol Appl httpsdoiorg101890025295
McLauchlan K K Gerhart L M Battles J J Craine J M Elmore A J
Higuera P E amp Perakis S S (2017) Centennial-scale reductions in nitrogen availability in temperate forests of the United States Scientific reports 7(1) 7856 httpsdoi101038s41598-017-08170-z
Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32ordmS 3050 masl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
Martiacuten-Foreacutes I Casado MA Castro I Ovalle C Pozo A Del Acosta-Gallo
B Saacutenchez-Jardoacuten L Miguel JM De 2012 Flora of the mediterranean basin in the chilean ESPINALES Evidence of colonisation Pastos 42 137ndash160
Marzecovaacute A Mikomaumlgi a Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54 httpsdoiorg103176eco2011105
Matesanz S Valladares F 2014 Ecological and evolutionary responses of
Mediterranean plants to global change Environ Exp Bot httpsdoiorg101016jenvexpbot201309004
64
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Menzel P Gaye B Wiesner MG Prasad S Stebich M Das BK Anoop A
Riedel N Basavaiah N 2013 Influence of bottom water anoxia on nitrogen isotopic ratios and amino acid contributions of recent sediments from small eutrophic Lonar Lake central India Limnol Oceanogr 58 1061ndash1074 httpsdoiorg104319lo20135831061
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Michelutti N Wolfe AP Cooke CA Hobbs WO Vuille M Smol JP 2015
Climate change forces new ecological states in tropical Andean lakes PLoS One httpsdoiorg101371journalpone0115338
Muntildeoz-Rojas M De la Rosa D Zavala LM Jordaacuten A Anaya-Romero M
2011 Changes in land cover and vegetation carbon stocks in Andalusia Southern Spain (1956-2007) Sci Total Environ httpsdoiorg101016jscitotenv201104009
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Poraj-Goacuterska A I Żarczyński M J Ahrens A Enters D Weisbrodt D amp
Tylmann W (2017) Impact of historical land use changes on lacustrine sedimentation recorded in varved sediments of Lake Jaczno northeastern Poland Catena 153 182-193 httpdxdoiorg101016jcatena201702007
Pongratz J Reick CH Raddatz T Claussen M 2010 Biogeophysical versus
biogeochemical climate response to historical anthropogenic land cover change Geophys Res Lett 37 1ndash5 httpsdoiorg1010292010GL043010
Prudhomme C Giuntoli I Robinson EL Clark DB Arnell NW Dankers R
Fekete BM Franssen W Gerten D Gosling SN Hagemann S Hannah DM Kim H Masaki Y Satoh Y Stacke T Wada Y Wisser D 2014 Hydrological droughts in the 21st century hotspots and uncertainties from a global multimodel ensemble experiment Proc Natl Acad Sci httpsdoiorg101073pnas1222473110
65
Ruumlhland KM Paterson AM Smol JP 2015 Lake diatom responses to
warming reviewing the evidence J Paleolimnol httpsdoiorg101007s10933-
015-9837-3
Sarricolea P Meseguer-Ruiz Oacute Serrano-Notivoli R Soto M V amp Martin-Vide
J (2019) Trends of daily precipitation concentration in Central-Southern Chile Atmospheric research 215 85-98 httpsdoiorg101016jatmosres201809005
Sernageomin 2003 Mapa geologico de chile version digital Publ Geol Digit Serrano-Notivoli R de Luis M Begueriumlia S 2017 An R package for daily
precipitation climate series reconstruction Environ Model Softw httpsdoiorg101016jenvsoft201611005
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Stine S 1994 Extreme and persistent drought in California and Patagonia during
mediaeval time Nature 369 546ndash549 httpsdoiorg101038369546a0
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL
Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Syvitski JPM Voumlroumlsmarty CJ Kettner AJ Green P 2005 Impact of humans
on the flux of terrestrial sediment to the global coastal ocean Science 308(5720) 376-380 httpsdoiorg101126science1109454
Thomas RQ Zaehle S Templer PH Goodale CL 2013 Global patterns of
nitrogen limitation Confronting two global biogeochemical models with observations Glob Chang Biol httpsdoiorg101111gcb12281
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Urbina Carrasco MX Gorigoitiacutea Abbott N Cisternas Vega M 2016 Aportes a
la historia siacutesmica de Chile el caso del gran terremoto de 1730 Anuario de Estudios Americanos httpsdoiorg103989aeamer2016211
Vargas G Ortlieb L Chapron E Valdes J Marquardt C 2005 Paleoseismic
inferences from a high-resolution marine sedimentary record in northern Chile
66
(23degS) Tectonophysics httpsdoiorg101016jtecto200412031 399(1-4) 381-398httpsdoiorg101016jtecto200412031
Valero-Garceacutes BL Latorre C Morelliacuten M Corella P Maldonado A Frugone M Moreno A 2010 Recent Climate variability and human impact from lacustrine cores in central Chile 2nd International LOTRED-South America Symposium Reconstructing Climate Variations in South America and the Antarctic Peninsula over the last 2000 years Valdivia p 76 (httpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-yearshttpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-years
Vicente-Serrano SM Gouveia C Camarero JJ Begueria S Trigo R
Lopez-Moreno JI Azorin-Molina C Pasho E Lorenzo-Lacruz J Revuelto J Moran-Tejeda E Sanchez-Lorenzo A 2013 Response of vegetation to drought time-scales across global land biomes Proc Natl Acad Sci httpsdoiorg101073pnas1207068110
Villa Martiacutenez R P 2002 Historia del clima y la vegetacioacuten de Chile Central
durante el Holoceno Una reconstruccioacuten basada en el anaacutelisis de polen sedimentos microalgas y carboacuten PhD
Villa-Martiacutenez R Villagraacuten C Jenny B 2003 Pollen evidence for late -
Holocene climatic variability at laguna de Aculeo Central Chile (lat 34o S) The Holocene 3 361ndash367
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Human alteration of the global nitrogen cycle sources and consequences Ecological applications 7(3) 737-750 httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Rein B Urrutia R amp Appleby P (2009) A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 The Holocene 19(6) 873-881 httpsdoiorg1011770959683609336573
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Widory D Kloppmann W Chery L Bonnin J Rochdi H Guinamant JL
2004 Nitrate in groundwater An isotopic multi-tracer approach J Contam Hydrol 72 165ndash188 httpsdoiorg101016jjconhyd200310010
67
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
68
Supplementary material
Facie Name Description Depositional Environment
F1 Organic-rich
mud
Massive to banded black
organic - rich (TOC up to 14 )
mud with aragonite in dm - thick
layers Slightly banded intervals
contain less OM (TOClt4) and
aragonite than massive
intervals High MnFe (oxic
bottom conditions) High CaTi
BrTi and BioSi (up to 5)
Distal low energy environment
high productivity well oxygenated
and brackish waters and relative
low lake level
F2 Massive to
banded silty clay
to fine silt
cm-thick layers mostly
composed by silicates
(plagioclase quartz cristobalite
up to 65 TOC mean=23)
Some layers have relatively high
pyrite content (up to 25) No
carbonates CaTi BrTi and
BioSi (mean=48) are lower
than F1 higher ZrTi (coarser
grain size)
Deposition during periods of
higher sediment input from the
watershed
69
F3 Banded to
laminated light
brown silty clay
cm-thick layers mostly
composed of clay minerals
quartz and plagioclase (up to
42) low organic matter
(TOC mean=13) low pyrite
and BioSi content
(mean=46) and some
aragonite
Flooding events reworking
coastal deposits
F4 Laminated
coarse silts
Thin massive layers (lt2mm)
dominated by silicates Low
TOC (mean=214 ) BrTi
(mean=002) MnFe (lt02)
TIC (lt034) BioSi
(mean=46) and TS values
(lt064) and high ZrTi
Rapid flooding events
transporting material mostly
from within the watershed
F5 Breccia with
coarse silt
matrix
A 17 cm thick (80-97 cm
depth) layer composed by
irregular mm to cm-long ldquosoft-
clastsrdquo of silty sediment
fragments in a coarse silt
matrix Low CaTi BrTi and
MnFe ratios and BioSi
Rapid high energy flood
events
70
(mean=43) and high ZrTi
(gt018)
Table Sedimentological and compositional characteristics of Laguna Matanzas
facies
71
CAPIacuteTULO 2 STABLE ISOTOPES TRACK LAND USE AND COVER
CHANGES IN A MEDITERRANEAN LAKE IN CENTRAL CHILE OVER THE
LAST 600 YEARS
72
Stable isotopes track land use and cover changes in a mediterranean lake in
central Chile over the last 600 years
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugone-Aacutelvarezabc Pablo
Sarricolead Leonardo Villaciacutese M Laura Carrevedob Blas Valero-Garceacutescg
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Departamento de Ecologiacutea Universidad de ChileLas Palmeras 3425 Nuntildeoa Santiago Chile
f Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding authors E-mail addresses clatorrebiopuccl magdalenafuentealbagmailcom
Key words Anthropocene lake sediment δsup1⁵N δsup1sup3C Great Acceleration organic
geochemistry watershedndashlake system Stable Isotope Analyses land usecover
change Nitrogen cycle mediterranean ecosystems central Chile
73
Abstract
Nutrient fluxes in many aquatic ecosystems are currently being overridden by
anthropic controls especially since the industrial revolution (mid-1800s) and the
Great Acceleration (mid-1900s) Land use and cover changes (LUCC) alter the
availability and fluxes of nutrients such as nitrogen that are transferred via runoff
and groundwater into lakes By altering lake productivity and trophic status these
changes are often preserved in the sedimentary record Here we use
geolimnological proxies and stable isotope analyses (δsup1⁵N δsup13C) on lake sediments
to reconstruct the lake-watershed nutrient transfer in a central Chilean lake (Lago
Vichuqueacuten) over the past 600 years We compare our multiproxy record from recent
lake sediments to the soilvegetation relationship across the watershed as well as
land usecover changes from 1975 to 2014 derived from satellite images Our results
show that lake sediment δsup1⁵N values increased with meadow cover but decreased
with tree plantations suggesting increased nitrogen retention when trees dominate
the watershed Our δsup1⁵N record from central Chile can thus be interpreted as a proxy
for nutrient availability over the last 600 years mainly derived from land use changes
coupled with climate drivers Although variable sources of organic matter and in situ
fractionation often hinder straightforward environmental interpretations of stable N
isotope signatures our study shows that lake sediment δsup1⁵N can be a useful tool for
assessing the contribution of past human activities in nutrient and nitrogen cycling
1 Introduction
Nitrogen (N) is a limiting nutrient in terrestrial and aquatic ecosystems (Vitousek
et al 1997) Changes in its availability can drive eutrophication and increase
pollution in these ecosystems (McLauchlan et al 2013) Although recent human
74
impacts on the global N cycle have been significant the consequences of increased
anthropogenic N on these ecosystems are unclear (Gerhart and Mclauchlan 2014
Battye et al 2017) Isotopic signatures are key to understand N dynamics in lakes
nevertheless in situ andor diagenetic fractionation along with multiple sources of
organic matter (OM) often hinder straightforward environmental interpretations from
isotopes Monitoring δ15N and δ13C values as components of the N cycle
specifically those related to the link between terrestrial and aquatic ecosystems can
help differentiate between effects from processes versus sources in stable isotope
values (eg from Particulate Organic Matter -POM- soil and vegetation) and
improve how we interpret variations in δ15N (and δ13C) values at longer temporal
scales
The main processes controlling stable N isotopes in bulk lake OM are source
lake productivity diagenesis and denitrification (Talbot 2001 Leng et al 2006
Fariacuteas et al 2009 Torres et al 2012 Brahney et al 2014) N sources depend on
contributions from the watershed (ie soil and biomass) the transfer of atmospheric
N to the lake (ie atmospheric N2 fixation) and nutrient recycling (Leng et al 2006)
Atmospheric N2 (isotopically defined as δ15N =0) is fixed by cyanobacteria with
minimal fractionation resulting in δ15NOM values that fluctuate from -2 to +2permil (Fogel
and Cifuentes 1993 Talbot 2001 Elliot and Brush 2006) Besides N2 fixation by
cyanobacteria phytoplankton species assimilate dissolved inorganic nitrogen (DIN)
and values typically range from +7 to +10permil (Xu et al 2016 Meyers et al 2003) In
addition seasonal changes in POM occur in the lake water column Gu et al (2006)
sampled the water column of Lake Wauberg (Florida USA) during a hydrologic year
and found a higher development of N fixing species during the summer A major
factor behind this increase are human activities in the watershed which control the
75
inputs of the sediment and nutrients to the lake (Woodward et al 2011) Some
studies have shown higher δ15N values in lake sediments from watersheds that are
highly affected by human activities (Bruland and Mackenzie 2010 Woodward et al
2011) Syntenic fertilizers displayed δ15N values between -4 to +4permil cattle manure
around +8 to +22permil and surface and groundwater OM between 0 to +10permil (Elliott
and Brush 2006 Leng et al 2006) Although relatively low δ15N values from
fertilizers constitute major N input to human-altered watersheds the elevated loss
of 14N via volatilization of ammonia and denitrification leaves the remaining total N
input enriched in 15N (Bruland and Mackenzie 2010)
In addition to the different sources and variations in lake productivity early
diagenesis at the sedimentndashwater interface in the sediment can further alter
sediment OM isotope values (Brahney et al 2014 Galman et al 2009) During
diagenetic loss of N in lacustrine systems kinetic fractionation occurs and the
remaining N is enriched in 15N leading to higher δ15N values (Galman et al 2006
Talbot 2001) Denitrification tends to enrich lake sediments in 15N due to the
assimilation of nitrates by denitrifying bacteria (Granger et al 2008) and is more
prevalent in anoxic environments (Brahney et al 2016 Lehman et al 2002)
Carbon isotopes in lake sediments can also provide useful information about
paleoenvironmental changes OM origin and depositional processes (Meyers et al
2003) Allochthonous organic sources (high CN ratios) produce isotope values
similar to values from catchment vegetation Autochthonous organic matter (low CN
ratio) is influenced by fractionation both in the lake and the watershed leading up to
carbon assimilation by lacustrine biota (Woodward et al 2011) Changes in
productivity greatly affect δ13COM values (Torres et al 2012) as during C uptake
plankton preferentially assimilate 12C leaving the dissolved inorganic carbon (DIC)
76
pool enriched in 13C (Gu et al 2006) with phytoplankton averaging ca 20 permil lower
than the respective δ13CDIC values (Leng et al 2006) Under conditions of low to
moderate primary productivity plankton preferentially uptake the lighter 12C
resulting in low δ13C values displayed by the OM (Torres et al 2012) Conversely
during high primary productivity phytoplankton will uptake 12C until its depletion and
is then forced to assimilate the heavier isotope resulting in an increase in δ13C
values Higher productivity in C-limited lakes due to slow water-atmosphere
exchange of CO2 also results in high δ13C values (Galman et al 2009) In these
cases algae are forced to uptake dissolved bicarbonate with δ13C values between
7 to 10permil higher than those of the atmospheric CO2 dissolved in water (Sun et al
2016 Torres et al 2012 Galman et al 2009)
Stable isotope analyses from lake sediments are thus useful tools to
reconstruct shifts in lake-watershed dynamics caused by changes in limnological
parameters and LUCC Our knowledge of the current processes that can affect
stable isotope signals in a watershed-lake system is limited however as monitoring
studies are scarce Besides in order to use stable isotope signatures to reconstruct
past environmental changes we require a multiproxy approach to understand the
role of the different variables in controlling these values Hence in this study we
carried out a detailed survey of current N dynamics in a coastal central Chilean lake
(Lago Vichuqueacuten) and a multiproxy study of a sediment record spanning the last
600 years The characterization of the recent changes in the watershed since 1970s
is based on satellite images to compare recent changes in the lake and assess how
these are related with climate variability and an ever increasing human footprint
(Lara et al 2012 Gallardo et al 2015 Steffen et al 2015) Our main goal is to
investigate how stable isotope values from lake sediment reflect changes in the lake
77
ndash watershed system during periods of high watershed disruption (eg Spanish
Conquest late XIX century Great Acceleration) and recent climate change (eg
Little Ice Age and current global warming)
2 Study Site
Lago Vichuqueacuten (34ordm49`S 72ordm04W 4 m asl 30 m of depth Fig 1) is a
mesotrophic to eutrophic lake with dominant anoxic bottom conditions that is
stratified year around (Frugone-Aacutelvarez et al 2017) The main inlets are the
Vichuqueacuten and Huintildee rivers and the only outlet is the Llico estuary that drains into
the Pacific Ocean High tides can sporadically shift the flow direction of the Llico
estuary which increases the marine influence in the lake Dune accretion gradually
limited ocean-lake connectivity until the estuary was almost completely closed off
by ca 10 ky BP and the current lake regime formed (Frugone-Aacutelvarez et al 2017)
The area is characterized by a mediterranean climate with cold-wet winters and
hot-dry summers and an annual precipitation of ~650 mm and a mean annual
temperature of 15ordmC During the austral winter months (June - August) precipitation
is modulated by north-west shifts of the South Pacific Anticyclone (SPA) followed by
an increased frequency of storm fronts stemming off the South Westerly Winds
(SWW) A strengthened SPA during austral summers (December - March) which
are typically dry and warm blocks the northward migration of storm tracks stemming
off the SWW (Fig1 Frugone-Aacutelvarez et al 2017)
78
Figure 1 a) The location of the Lago Vichuqueacuten in central Chile b) Map of land
uses and cover in 2016 c) Ombrothermic diagram mediterranean climates are
characterized by cold-wet winters with surplus moisture from June to August and
hot-dry summers d) Lake bathymetry showing location of cores and water sampling
sites used in this study
Although major land cover changes in the area have occurred since 1975 to the
present as the native forests were replaced by tree (Monterey pine and eucalyptus)
plantations the region was settled before the Spanish conquest (Frugone-Alvarez
et al 2018) Historical documents indicate that Vichuqueacuten was occupied by a
Mitimaes colony as part of the Inka empire (Odone 1998) Unlike other Andean
areas during the Inka period the introduction of corn and quinoa in the Vichuqueacuten
watershed do not seem to have intensified land use The Spanish colonial period in
Chile lasted from 1542 CE to the independence in 1810 CE The first historical
document (1550 CE) shows that the areas around Vichuqueacuten were settled by the
Spanish early on as the Vichuqueacuten watershed was given as an ldquoencomiendardquo
system to Juan Cuevas The ldquoencomiendardquo system provided the owners with land
79
and indigenous people to work but also the introduction of wheat wine cattle
grazing and logging of native forests for lumber extraction and increasing land for
agricultural activities (Odone et al 1998 Armesto et al 2010) During the 19th
century (the Republic) the export of wheat to Australia and Canada generated
intensive changes in land cover use The town of Vichuqueacuten became the regional
capital and plans to build a maritime port in the bay of Vichuqueacuten were drawn
However the fall of international markets in 1880 paralyzed these plans During the
20th century the plantation of trees (Pinus radiata and Eucalyptus globulus) in areas
cleared of native forests was subsidized by two national laws DFL nordm265 (1931) and
DFL nordm 701 (1974) both of which provided funds for such plantations During the last
decades the urbanization with summer vacation homes along the shorelines of
Lago Vichuqueacuten has greatly increased Extensive lake eutrophication is currently a
large environmental problem (EULA 2008)
3 Methods
Coring campaigns were organized from 2011 to 2015 (Fig 1d) We recovered
12 short cores and 4 long cores (up to 14 m long) at several sites using a hammer-
modified UWITEC gravity corer Sediment cores (VIC11-2A 170 cm VIC13-2B 170
cm VIC13-2D 1200 cm and VIC15-1A 119 cm) were split photographed sub-
sampled and stored in the Pyrenean Institute of Ecology (IPE-CSIC Spain) Core
VIC13-2B was selected for detailed multiproxy analyses (including elemental
geochemistry C and N isotopes XRF and 14C dating) Stable isotope analyses
(δ13C δ15N) were performed at the Laboratory of Biogeochemistry and Applied
Stable Isotopes (LABASI-PUC Chile) Sediment samples for δ13Corg were pre-
treated with 50 ml of HCl and 50 ml of deionized water and dried at 60ordmC for 4 hr to
remove carbonates (Harris et al 2011) Isotope analyses were conducted using a
80
Delta V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via
a Conflo IV interface Isotope results are expressed in standard delta notation (δ)
and expressed in per mil (permil) relative to the Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Total carbon (TC)
Total Organic Carbon (TOC) Total Inorganic Carbon (TIC) and Total Sulfur (TS)
were measured every 2 cm with a LECO SC 144 DR at the IPE-CSIC
An AVAATECH X-Ray Fluorescence II core scanner with a Rh X-ray tube from
the University of Barcelona was used to obtain XRF logs every 4 mm of resolution
Results are expressed as element intensities in counts per second (cps) Tube
voltage was operated at 30 kV and 10 kV to obtain the abundances of 18 elements
(Pb Al Si S Cl K Ca Ti V Mn Fe Co Zn Br Rb Sr Y Zr) with an average of
at least of 1800 cps (less for Pb=437 and Zn=934 cps) Pb was included due to
similar behavior with Co and Fe Element ratios were calculated to describe changes
in redox conditions (MnFe) endogenic productivity (BrTi) carbonate formation
(SrCa and CaTi) and sediment input (ZrTi) (Barreiro-Lostres et al 2014
Carrevedo et al 2015 Frugone-Aacutelvarez et al 2017 Kylander et al 2011 Moreno
et al 2007a)
Several campaigns were carried out to sample the POM from the water column
two per hydrologic year from November 2015 to August 2018 A liter of water was
recovered in three sites through to the lake two are from the shallower areas (with
samples taken at 2 and 5 m depth at each site) and one in the deeper central portion
(at 2 5 and 20 m depth) of the lake (Fig 1d) The samples were filtered using glass
fiber filters (25 m) and then frozen to obtain the stable carbon and nitrogen isotope
signal of lacustrine POM Additionally soil and vegetation samples from the
following communities native species meadow hydrophytic vegetation and
81
Monterrey pine (Pinus radiata) were taken for stable isotope analyses (see in
supplementary material)
The age model for the complete Lago Vichuqueacuten sedimentary sequence is
based on Frugone-Aacutelvarez et al (2017) with the last 1000 years based on
210Pb137Cs dating techniques in VIC15-1A and 14C AMS dates on bulk sediment
samples (Supplementary Table S1) The 14C measurements of lake water DIC show
a moderate reservoir effect of 180 plusmn 25 14C yr) The revised age-depth model used
here includes three more 14C AMS dates performed with the program Bacon to
establish the deposition rates along the core (Blaauw and Christen 2011 Fig 2)
The age-depth model indicates that average resolution between 0 to 87 cm is lt2
cm per year and from 88 to 170 cm it is lt47 cm per year
82
Figure 2 Chronological age-depth model for the Lago Vichuqueacuten sedimentary
sequence spanning the last 1000 yr (see also Frugone-Alvarez et al 2017)
To estimate land use changes in the watershed we use Landsat MSS images
for 1975 and Landsat TM for 1989 and 2014 all taken in either summer or autumn
(Table 1) We performed supervised classification of land uses (maximum likelihood
83
algorithm) for each year (1975 1989 and 2014) and results were mapped using
ArcGIS 102
Table 1 Images using for LUCC reconstruction
Source of LUCC
Acquisition
Date Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat TM 19991226 30 m
CONAF 2009 30 m
Land cover Chile 2014 30 m
CONAF 2016 30 m
Previous Work on Lago Vichuqueacuten sedimentary sequence
The sediments are organic-poor dark brown to brown laminated silt with some
intercalated thin coarser clastic layers Lacustrine facies have been classified
according to elemental composition (TOC TS TIC and TN) grain size and
sedimentary textures (Frugone-Aacutelvarez et al 2017) Four main types of lacustrine
facies were identified in this short core Facies L1 is a laminated (1cm) black to dark
brown diatom-poor silt with relatively higher TOC percentages (TOC about 185)
TS (about of 18) and TOCTN (about 92) Facies L2 is characterized by a
homogeneous black organic-poor silty clay with low TOC (mean= 13) TS (mean=
13) TOCTN (mean= 88) values Facies L3 is a laminated (1cm) black organic-
poor (TOC mean 14) silts with higher TS (mean= 21) and TOCTN ratios
(mean= 88) Facies L3 is interpreted as ldquobaselinerdquo deposition on the distal areas
84
of Vichuqueacuten lake Mineralogy is dominated by mica (muscovite) clay minerals
(chlorite) and plagioclase (albite) Quartz and calcite occur at the top and pyrite
occurs in the lower part of the sequence Facies T is composed by massive banded
sediments 4 cm thick and coarse grain size It is interpreted as an instantaneous
depositional event (for more detail see Frugone-Aacutelvarez et al 2017) In this work
we identified four subunits based on geochemical and stable isotope signals
4 Results
41 Geochemistry and PCA analysis
High positive correlations exist between Al Si K and Ti (r = 078 ndash 096
supplementary Fig S1) and between Zn Zr Rb and Y (r = 072 - 096) which reflect
the silicate content of the watershed-derived sediments (Marzecoacuteva et al 2011) Zr
is commonly associated with minerals more abundant in coarser deposits Thus the
ZrTi profile could reflect grain size (Cuven et al 2010) showing higher variability
in the upper part of the Lake Vichuqueacuten sequence and in the alternation between
facies L1 and L2 Another group of elements made up of Co Fe and Pb displayed
positive correlations (r = 067ndash 097) and represents the input of heavy metals Br
Cl and Ca show a positive correlation (r = 049 ndash 06 p-value = 00001) BrTi ratio
is interpreted as a productivity indicator due to Br having a strong affinity with humic
and organic compounds (Marzecoacuteva et al 2011 Frugone-Aacutelvarez et al 2017) In
our Lago Vichuqueacuten sequence BrTi values are high from 170 to 132 cm and from
36 cm to the top of core where it shows several sharps peaks (Fig 3) The MnFe
ratio is indicative of lake bottom oxygenation (Naecher et al 2013) as under
reducing conditions Mn tends to become more mobile than Fe leading to a
decreased MnFe (Marzecoacuteva et al 2011) Several sharp peaks in MnFe occurred
85
from 127 to 120 cm and from 57 to 30 cm (MgFegt03) The silicate group and the
Br Cl Ca Mn group are negatively correlated (r= -012 and -066)
Principal Component Analysis (PCA) was undertaken on the XRF
geochemical data to investigate the main factors controlling sediment deposition in
Lago Vichuqueacuten (Fig 3) The four main eigenvectors explain 801 of the variance
(supplementary material Table S2) The principal component (PC1) explain 437
of the total of variance and grouped elements are associated with terrigenous input
to the lake Positive values of the biplot have been attributed to higher heavy metals
deposition (Pb Co and Fe) and negative values to higher silicate input (Al K Ti and
Si) in a high energy catchment environment (Zr) The PC2 explains 198 of the
total of variance and highlights the endogenic productivity in the lake The positive
loading is compound by Br Cl Ca and Sr and to a lesser extent by Y Zn Rb and
Mn The PC2 may explain both the precipitation of carbonates (Ca Sr) and biological
production (Br)
86
Figure 3 Principal Component Analysis of XRF geochemical measurements in
VIC13-2B Lago Vichuqueacuten lake sediments
42 Sedimentary units
Based on geochemical and stable isotope analysis we identified four
lithological subunits in the short core sedimentary sequence Our PCA analyses and
Pearson correlations pointed out which variables were better for characterizing the
subunits (Fig 4) in terms of endogenic productivity heavy metals and terrestrial
input The unit 2c (170-131 cm) is characterized by the occurrence of some clastic
layers (L2 from 160 and 150 cm) high δsup1⁵N values (mean= +59 plusmn 04permil) with
Magnetic Susceptibility (MS) BrTi ZrTi and SrCa values increase towards the top
Also it has relatively low δsup1sup3C values (mean= -265 plusmn 11permil) and CNmolar ratios
(mean=78 plusmn 04) The ZrTi show upwards increasing values reaching the highest
values of the sequence at the top of this unit suggesting a coarsening upward trend
and relatively higher depositional energy The MS trend also indicates higher
erosion in the watershed and enhanced delivery of ferromagnetic minerals likely
from soil particles particularly from 150 to 160 cm (Frugone-Aacutelvarez at al 2017)
The subunit 2b (130-118 cm) is also composed of black silts but it has the
lowest MS values of the whole sequence and its onset is marked by a sharp
decrease in ZrTi This subunit is marked by sharp peaks of MnFe (at 126 and 120
cmgt027) SrCa (at 125 and 120 cmgt033) CaTi (at 124 and 120 cmgt199) TOC
(about 26 at 130 124 122 and 118 cm) and δsup1⁵N (gt+67permil at 126 and 120 cm)
BrTi oscillates around low values (about 01) whereas the δsup1sup3C values range
between -262 and -282permil
87
The unit 2a (58-117 cm) shows increasing and then decreasing MS values
and displayed sharps peaks of δsup1⁵N (gt64 at 102 to 99 87 and 75 cm) and CN
(about 10 at 86 74 66 and 60 cm) SrCa (mean = 04 plusmn 013) BrTi (mean = 008
plusmn 002) MnFe (mean = 002 plusmn 001) and δsup1sup3C (mean = -260 plusmn 07permil) oscillate in
low values The upper part of this unit (72 ndash 57 cm) shows an increase of SrCa
(from 03 to 05)
The upper unit 1a (0 - 57 cm) starts with a fine clastic layer (T at 57 to 54
cm) interpreted as deposition during a high-energy event It is characterized by
lower BrTi (mean = 003 plusmn 002) TOC (mean = 12 plusmn 03) and δsup1sup3C (mean= -
266 plusmn 06permil) higher δsup1⁵N (mean = +55 plusmn 08 permil) This unit is made of alternating
fine (lt 1cm) black and dark brown laminae with carbonate (L1 facies) Recently
deposited sediments show decreasing MS values increasing TOC (mean= 15 plusmn
04 ) sharp peaks of BrTi (gt015 at 33 21 5 and 1 cm) and the lowest δsup1⁵N values
of the entire record (mean= 45 plusmn 06 permil) and δsup1sup3C values (mean= -277 plusmn 09 permil)
Heavy metal content increases abruptly in the upper section of the core (2 - 20 cm)
(peaks of FeTi CoTi and PbTi)
88
Figure 4 Sedimentary units of Lago Vichuqueacuten sedimentary sequence Selected
variables illustrate watershed input (Magnetic Susceptibility ZrTi CoTi and MnFe)
endogenic productivity (SrCa CaTi TS) organic matter source (BrTi TOC
CNmolar and stable isotope records (δ13Corg and δ15Nbulk)
43 Recent seasonal changes of particulate organic matter on water column
The average of δ15NPOM from the three sites across to the lake was +86 plusmn 58
permil oscillating widely from -14 to +193permil (Fig 5) Important seasonal differences
occur at 2 and 5 m depth with more negative values during summer (~53 plusmn 52 permil)
than winter (~122 plusmn 40permil) At 20 m depth δ15NPOM values exhibit small seasonal
ranges (mean = +10permil for winter and +87permil for summer) The δ13CPOM average was
-296 plusmn 33permil with slightly seasonal and water column depth differences However
more positive δ13CPOM values occurred during winter (mean=-290 plusmn 41permil) than in
summer (mean= -301 plusmn 19 permil) Like the δ15NPOM values CNPOM (mean= 83 plusmn 17)
displayed important seasonal and water depth differences Lower CNPOM ratios
89
occurred during summer at 2 and 5 m depth (mean= 95 plusmn 17) and higher and more
constant values during winter (mean = 73 plusmn 09) at the same depth At 20 m CNPOM
shows similar values in both winter (70) and summer (74)
Figure 5 Seasonal POM averages from samples taken of the Lago Vichuqueacuten
water column Samples were taken from 2015 to 2018 at 2 (n=16) 5 (n=14) and 20
(n=8) meters depth
44 Stable isotope values across the Lake Vichuqueacuten watershed
Figure 6 shows modern vegetation soil and sediment isotope values found for
the watershed Huintildee tributary and Vichuqueacuten lake The δsup1⁵NFoliar values from
meadow plantations and macrophytes have similar range values with a mean of
+89 plusmn50 permil (n=18) +74 plusmn 75 permil (n=5) +82 plusmn 36 permil (n=6) respectively Native
vegetation has overall lower δsup1⁵NFoliar values with a mean of 14 plusmn 21 permil (n=28 see
Table 2 in supplementary material) The δsup1sup3CFoliar from terrestrial samples exhibit
similar values across the different plant communities (tree plantation mean=-274 plusmn
13permil meadow mean = -275 plusmn 23 permil native species mean=-272 plusmn 14permil) whereas
macrophytes display slightly more negative values with a mean of -287 plusmn 23permil
Macrophytes substrate displayed the highest δsup1⁵N values with a mean of +60 plusmn
14permil (n=3) Similar and relatively high δsup1⁵N values for soil of plantation (mean= +54
plusmn 13permil n=4) meadow (mean= +49 plusmn 04permil n=8) and surface river sediment
90
(mean= +52 plusmn 17permil n=10) Native forest soils oscillate around slightly more
negative values with a mean of +45 plusmn 12permil (n=4) Slightly more positive soil δsup1sup3C
values occur both underneath native forests and in tree plantations with means of -
284 plusmn 23permil and -298 plusmn 13permil respectively compared to either meadow soils
(mean= -302 plusmn 11permil) aquatic sediments underneath macrophytes (-311 plusmn 10permil)
or from surface river sediments (mean= -312 plusmn 10permil)
Figure 6 Stable isotope content of modern soil river sediment and foliar vegetation
used as end members in the sedimentary sequence of Lago Vichuqueacuten a)
Scatterplot of foliar δsup1⁵N and δsup13C values for vegetation from Lago Vichuqueacuten
watershed (plantation meadow and native species) and macrophytes on Lake
Vichuqueacuten See supplementary material for more detail of vegetation types b)
Scatterplot of soil δsup1⁵N and δsup13C values across the watershed the sediments of the
Huintildee and Vichuqueacuten rivers and the fluvial sediment corresponding to the
macrophyte vegetation
45 Land use and cover change from 1975 to 2014
Major land use changes between 1975 CE and 2016 CE in the Lago
Vichuqueacuten watershed are summarized in Figure 7 The watershed has a surface
area of 535329 km2 of which native vegetation (26) and shrublands (53)
represent 79 of the total surface in 1975 Meadows are confined to the valley and
91
represent 17 of watershed surface Tree plantations initially occupied 1 of the
watershed and were first located along the lake periphery By 1989 the areas of
native forests shrublands and meadows had decreased to 22 31 and 14
respectively whereas tree plantations had expanded to 30 These trends
continued almost invariably until 2016 when shrublands and meadows reached 17
and 5 of the total areas while tree plantations increased to 66 Native forests
had practically disappeared by 1989 and then increased up to 7 of the total area
in 2016 (Fig 6) preferentially occupying the southeastern sector of the watershed
Figure 7 Changes in land use and cover between 1975 and 2016 in the Lago
Vichuqueacuten watershed as measured from satellite images The major change is
represented by the replacement of native forest shrubland and meadows by
plantations of Monterrey pine (Pinus radiata)
Figure 8 shows correlations between lake sediment stable isotope values and
changes in the soil cover from 1975 to 2013 Positive relationships occurred
between sediment δsup1⁵N and δsup13C values from core VIC13-2B (unit 1) and the
92
percentage of native forest (r=089 for δsup1⁵N 082 for δsup13C) shrublands (r = 071 for
δsup1⁵N 057 for δsup13C) and meadow cover (r = 088 for δsup1⁵N 081 for δsup13C) All of these
correlations are significant (p value lt 0001) In contrast significant negative
correlations (p lt0001) occurred between tree plantation cover and lake sediment
stable isotope values (δsup1⁵N r = -082 and δsup13C r = -067) native forest (r = -097)
meadows (r = -086) and shrubland (r =-093)
Figure 8 Correlation plots of land use and cover change versus lake sediment
stable isotope values The δsup1⁵N values are positively correlated with native forests
agricultural fields and meadow cover across the watershed Total Plantation area
increases are negatively correlated with native forest meadow and shrubland total
area Significance levels are indicated by the symbols p-values (0 0001 001
005 01 1) lt=gt symbols ( )
93
5 Discussion
51 Seasonal variability of POM in the water column
The stable isotope values of POM can vary during the annual cycle due to
climate and biologic controls namely temperature and length of the photoperiod
which affect phytoplankton growth rates and isotope fractionation in the water
column (Gu et al 2006 Leng et al 2006) POM from Lago Vichuqueacuten surface
samples (2 and 5 m depth) displayed lower δ13CPOM values during the summer than
in winter During C uptake phytoplankton preferentially utilize 12C leaving the
DICpool enriched in 13C Therefore as temperature increases during the summer
phytoplankton growth generates OM enriched in 12C until this becomes depleted
and then the biomas come to enriched u At the onset of winter the DICpool is now
enriched in 13C and despite an overall decrease in phytoplankton production the
OM produced will also be enriched in 13C In contrast δ13CPOM values at 20 m depth
did not reflect these seasonal differences probably due to water-column
stratification that maintains similar temperatures and biological activity throughout
the year
Lake N availability depends on N sources including inputs from the
watershed and the atmosphere (ie deposition of N compounds and fixation of
atmospheric N2) which varies during the hydrologic year The fixation of atmospheric
N2 is an important natural source of N to the lake occurring mainly during the
summer season associated with higher temperature and light (Gu et al 2006)
Because the δ15NPOM of N2 fixers is close to 0permil with almost no overall isotope
fractionation (Leng et al 2006) when N2 is the major N source δ15NPOM values are
typically low However when DIN concentrations are high or alternatively when little
94
N2 fixation occurs δ15NPOM values will be higher (Gu et al 2006) δ15NPOM values
from summer Lago Vichuqueacuten samples were lower than those from winter with large
differences between the seasons (~5permil Fig 5 and 9) Furthermore δ15NPOM values
were high when monthly average temperature was low and monthly precipitation
was high (Fig 9) This pattern is consistent with the dominance of high N2 fixation
by cyanobacteria associated with increased summer temperatures This correlation
of δ15NPOM values with temperature further suggests a functional group shift i e
from N fixers to phytoplankton that uptake DIN The correlation between wetter
months and higher δ15NPOM values could be caused by increased N input from the
watershed due to increased runoff during the winter season The lack of data of the
δ15NPOM between the 30 mm and the 160 mm of precipitation is due to the
mediterranean-type climate that concentrates precipitations in the winter months
Finally Lago Vichuqueacuten is characterized by pronounced seasonality leading to
higher phytoplankton biomass in summer characterized by low δ15NPOM In winter
low biomass production and increased input from watershed is associated to high
δ15NPOM
95
Figure 9 Seasonal changes can drive δ15NPOM values in Lago Vichuqueacuten Data
correspond to average monthly temperature and total monthly precipitation for the
months when the water samples were taken (years 2015 - 2018) P-valuelt005
52 Stable isotope signatures in the Lake Vichuqueacuten watershed
The natural abundance of 15N14N isotopes of soil and vegetation samples
from the Lago Vichuqueacuten watershed appear to result from a combination of factors
isotope fractionation different N sources for plants and soil microorganisms (eg N2
fixation Nitrates) depth in the soil where N uptake by vegetation occurs and N loss
mechanisms (ie denitrification leaching and ammonia volatilization Hogberg
1997) The lowest δsup1⁵Nfoliar values are associated with native species and are
probably due to the contribution of N-fixing species (eg Sophora) (Fig6 and for
more detail see Table S3 in supplementary material) The number native N-fixers
species present in the Chilean mediterranean vegetation are not well known
however so there could be other N-fixing species in this group Alternatively δsup1⁵Nfoliar
values reflect soil N uptake (Kahmen et al 2008) In environments limited by N
plants have δsup1⁵N values similar to the soil δsup1⁵N values however the denitrification
and volatilization of ammonia can lead to the remain N of soil to come enriched in
15N (Cifuentes et al 1989 Craine et al 2009 Bruland and Mckenzie 2010) Soil N
isotope samples from native species communities tends to display relatively high
δsup1⁵N values respect to foliar samples due to loss of N-soil
The higher foliar and soil δsup1⁵N values obtained from samples of meadows
aquatic macrophytes and tree plantations can be attributed to the presence of
greater amounts of nitrate available for plant uptake (Fig 6) Houmlgberg (1997)
suggests that the availability of different N sources in soils (ie nitrates versus
96
ammonia) with different residence times can also explain these δsup1⁵NFoliar values
Indeed Feigin et al (1974) described differences of up to 20permil between ammonia
and nitrates sources Denitrification and nitrification discriminate much more against
15N than N2 fixation does (Craine et al 2009) leading to soils (and plants after
uptake) enriched in 14N
In general multiple processes that affect the isotopic signal result in similar
δsup1⁵N values between the soil of the watershed and the sediments of the river
However POM isotope fluctuations allow to say that more negative δsup1⁵N values are
associated to lake productivity while more positive δsup1⁵N values are associated with
N input from the watershed
δ13C values in Lago Vichuqueacuten watershed reflect CO2 gas exchange between
C3 plants and algae with the atmosphere During photosynthesis plants
discriminate against the heavier stable isotope of carbon (13C) in favor of the lighter
isotope 12C which results in δ13CFoliar values between -220permil and -320permil (Browman
and Cook 2002 Liu et al 2007) Likewise the δ13CFoliar values from Lago Vichuqueacuten
oscillate around -27permil (Fig 6) Plants transform atmospheric CO2 into soil organic
carbon (C) which in turn reflects this initial discrimination against 13C during C
uptake and post photosynthetic fractionation (Farquhar et al 1982 1989 Badeck
et al 2005 Pausch and Kuzyakov 2017) Slightly more positive δ13CFoliar values
(about 15permil) were measured in comparison with their δ13CSoil values This may be
reflecting the C transference from plants to the soil but also a soil-atmosphere
interchange The preferential assimilation of the light isotopes (12C) during soil
respiration carried by the roots and the microbial biomass that is associated with the
decomposition of litter roots and soil organic matter explain this differential
(Berhardt et al 2006 Blagodatskaya et al 2010 Midwood and Millard 2011)
97
In general the δ13CSoil values from the Lago Vichuqueacuten watershed oscillated
around -290permil and did not vary with our plant classification types Here we use
these values as terrestrial-end members to track changes in source OM (Fig 6)
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from the terrestrial watershed By the other hand more positive δ13C
values most likely reflect an increased aquatic OM component as indicated by POM
isotope fluctuations (Fig 9)
53 Recently land use and cover change and its influences on N inputs to the lake
Tree plantations in the Lago Vichuqueacuten watershed have increased greatly in
the last few decades (from 1 in 1975 to 66 in 2016 Fig 7) replacing previous
native forests (from 26 to 7) meadows (17 to 5) and shrublands (53 to
17) In 1975 tree plantations were confined to the lake perimeter with discrete
patches distributed throughout the watershed The ldquoForest Lawrdquo (DFL 701- passed
in 1974) allocated state funding to afforestation efforts and management of tree
plantations which greatly favored the replacement native forests by introduced trees
This increase is marked by a sharp and steady decrease in lake sediment δ15N and
δ13C values because tree plantations function as a nutrient sink whereas other land
uses such as meadows favor nutrient loss from the watershed (Fig8) Bruland and
Mackenzie (2014) noted a decrease in wetland δ15N values when watershed
forested cover increased and concluded that N inputs to the wetlands are lower from
the forested areas as they generally do not export as much N as agricultural lands
A positive correlation between native vegetation and δ15Ncore values can be
explained by the relatively scarce arboreal cover in the watershed in 1975 when
native forest occupied just 26 of the watershed surface whereas shrublands and
98
meadows occupied more than the 70 of the surface of the watershed with the
concomitant elevated loss of N (Fig 7 and Fig 8)
54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years
Sedimentological compositional and geochemical indicators all show changes in
the N cycle associated to LUCC lake dynamics and climate changes (Fig 10) From
the pre-Columbian indigenous settlement including the Spanish colonial period up
to the start of the Republic (1300 - 1800 CE) the introduction of crops such as
quinoa and wheat but also the clearing of land for extensive agriculture would have
favored the entry of N into the lake Conversely major changes observed during the
last century were characterized by a sharp decrease of N input that were coeval
with the increase of tree plantations
From 1365 until 1550 CE (unit 2c 2b and half of 2a Fig 3 and Fig 10) pre-
Columbian agricultural societies planted mostly crops of corn and quinoa (Ramirez
and Vidal 1985) This period is characterized by large peaks seen in the δsup1⁵N record
(eg 1403 1452 1476 and 1505) with overall high δsup1⁵N values (up to 5permil) indicating
that N input from watershed was elevated and oscillating to the beat of the NT These
positive δsup1⁵N peaks could be due to several causes including a) the clearing of land
for farming b) N loss via denitrification which would be generally augmented in
anoxic environments (Diebel et al 2009) such as Lago Vichuqueacuten (low MnFe
values about 0001 Fig 4) or c) climate especially cold-wet winters or hot-dry
summers can also exert control on the δsup1⁵N record Indeed tree-ring records and
summer temperature reconstructions show overall wetcold conditions during this
period (Christie et al 2009 Von Gunten et al 2009 Fig 10) Increased
precipitation would bring more sediment (and nutrients) from the watershed into the
99
lake and increase lake productivity which is also detected by the geochemical
proxies for productivity (peaks of BrTi SrCa and TOC Fig 4 and Fig 10) (see also
Frugone-Alvarez et al 2017)
Figure 10 Changes in the N availability during the last six centuries in Lago
Vichuqueacuten a) lake sediment δsup1⁵N values displayed more positive values during the
prehistoric period Spanish Colony and the starting 19th century which is associated
with enhanced N input from the watershed by extensive clearing and crop
plantations The inset shows this relationship between sediment δsup1⁵N and
100
percentage of meadow cover over the last 30 years b) Summer temperature
reconstruction from central Chile (von Gunten et al 2009) showing a
correspondence with cold phases and high δsup1⁵N values at Vichuqueacuten except for the
last 100 years c) Palmer Drought Severity Index (PDSI) is an interannual moisture
variability reconstruction for late springndashearly summer during the last six centuries
(Christie et al 2009) Grey shadow indicating higher precipitation periods
From 1550 and 1810 CE (unit 2a) and during the Colonial period several peaks
of δ15N could be further related not only to high precipitation (Fig 10 grey shadow)
but also pulses of enhanced N input from the watershed linked to human land use
In 1550 CE Juan Cuevas was granted lands and indigenous workers under the
encomienda system for agricultural and mining development of the Vichuqueacuten
village (Ramiacuterez and Vidal 1985) Historical documents show that around 1579 CE
the Vichuqueacuten watershed was occupied by indigenous communities dedicated to
wheat plantations and vineyards wood extraction and gold mining (Odone 1998)
The introduction of the Spanish agricultural system implied not just a change in the
types of crops used (from quinoa to vineyards and wheat) but also a clearing of
native species for the continuous increase of agricultural surface and wood
extraction During the Colonial period wheat from Vichuqueacuten was exported to Peru
(Ramiacuterez and Vidal 1985) Armesto et al (2009) indicate that during the XVIII and
XIX centuries the extraction of wood for mining operations was important enough to
cause extensive loss of native forests The independence and instauration of the
Chilean Republic did not change this prevailing system Increases in the
contributions of N to the lake during the second half of the XIX century (peaks in
δsup1⁵N ca 1870 CE) could be due to expanding farm land dedicated for wheat
101
production and increased commercial trade with California and Canada (Ramiacuterez
and Vidal 1985)
In contrast LUCC in the last century are clearly related to the development of
large-scale tree plantations (Fig 7) The δsup1⁵N record reaches the lowest values of
the entire sequence in the last few decades (Fig 10) A marked increase in lake
productivity NT concentration and decreasing sediment input is synchronous (unit
1 Fig 4) with trees replacing meadows shrublands and areas with native forests
(Fig 7 and 8) The implementation of the lsquoForest Lawrsquo in 1974 had a stronger impact
on the landscape and lake ecosystem dynamics than the impacts of ongoing climate
change in the region which is much more recent (Garreaud et al 2018) although
the prevalence of hot dry summers seen over the last decade would also be
associated with low δsup1⁵N values and increase of NT (Fig 10) Heavy metals ratios
(FeTi CoTi and PbTi) show notable increases in lake sediments from 1992 to 2011
CE (Fig 4) Although this could be related to mining in the El Maule region the
closest mines are 60 Km away (Pencahue and Romeral) so local factors related to
shoreline urbanization for the summer homes and an increase in tourist activity
could also be a major factor
6 Conclusions
The N isotope signal in the watershed depends on the rates of exchange
between vegetation and soil cover (Fig 11) Plants preferentially uptake 14N and the
underlying soils become enriched in 15N especially when the terrestrial ecosystem
is N-limited andor significant N loss occurs (ie denitrification andor ammonia
volatilization) LUCC in the Lago Vichuqueacuten watershed have heavily modified the
links between terrestrial and aquatic ecosystems with agriculture practices
102
contributing more N to the lake than tree plantations or native forests In situ lake
processes can also fractionate N isotopes An increase of N-fixing species results
in OM depleted in 15N which results in POM with lower δsup1⁵N values during these
periods During winter phytoplankton is typically enriched in 15N due to the
decreased abundance of N-fixing species and increased N input from the
watershed This in turn generates the highest δsup1⁵NPOM values in Lago Vichuqueacuten
Periods of elevated productivity tend to deplete the dissolved inorganic N of 14N
resulting in even higher δ15N values
Our study illustrates the usefulness of δsup1⁵N as a tool for reconstructing the past
influence of LUCC on N availability in lake ecosystems To constrain the relative
roles of the diverse forcing mechanisms that can alter N cycling in mediterranean
ecosystems all main components of the N cycle should be monitored seasonally
(or monthly) including the measurements of δ15N values in land samples
(vegetation-soil) as well as POM
103
Figure 11 Summary of human and environmental factors controlling the δ15N
values of lake sediments Particulate organic matter(POM) δ15N values in
mediterranean lakes are driven by N input from the watershed that in turn depend
on land use and cover changes (ie forest plantation agriculture) andor seasonal
changes in N sources andor lake ecosystem processes (ie bioproductivity redox
condition denitrification) Arrows indicate the direction of N fluxes (inputoutput from
the N cycle) N cycle processes that deplete lake sediments of 15N are shown in
blue whereas those that enrich sediments in 15N are shown in red
104
Supplementary material
Figure S1 Pearson correlate coefficient between geochemical variables in core
VIC13-2B Positive and large correlations are in blue whereas negative and small
correlations are in red (p valuelt0001)
Figure S2 Principal Component Analysis of geochemical elements from core
VIC13-2B
105
Table S1 Lago Vichuqueacuten radiocarbon samples
RADIOCARBON
LAB CODE
SAMPLE
CODE
DEPTH
(m)
MATERIAL
DATED
14C AGE ERROR
D-AMS 029287
VIC13-2B-
1 043 Bulk 1520 24
D-AMS 029285
VIC13-2B-
2 085 Bulk 1700 22
D-AMS 029286
VIC13-2B-
2 124 Bulk 1100 29
Poz-63883 Chill-2D-1 191 Bulk 945 30
D-AMS 001133
VIC11-2A-
2 201 Bulk 1150 44
Poz-63884
Chill-2D-
1U 299 Bulk 1935 30
Poz-64089
VIC13-2D-
2U 463 Bulk 1845 30
Poz-64090
VIC13-2A-
3U 469 Bulk 1830 35
D-AMS 010068
VIC13-2D-
4U 667 Bulk 2831 25
Poz-63886
VIC13-2D-
4U 719 Bulk 3375 35
106
D-AMS 010069
VIC13-2D-
5U 775 Bulk 3143 27
Poz-64088
VIC13-2D-
5U 807 Bulk 3835 35
D-AMS-010066
VIC13-2D-
7U 1075 Bulk 6174 31
Poz-63885
VIC13-2D-
7U 1197 Bulk 6440 40
Poz-5782 VIC13-15 DIC 180 25
Table S2 Loadings of the trace chemical elements used in the PCA
Elementos PC1 PC2 PC3 PC4
Zr 0922 0025 -0108 -0007
Zn 0913 -0124 -0212 0001
Rb 0898 -0057 -0228 0016
K 0843 0459 0108 0113
Ti 0827 0497 0060 -0029
Al 0806 0467 0080 0107
Si 0803 0474 0133 0136
Y 0784 -0293 -0174 0262
V 0766 0455 0090 -0057
Br 0422 -0716 -0045 0226
Ca 0316 -0429 0577 0489
Sr 0164 -0420 0342 -0182
Cl 0151 -0781 -0397 0162
107
Mn -0121 -0091 0859 0095
S -0174 -0179 -0051 0714
Pb -0349 0414 -0282 0500
Fe -0700 0584 -0023 0280
Co -0704 0564 -0107 0250
Table S3 Stable Isotope values from vegetation of Lago Vichuqueacutenacutes watershed
Taxa Classification δsup1⁵N δsup1sup3C CN
molar
Poaceae Meadow 1216 -2589 3602
Juncacea Meadow 1404 -2450 3855
Cyperaceae Meadow 1031 -2596 1711
Taraxacum
officinale Meadow 836 -2400 2035
Poaceae Meadow 660 -2779 1583
Poaceae Meadow 453 -2813 1401
Poaceae Meadow 966 -2908 4010
Juncus Meadow 1247 -2418 3892
Poaceae Meadow 747 -3177 6992
Poaceae Meadow 942 -2764 3147
Poaceae Meadow 1479 -2634 2895
Poaceae Meadow 1113 -2776 1795
Poaceae Meadow 2215 -2737 7971
Poaceae Meadow 1121 -2944 2934
Poaceae Meadow 638 -3206 1529
108
Macrophytes Macrophytes 886 -3044 2286
Macrophytes Macrophytes 1056 -2720 2673
Macrophytes Macrophytes 769 -3297 1249
Macrophytes Macrophytes 967 -2763 1442
Macrophytes Macrophytes 959 -2670 2105
Macrophytes Macrophytes 334 -2728 1038
Acacia dealbata
Introduced
species 656 -2696 1296
Acacia dealbata
Introduced
species 487 -2941 1782
Acacia dealbata
Introduced
species 220 -2611 3888
Luma apiculata Native species 433 -2542 4135
Luma apiculata Native species 171 -2664 7634
Luma apiculata Native species -001 -2736 6283
Luma apiculata Native species 029 -2764 6425
Azara sp Native species 159 -2868 8408
Azara sp Native species 101 -2606 2885
Baccharis concava Native species 104 -2699 5779
Baccharis concava Native species 265 -2488 4325
Baccharis concava Native species 287 -2562 7802
Baccharis concava Native species 427 -2781 5204
Baccharis linearis Native species 190 -2610 4414
Baccharis linearis Native species 023 -2825 5647
109
Peumus boldus Native species 042 -2969 6327
Peumus boldus Native species 205 -2746 4110
Peumus boldus Native species 183 -2743 6293
Chusquea quila Native species 482 -2801 4275
Poaceae meadow 217 -2629 7214
Lobelia sp Native species 224 -2645 3963
Lobelia sp Native species -091 -2565 4538
Aristotelia chilensis Native species -035 -2785 5247
Aristotelia chilensis Native species -305 -2889 2305
Aristotelia chilensis Native species 093 -2836 5457
Chusquea quila Native species 173 -2754 3534
Chusquea quila Native species 045 -2950 6739
Quillaja saponaria Native species 223 -2838 9385
Scirpus meadow 018 -2820 7115
Sophora sp Native species -184 -2481 2094
Sophora sp Native species -181 -2717 1721
Pinus radiata
Introduced
trees 1581 -2602 3679
Pinus radiata
Introduced
trees 1431 -2784 4852
Pinus radiata
Introduced
trees -091 -2708 9760
Pinus radiata
Introduced
trees 153 -2568 3470
110
Salix sp
Introduced
trees 632 -2878 1921
LITERATURE CITED
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A historical
framework for land cover change in southwestern South America in the past 15000
years Land use policy 27 148ndash160
httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the next
carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014
httpsdoiorg101002eft2235
Blaauw M Christeny JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474
httpsdoiorg10121411-BA618
Bowman DMJS Cook GD 2002 Can stable carbon isotopes (δ13C) in soil
carbon be used to describe the dynamics of Eucalyptus savanna-rainforest
boundaries in the Australian monsoon tropics Austral Ecol
httpsdoiorg101046j1442-9993200201158x
Brahney J Ballantyne AP Turner BL Spaulding SA Otu M Neff JC 2014
Separating the influences of diagenesis productivity and anthropogenic nitrogen
deposition on sedimentary δ15N variations Org Geochem 75 140ndash150
httpsdoiorg101016jorggeochem201407003
111
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409
httpsdoiorg102134jeq20090005
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego R
Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and
environmental change from a high Andean lake Laguna del Maule central Chile
(36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the
Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from
tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Craine JM Elmore AJ Aidar MPM Dawson TE Hobbie EA Kahmen A
Mack MC Kendra K Michelsen A Nardoto GB Pardo LH Pentildeuelas J
Reich B Schuur EAG Stock WD Templer PH Virginia RA Welker JM
Wright IJ 2009 Global patterns of foliar nitrogen isotopes and their relationships
with cliamte mycorrhizal fungi foliar nutrient concentrations and nitrogen
availability New Phytol 183 980ndash992 httpsdoiorg101111j1469-
8137200902917x
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty
Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-
010-9453-1
Diebel M Vander Zanden MJ 2012 Nitrogen stable isotopes in streams stable
isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19
1127ndash1134 httpsdoiorg10189008-03271
112
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in freshwater
wetlands record long-term changes in watershed nitrogen source and land use SO
- Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash
2916
Fariacuteas L Castro-Gonzales M Cornejo M Charpentier J Faundez J
Boontanon N Yoshida N 2009 Denitrification and nitrous oxide cycling within the
upper oxycline of the oxygen minimum zone off the eastern tropical South Pacific
Limnol Oceanogr 54 132ndash144
Farquhar GD Ehleringer JR Hubick KT 1989 Carbon Isotope Discrimination
and Photosynthesis Annu Rev Plant Physiol Plant Mol Biol
httpsdoiorg101146annurevpp40060189002443
Farquhar GD OrsquoLeary MH Berry JA 1982 On the relationship between
carbon isotope discrimination and the intercellular carbon dioxide concentration in
leaves Aust J Plant Physiol httpsdoiorg101071PP9820121
Fogel ML Cifuentes LA 1993 Isotope fractionation during primary production
Org Geochem httpsdoiorg101007978-1-4615-2890-6_3
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A
Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-
resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS)
implications for past sea level and environmental variability J Quat Sci 32 830ndash
844 httpsdoiorg101002jqs2936
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924
httpsdoiorg104319lo20095430917
113
Gerhart LM McLauchlan KK 2014 Reconstructing terrestrial nutrient cycling
using stable nitrogen isotopes in wood Biogeochemistry 120 1ndash21
httpsdoiorg101007s10533-014-9988-8
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and oxygen
isotope fractionation during dissimilatory nitrate reduction by denitrifying bacteria 53
2533ndash2545 httpsdoiorg10230740058342
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater
eutrophic lake Limnol Oceanogr 51 2837ndash2848
httpsdoiorg104319lo20065162837
Harris D Horwaacuteth WR van Kessel C 2001 Acid fumigation of soils to remove
carbonates prior to total organic carbon or CARBON-13 isotopic analysis Soil Sci
Soc Am J 65 1853 httpsdoiorg102136sssaj20011853
Houmlgberg P 1997 Tansley review no 95 natural abundance in soil-plant systems
New Phytol httpsdoiorg101046j1469-8137199700808x
Kylander ME Ampel L Wohlfarth B Veres D 2011 High-resolution X-ray
fluorescence core scanning analysis of Les Echets (France) sedimentary sequence
New insights from chemical proxies J Quat Sci 26 109ndash117
httpsdoiorg101002jqs1438
Lara A Solari ME Prieto MDR Pentildea MP 2012 Reconstruccioacuten de la
cobertura de la vegetacioacuten y uso del suelo hacia 1550 y sus cambios a 2007 en la
ecorregioacuten de los bosques valdivianos lluviosos de Chile (35o - 43o 30acute S) Bosque
(Valdivia) 33 03ndash04 httpsdoiorg104067s0717-92002012000100002
Lehmann M Bernasconi S Barbieri A McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during
114
simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66
3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007
Liu G Li M An L 2007 The Environmental Significance Of Stable Carbon
Isotope Composition Of Modern Plant Leaves In The Northern Tibetan Plateau
China Arctic Antarct Alp Res httpsdoiorg1016571523-0430(07-505)[liu-
g]20co2
Marzecovaacute A Mikomaumlgi A Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54
httpsdoiorg103176eco2011105
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash
1643 httpsdoiorg1011770959683613496289
Meyers PA 2003 Application of organic geochemistry to paleolimnological
reconstruction a summary of examples from the Laurention Great Lakes Org
Geochem 34 261ndash289 httpsdoiorg101016S0146-6380(02)00168-7
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland
Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Odone C 1998 El Pueblo de Indios de Vichuqueacuten siglos XVI y XVII Rev Hist
Indiacutegena 3 19ndash67
Pausch J Kuzyakov Y 2018 Carbon input by roots into the soil Quantification of
rhizodeposition from root to ecosystem scale Glob Chang Biol
httpsdoiorg101111gcb13850
115
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98
httpsdoiorg1011772053019614564785
Sun W Zhang E Jones RT Liu E Shen J 2016 Biogeochemical processes
and response to climate change recorded in the isotopes of lacustrine organic
matter southeastern Qinghai-Tibetan Plateau China Palaeogeogr Palaeoclimatol
Palaeoecol httpsdoiorg101016jpalaeo201604013
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of
different trophic status J Paleolimnol 47 693ndash706
httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl
httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M 2009 High-resolution quantitative climate
reconstruction over the past 1000 years and pollution history derived from lake
sediments in Central Chile Philos Fak PhD 246
Woodward C Chang J Zawadzki A Shulmeister J Haworth R Collecutt S
Jacobsen G 2011 Evidence against early nineteenth century major European
induced environmental impacts by illegal settlers in the New England Tablelands
south eastern Australia Quat Sci Rev 30 3743ndash3747
httpsdoiorg101016jquascirev201110014
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K Yeager
KM 2016 Different responses of sedimentary δ15N to climatic changes and
116
anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau
J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
117
DISCUSION GENERAL
El Nitroacutegeno (N) es un elemento clave que controla la dinaacutemica y
funcionamiento de ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al
1997) Pese a que el N (N2) es muy abundante en la atmoacutesfera (78) es escaso
en los ecosistemas pues la mayoriacutea de la biota no puede acceder a su forma
molecular (Galloway et al 1995 Zaehle et al 2013) En este sentido la entrada
natural de N en los ecosistemas es viacutea fijacioacuten bioloacutegica del N2 por bacterias que lo
convierten en compuestos bioloacutegicamente disponibles (Battye et al 2017) Debido
a que la fijacioacuten es un proceso energeacuteticamente costoso los ecosistemas
comuacutenmente estaacuten limitados por N (Vitousek and Howarth 1991) Los LUCC
contribuyen al incremento del N disponible y son una de las principales causas de
eutroficacioacuten y contaminacioacuten de los cuerpos de agua (Woodward et al 2012)
En Chile central los LUCC principalmente relacionados con las actividades
agriacutecolas y forestales han causado grandes cambios en el ciclo del N debido al
118
reemplazo de la cobertura natural por un monocultivo y al uso de fertilizantes que
modifican los aportes de MO y N a los cuerpos de agua El programa de
estabilizacioacuten de laderas impulsada por el gobierno (Albert 1900) y la ley forestal
de 1974 han fomentado la forestacioacuten con plantaciones de Pinus radiata y
Eucalyptus globulus y en el caso de Lago Vichuqueacuten estas actualmente cubren maacutes
del 60 de su cuenca (Cap2 Fig 5) Ademaacutes en Laguna Matanzas la
sobreexplotacioacuten del agua en conjunto con el cambio climaacutetico actual el que ha
conducido a una de las megasequiacuteas maacutes severas en las uacuteltimas deacutecadas
(Garreaud et al 2017) generoacute una combinacioacuten letal que secoacute el lago en tan solo
10 antildeos (Cap1 Fig1) Las evidencias obtenidas a partir de estos dos lagos
permiten identificar las huellas del Antropoceno en Chile central basadas en el
registro sedimentario lacustre
La transferencia de N desde los ecosistemas terrestres a acuaacuteticos es un
proceso natural con controles bioacuteticos climaacuteticos y geoloacutegicos El enlace
hidroloacutegico entre los ecosistemas terrestres y acuaacuteticos hace que el estado troacutefico
de los lagos esteacute vinculado al de su cuenca (McLauchlan et al 2013) En Chile
central se sabe poco del impacto de los LUCC sobre el ciclo de nutrientes en los
ecosistemas lacustres aunque al menos desde la colonizacioacuten espantildeola se tienen
registros de influencia humana en las cuencas Durante la colonia espantildeola
Laguna Matanzas fue una hacienda ganadera que exportaba productos caacuternicos al
Peruacute (Contreras-Lopez et al 2014) A su vez en el Lago Vichuqueacuten se realizaban
extensas actividades agriacutecolas hasta comienzos del s XX como por ejemplo
cultivos de vid y trigo ademaacutes de extraccioacuten de madera y mineriacutea de oro (Odone
1997) En las uacuteltimas deacutecadas los LUCC han estado vinculados principalmente con
el desarrollo forestal al compaacutes de una estrategia gubernamental de fomentar con
119
incentivos financieros (Ley de Bosques DFL nordm265 de 1931 y DFL 701 de 1974)
esta actividad El incremento de la superficie forestal es especialmente fuerte en
ambos lagos a partir de la deacutecada de 1980 y hasta 2016 (Laguna de Matanzas 0-
17 Lago Vichuqueacuten 1-66) Este aumento habriacutea sido en detrimento del bosque
nativo y los pastizales (Cap 1 Fig 5 y Cap 2 Fig 8) El incremento de la superficie
forestal generariacutea una disminucioacuten de los aportes de sedimentos y nutrientes al lago
y en este sentido un cambio de estado en los flujos de N (e g tipping points) que
a su vez ha quedado registrado en la sentildeal isotoacutepica (δ15N) y la acumulacioacuten de
MO en los sedimentos lacustres
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central
Los sedimentos lacustres son verdaderos ldquosumiderosrdquo que pueden llegar a
registrar lo que ocurre en cuanto a la historia ambiental de su cuenca En esta tesis
se utilizan los isoacutetopos estables de N (15N y 14N) en sedimentos lacustres para
reconstruir los cambios en la disponibilidad de N en el tiempo y establecer la
magnitud de impacto generado por actividades humanas El fraccionamiento
cineacutetico en los lagos es mediado por procesos bioloacutegicos que mediante la
asimilacioacuten de N genera un producto (eg fitoplancton) maacutes ldquoligerordquo (valores maacutes
bajos de 15N) que su remanente (eg DIN) (Fogel y Cifuentes 1993) Sin embargo
en periacuteodos en que los lagos estaacuten limitados por N el fitoplancton discrimina menos
y la MO resultante queda enriquecida en 15N (Fig 1 a y b) Ademaacutes la
desnitrificacioacuten acentuada en lagos de fondo anoacutexicos (eg Laguna Matanzas
entre 1800 y 1940 Cap1 Fig 6) conduce a un enriquecimiento en 15N de los
sedimentos y de la columna de agua Esta informacioacuten queda reflejada en la MO
120
de los sedimentos lacustres lo que permite conocer la disponibilidad del N en el
tiempo a partir de las variaciones de 15N
En esta tesis analizamos la MO de los sedimentos lacustres para reconstruir
la influencia de los LUCC en los aportes de sedimentos y N al lago En teacuterminos de
asimilacioacuten de N se puede distinguir entre dos grupos principales de productores
primarios que componen el POM (Fig1)
1 Los que asimilan el DIN compuesto por amonio nitratos y nitritos Aquiacute el
δ15N resultante fluctuaraacute en torno a valores maacutes positivos en la medida que
la productividad se incrementa o no hay reposicioacuten del DIN (Fig 1)
2 Los fijadores de N Dado que es un proceso energeacuteticamente costoso en
ambientes que no estaacuten limitados por N muchas veces son excluiacutedas
competitivamente por el resto del fitoplancton Si el DIN queda agotado por
el incremento de la productividad (o bien si el ambiente estaacute limitado ya sea
por N o por otros nutrientes) los fijadores de N pueden florecer y la MO que
se genera a partir de ello tendraacute un δ15N con valores en torno a 0permil
De este modo la MO en los sedimentos lacustres dependeraacute de la
composicioacuten de productores primarios (ie fijadores vs asimiladores del DIN)
ademaacutes de los aportes desde la cuenca tanto de MO como del N de la cuenca que
pasa a formar parte del DIN (eg fertilizantes estieacutercol) (Fig 1)
La MO de los lagos estudiados en esta tesis ha sido analizada a partir de
variables geoquiacutemicas isotoacutepicas frotis y estaacute compuesta principalmente por
diatomeas y MO amorfa A partir de los datos obtenidos en esta tesis los valores
de δ15N aumentan cuando hay una mayor cobertura agriacutecola o pastizales A su vez
tiende a disminuir cuando aumenta la cobertura arboacuterea independiente si esta es
por plantaciones forestales o por bosque nativo
121
Por otra parte la dinaacutemica del lago tambieacuten imprime caracteriacutesticas
especiales en el POM observaacutendose variaciones estacionales en los valores
δ15NPOM Durante la estacioacuten caacutelida y seca se observan valores maacutes bajos que
durante la estacioacuten friacutea y huacutemeda posiblemente relacionado con el incremento de
la contribucioacuten de especies fijadoras de N (Lago Vichuqueacuten Fig 9 Cap 2) durante
el verano En la estacioacuten friacutea y huacutemeda el DIN seriacutea la principal fuente de N Las
mayores entradas de MO y N terrestre debidos a un incremento del lavado de la
cuenca como consecuencia de las precipitaciones y la mineralizacioacuten de la MO
podriacutean estimular la produccioacuten de MO lacustre por crecimiento del fitoplancton
Como consecuencia se observan tendencias decrecientes de los valores de
δ15N en la MO sedimentaria cuando la productividad en el lago estaacute relacionada
con especies fijadoras de N mientras que el δ15N muestra aumentos cuando la
productividad del lago estaacute asociada principalmente al consumo del DIN pero
tambieacuten cuando predominan procesos de perdida de N (eg desnitrificacioacuten Fig
1)
Hemos identificado tres fases en la dinaacutemica del δ15N para ambos lagos
Entre 1800 y 1940 CE la cuenca de laguna de Matanzas estaba dominada por
actividades agriacutecolas y ganaderas (δ15Nsuelo gt 4permil Cap 1 Fig 4 y 6) Las entradas
de sedimentos y MO desde la cuenca son altas (δ13C lt -209permil ZrTi ~029 AlTi
~013) y coincide con un clima maacutes huacutemedo y friacuteo (Von Gunten et al 2009
Christiansen et al 2011) El lago teniacutea un nivel de agua maacutes profundo dominado
por condiciones anoacutexicas (MnFe ~0013) que contribuiriacutean a mayores valores de
δ15N en la columna de agua y por ende de los sedimentos acumulados en el fondo
debido a la peacuterdida diferencial de 14N viacutea desnitrificacioacuten (Fig 7 Cap1) La baja
122
produccioacuten de MO lacustre (BrTi ~002 TOC~17) coincide con un bajo contenido
de NT (~478 μg) y valores maacutes positivos de δ15N (gt 4permil)
La actividad agriacutecola tambieacuten dominaba en la cuenca del Lago Vichuqueacuten
durante este periacuteodo (δ15Nsuelo gt4permil Cap2 Fig 6) El aporte sedimentario desde la
cuenca es alto (AlTi ~029 ZrTi ~032) y fue favorecido por mayores
precipitaciones a las actuales (Christiansen et al 2011) El Lago Vichuqueacuten es un
lago estratificado anoacutexico (MnFe ~002) y profundo (~30 m) por lo que la
desnitrificacioacuten juega un rol importante en el enriquecimiento de la MO
sedimentaria La baja produccioacuten de MO (BrTi ~007 TOC ~12) de origen
lacustre (δ13C ~ -268permil) coincide con un bajo contenido de NT (~309μg +6) y
valores maacutes positivos de δ15N (56permil +03)
Durante esta fase en ambos lagos los aportes de N de la cuenca parecen
ser los principales responsables en las fluctuaciones del δ15N Esta sentildeal podriacutea
estar amplificada por bajas temperaturas (que afectan la productividad lacustre) y
altas precipitaciones que favorecen el lavado de la cuenca y aumentan el aporte de
sedimentos y MO desde la cuenca predominantemente agriacutecola
Entre 1940 y 1980 CE en Laguna Matanzas se observa un incremento en
la acumulacioacuten de MO algal (TOC ~2 +1 δ13C ~-230+09permil) El ambiente
deposicional es ligeramente menos turbulento que el anterior (AlTi ~011+001
ZrTi ~03+001) y el lago estaacute en una fase de transicioacuten hacia condiciones maacutes
oacutexicas (MnFe ~002+0004) y maacutes productivas (BrTi ~004+0007) A su vez δ15N
tiende hacia valores maacutes negativos (δ15N ~30permil+03) mientras que el NT aumenta
oscilando en antifase con el δ15N
En Lago Vichuqueacuten en cambio se observa un ligero incremento en la
acumulacioacuten de MO (TOC ~13 +02) de origen terrestre (δ13C ~-27permil +04) La
123
productividad es similar al periacuteodo anterior (BrTi ~007+003) y el ambiente
deposicional es menos turbulento (AlTi ~02 +009 Zrti ~033 +007) El δ15N y el
NT oscilan en torno a valores ligeramente maacutes bajos (δ15N ~51permil +04 NT ~292μg
+6) Este periacuteodo se caracteriza por ser maacutes caacutelido que el anterior lo que
posibilitariacutea un incremento en la productividad lacustre en Laguna Matanzas pero
que no es observada en el Lago Vichuqueacuten
Entre 1980 y la actualidad en Laguna Matanzas habriacutea un incremento de la
acumulacioacuten de MO (TOC ~59 + 3) asociada a un incremento en la productividad
del lago (BrTi ~009+005) y a los aportes de MO terrestres (δ13C ~-24 + 2permil) El
lago se vuelve maacutes oacutexico (Mn ~0003 +0006) y las entradas de sedimento
disminuyen (AlTi~ 009 +002) El δ15N tiende a valores maacutes negativos (δ15N ~13permil
+1) y el NT aumenta de 20 a 55 μg (NT ~346 + 9 μg) tomando valores sin
precedentes en todo el registro En los sedimentos lacustres del Lago Vichuqueacuten
tambieacuten se observa un incremento de la acumulacioacuten de MO (TOC ~17 + 05)
asociada a un incremento en la productividad del lago (BrTi ~01 + 003) y las
entradas de sedimento desde la cuenca disminuyen (AlTi ~005 + 005) El δ15N
(~43 + 05permil) tiende a valores maacutes negativos y el NT incrementa de 20 a 55 μg (NT
~346 + 9 μg) oscilando en antifase
Durante esta fase en ambos lagos se observa un aumento en la
acumulacioacuten de MO sedimentaria un incremento en el NT y valores maacutes negativos
de δ15N que coincide con el incremento de la superficie forestal de las cuencas
(Laguna Matanzas gt17 Lago Vichuqueacuten gt60)
124
Figura 2 Comparacioacuten de los cambios del ciclo del N entre Laguna Matanzas y
Lago Vichuqueacuten con los procesos biogeoquiacutemicos contrastantes en rojo En L
Matanzas las condiciones oacutexicas podriacutean estar favoreciendo la nitrificacioacuten del
amonio En Vichuqueacuten las condiciones anoacutexicas favorecen un aumento en δ15N de
la MO en los sedimentos por desnitrificacioacuten y mineralizacioacuten
Los ambientes mediterraacuteneos en el que los lagos del presente estudio se
encuentran insertos estaacuten caracterizados por una estacioacuten seca prolongada y las
precipitaciones ocurren en eventos puntuales alcanzando altos montos
pluviomeacutetricos en poco tiempo Las lluvias favorecen un lavado de la cuenca la
perdida de N y otros nutrientes Por lo que es esperable que los sedimentos del
lago reflejen mejor los valores isotoacutepicos de los suelos de la cuenca durante los
periacuteodos con altos montos pluviomeacutetricos En el Lago Vichuqueacuten por ejemplo el
125
POM superficial (2 y 5 m de profundidad) despliega valores isotoacutepicos maacutes
positivos en invierno presumiblemente como resultado de mayores aportes de MO
y N de su cuenca (Fig 1b) En ambos lagos los valores maacutes positivos en los
sedimentos lacustres ocurren durante periacuteodos con mayores montos pluviomeacutetricos
(Cap1 Fig 6 y Cap 2 Fig 12)
Ademaacutes del efecto de la precipitacioacuten en el aporte de nutrientes al lago en
esta tesis hemos encontrado que los mayores aportes de N desde la cuenca se dan
cuando la cobertura vegetal del suelo es menor En Laguna Matanzas el desarrollo
de la actividad ganadera desde la llegada de los espantildeoles a la cuenca habriacutea
incrementado los aportes de N al lago Los valores de δ15N en los sedimentos
lacustres depositados durante este periacuteodo son los maacutes positivos de todo el registro
(~4permil) A su vez en Lago Vichuqueacuten los valores isotoacutepicos maacutes positivos se
registran entre 1300 y 1900 CE que coinciden con el mayor desarrollo de
actividades agriacutecola y de extraccioacuten de maderas de la cuenca (Fig 10 Cap 2)
Los recientes LUCC (a partir de 1975) en las cuencas de Laguna Matanzas
y Lago Vichuqueacuten estaacuten caracterizados por un incremento de la superficie forestal
y agriacutecola en reemplazo de la cobertura de pastizal (Laguna Matanzas) y bosque
nativo (Lago Vichuqueacuten) Los valores de δ15N en los testigos sedimentarios fueron
maacutes positivos cuanto mayor la superficie agriacutecola de la cuenca y maacutes negativos
cuanto mayor la superficie forestal Aunque por los alcances de esta tesis no
podemos ser concluyentes respecto del origen del NT (aloacutectonoautoacutectono) parece
ser que esta maacutes ligado a los aportes de la cuenca pese a la disminucioacuten del aporte
sedimentario observado en ambos lagos Las plantaciones forestales a diferencia
del bosque nativo son fertilizadas con una mezcla de N P y K (Toral et al 2005)
Por lo que pese a la disminucioacuten del aporte sedimentario la concentracioacuten de
126
nutrientes en el aporte al lago puede verse incrementada por el tipo de manejo
forestal con respecto al bosque nativo
Los resultados del primer capiacutetulo demuestran que 1) las plantaciones
forestales yo bosque nativo retiene maacutes sedimentos y MO mientras que el uso de
suelo agriacutecola y los pastizales destinados al desarrollo de la actividad ganadera lo
libera 2) Al parecer la desnitrificacioacuten es el proceso natural maacutes importante de
perdida de N en lagos costeros y favorece el enriquecimiento de 15N tanto en la
columna de agua (ie 15N-DIN) como en los sedimentos una vez enterrados y 3) la
desnitrificacioacuten depende de los cambios en las condiciones REDOX en el lago La
oxigenacioacuten en lagos poco profundos -como Laguna Matanzas- es altamente
fluctuante a escalas decadal-centenal depende de la profundidad de la laacutemina de
agua y de la variabilidad de la precipitacioacuten En este sentido en Laguna Matanzas
habriacutea condiciones anoacutexicas en ambientes cuando el lago alcanza niveles maacutes
altos Estas condiciones se observan entre 1820 y 1940 CE y son coetaacuteneas con
episodios maacutes huacutemedos y friacuteos (von Gunten et al 2009 Christie et al 2009) pero
tambieacuten con una fuerte actividad ganadera en la cuenca
Las fuentes variables de N y su fluctuacioacuten en el tiempo hacen necesario
contar con un anaacutelogo moderno que permite usar al δ15N de los sedimentos
lacustres como un indicador indirecto de los cambios en la disponibilidad de N en
el tiempo Por ello en el segundo capiacutetulo integramos muestras de suelo-
vegetacioacuten y un monitoreo de POM de la columna de agua en verano e invierno La
composicioacuten isotoacutepica de POM estariacutea reflejando cambios de la fuente de N (fijacioacuten
vs consumo del DIN) que variacutea durante el antildeo hidroloacutegico Durante el verano la
mayor temperatura y horas-luz favoreceriacutean las entradas de N al lago viacutea fijacioacuten
bioloacutegica de N2 atmosfeacuterico resultando en un incremento de la MO con un valor
127
isotoacutepico bajo (POM verano~5permil Fig 1b) El N2 (δ15N = 0permil) es fijado praacutecticamente
sin fraccionamiento isotoacutepico generando un δ15N de la OM cercano a 0permil (Leng et
al 2006) En invierno las precipitaciones pueden estar favoreciendo un incremento
en la entrada de nutrientes al lago El DIN de la columna de agua llega a contener
valores de δ15N maacutes altos reflejando estos mayores aportes de la cuenca y el POM
del Lago Vichuqueacuten queda enriquecido en 15N (δ15N ~11permil Cap 2 Fig 9) Estas
variaciones estacionales pueden estar evidenciando un cambio en la composicioacuten
de especies de POM desde especies fijadoras a especies que consumen el N de la
columna de agua Sin embargo para corroborar esta informacioacuten seriacutea deseable
contar con el anaacutelisis de la composicioacuten de especies de las muestras de agua
extraidas
Entre los resultados maacutes importantes estaacuten los valores isotoacutepicos del suelo
y la biomasa representativa de la cuenca que incluye un listado de las especies
nativas de la zona que hasta ahora no se contaba (Cap 2 Tabla en material
suplementario) Las especies colectadas despliegan valores de δ15Nfoliar maacutes
positivos (plantaciones forestales y pastizal ~89permil) que el suelo (35permil) salvo por
las especies nativas las que oscilan en torno a valores cercanos al atmosfeacuterico
(~14permil) La razoacuten isotoacutepica 15N14N refleja la dinaacutemica de intercambio de N entre la
vegetacioacuten y el suelo (Koba et al 2002) La discriminacioacuten es positiva en la mayoriacutea
de los sistemas bioloacutegicos por lo que las plantas tienen valores de δ15N maacutes bajos
que el suelo (Evans et al 2001) como sucede con las especies nativas en Lago
Vichuqueacuten En consecuencia las plantas quedan empobrecidas en 15N y conducen
a un enriquecimiento en 15N del suelo sobretodo si no hay reposicioacuten de N por otras
viacuteas (eg fijacioacuten de N) Por lo tanto los valores maacutes negativos de δsup1⁵Nfoliar de las
especies nativas pueden estar relacionados con el consumo preferencial de 14N del
128
suelo donde la discriminacioacuten isotoacutepica de la vegetacioacuten contra el 15N conduce a
valores del δsup1⁵Nsuelo maacutes altos (Houmlgberg 1997) Por el contrario los valores maacutes
positivos de δsup1⁵Nfoliar de las plantaciones forestales y pastos con respecto al suelo
puede deberse por una parte que el suelo no cuenta con mecanismos naturales de
reposicioacuten de N salvo por la descomposicioacuten de la hojarasca que puede ser maacutes
lenta en Pinus radiata que en bosque nativo (Lusk et al 2001) y por otra por el alto
impacto de los aportes de N (y otros nutrientes) derivado de las actividades
humanas (eg uso de fertilizantes) en el suelo
El alcance maacutes significativo de esta tesis se relaciona con un cambio en la
tendencia maacutes negativa de δsup1⁵N y el incremento del NT a partir de 1940 y a partir
de 1970 CE en ambos lagos La causa maacutes probable de esta tendencia es el
reemplazo de la cobertura natural o preexistente a 1940 CE por plantaciones
forestales
En la figura 2 se observa una siacutentesis de los principales procesos que
afectan la sentildeal isotoacutepica de la MO en los sedimentos lacustres de L Matanzas y
L Vichuqueacuten La entrada de N y MO desde la cuenca depende del tipo de cobertura
Las plantaciones forestales y el bosque nativo retienen maacutes nutrientes y sedimentos
en la cuenca mientras que la agricultura los favorece Sin embargo las praacutecticas
de manejo que incluyen la aplicacioacuten de N (y otros nutrientes) pueden aportar maacutes
nutrientes al lago que la cobertra de bosque nativo Cuando las actividades
forestales cubren una mayor superficie de la cuenca (1940 en adelante) δsup1⁵N oscila
en valores maacutes negativos y el NT tiende a incrementarse en los sedimentos
lacustres Cabe destacar que los valores de δsup1⁵N observados en los testigos
sedimentarios a fines del s XX no tienen precedente durante la conquista y colonia
espantildeola o durante el resto del periodo de la Repuacuteblica
129
Figura 2 Modelo de transferencia de N entre los ecosistemas terrestres y
acuaacuteticos El δ15N de los sedimentos lacustres depende de la interaccioacuten entre los
aportes de la cuenca y porcesos ldquoin-lakerdquo Flechas indican la direccioacuten del flujo de
N En azul las salidasentradas de 14N en rojo las salidasentradas de 15N δ15N de
la interaccioacuten plantas-suelo depende del tipo de cobertura de suelo
130
CONCLUSIONES GENERALES
La transferencia de N entre cuencas y lagos es un factor de control del ciclo
del N en los lagos y queda reflejado en la sentildeal isotoacutepica de los sedimentos
lacustres (Leng et al 2006 McLauchlan et al 2007) En Laguna Matanzas el
suelo de las especies nativas y las plantaciones forestales despliegan valores de
δ15N maacutes bajos (~1permil) que los de Lago Vichuqueacuten (~35permil) Coincidentemente los
sedimentos lacustres de Laguna Matanzas oscilan con valores de δ15N maacutes bajos
(-15 a +45permil Fig 1) que los de Lago Vichuqueacuten (+3 a +8permil)
Por otra parte el estudio de la MO de los sedimentos lacustres ha permitido
reconstruir los cambios en los aportes de MO y N desde la cuenca Aunque no es
posible afirmar queacute tipo de N estaacute entrando al lago (eg Nitroacutegeno orgaacutenico e
inorganico) si podemos decir que los cambios en las fluctuaciones de δ15N son
coincentes con el crecimiento de la actividad forestal (1980 - hasta 2016 L
Matanzas 0-17 y L Vichuqueacuten 1-66) El δ15N fluctuacutea hacia valores maacutes
negativos Ademaacutes se observa un incremento del NT en los sedimentos lacustres
cuanto mayor es la superficie forestal
Entre 1800-1940 las cuencas eran fundamentalmente agriacutecolas y
ganaderas (Cap1 Fig 7 y Cap 2 Fig 12) el δ15N de los sedimentos lacustres
oscila con valores maacutes positivos (L Matanzas +40 plusmn 05permil y L Vichuqueacuten 56 plusmn
033permil Fig 1) hay una baja bioproductividad en el lago (BrTi ~002 TOC~17)
lo que coincide con un bajo contenido de NT (~478 μg) Este es un periacuteodo de altas
precipitaciones (Christie et al 2009) que tienden a favorecer el lavado de la cuenca
131
y con ello el aporte de MO sedimentos y nutrientes desde la cuenca tiende a verse
favorecido Aunque las principales actividades humanas en estas cuencas son
diferentes (ganaderiacutea en Laguna Matanzas Contreras-Lopez et al 2014
agricultura en Lago Vichuqueacuten Odone 1997) en ambos casos hubo un reemplazo
de la cobertura natural de bosques nativo que favorecioacute mayores aportes de N y
sedimentos desde la cuenca en un efecto sumado con el aumento de las
precipitaciones
A partir de 1980 CE en ambos lagos se observa una disminucioacuten en los
valores de δ15N y un incremento del NT que no tiene precedentes en todo el registro
y pese a que ambos lagos son limnologicamente muy diferentes En Lago
Vichuqueacuten el δ15N tiende a oscilar en antifase con el NT (Fig 1) En el caso de
Laguna Matanzas se observa una tendencia hacia valores maacutes negativos y a partir
de 1990s a valores maacutes positivos mientras que el NT tiende a incrementarse de
manera sostenida Ello podriacutea estar relacionado con el crecimiento de la actividad
forestal habriacutea una mayor retencioacuten de N y sedimentos en la cuenca ligado al
incremento de la superficie forestal Ademaacutes en el caso de L Matanzas el
incremento de la actividad agriacutecola (sumado al de las plantaciones forestales)
podriacutea expicar el cambio en la tendencia en la deacutecada de los 1990s
En el contexto de Antropoceno esta tesis nos permite identificar un gran
impacto de las actividades humanas en Chile central a partir de la deacutecada de 1940
y una Gran Aceleracoacuten a partir de la deacutecada de 1970s En el registro sedimentario
de los lagos en estudio se observa una gran acumulacioacuten de MO el δ15N oscila
hacia valores maacutes negativos y un gran incremento del NT y el crecimiento de la
actividad forestal parece ser la principal responsable Ademaacutes la sobreexplotacioacuten
132
del agua junto al cambio climaacutetico global ha generado una combinacioacuten letal para
los lagos costeros de Chile central
Figura 1 Comparacioacuten de las fluctuaciones de δ15N durante los uacuteltimos 300
antildeos en Laguna Matanzas y Lago Vichuqueacuten
133
Referencias
Battye W Aneja VP Schlesinger WH 2017 Earthrsquos Future Is nitrogen the next carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia
UC de V 2014 Elementos de la historia natural del An Mus Hist Natulas Vaplaraiso 27 51ndash67
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Evans RD Evans RD 2001 Physiological mechanisms influencing
plant nitrogen isotope composition 6 121ndash126 Fogel ML Cifuentes LA 1993 Isotope fractionation during primary
production Org Geochem 73ndash98 httpsdoiorg101007978-1-4615-2890-6_3 Galloway JN Schlesinger WH Ii HL Schnoor L Tg N 1995
Nitrogen fixation Anthropogenic enhancement-environmental response reactive N Global Biogeochem Cycles 9 235ndash252
Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J
Christie D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Koba K Koyama LA Kohzu A Nadelhoffer KJ 2003 Natural 15 N
Abundance of Plants Coniferous Forest httpsdoiorg101007s10021-002-0132-6
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos
Transactions American Geophysical Union httpsdoiorg1010292007EO070007 McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE
2007 Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
134
Phytologist N Oct N 2006 Tansley Review No 95 15N Natural Abundance in Soil-Plant Systems Peter Hogberg 137 179ndash203
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Turner BL Kasperson RE Matson PA McCarthy JJ Corell RW
Christensen L Eckley N Kasperson JX Luers A Martello ML Polsky C Pulsipher A Schiller A 2003 A framework for vulnerability analysis in sustainability science Proc Natl Acad Sci U S A httpsdoiorg101073pnas1231335100
Vitousek PM Aber JD Howarth RW Likens GE Matson PA
Schindler DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
Vitousek PM Howarth RW 1991 Nitrogen limitation on land and in the
sea How can it occur Biogeochemistry httpsdoiorg101007BF00002772 von Gunten L Grosjean M Rein B Urrutia R Appleby P 2009 A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 Holocene 19 873ndash881 httpsdoiorg1011770959683609336573
Xu R Tian H Pan S Dangal SRS Chen J Chang J Lu Y Maria
Skiba U Tubiello FN Zhang B 2019 Increased nitrogen enrichment and shifted patterns in the worldrsquos grassland 1860-2016 Earth Syst Sci Data httpsdoiorg105194essd-11-175-2019
Zaehle S 2013 Terrestrial nitrogen-carbon cycle interactions at the global
scale Philos Trans R Soc B Biol Sci 368 httpsdoiorg101098rstb20130125
9
RESUMEN
El ldquoAntropocenordquo se caracteriza por pertubaciones de origen humano que
conllevan impactos globales Por ejemplo los cambios de usocobertura del suelo
(LUCC) son conocidos por perturbar el ciclo del N a partir de la Revolucioacuten Industrial
pero especialmente desde la Gran Aceleracioacuten (1950 CE) en adelante Sin
embargo existen incertezas asociadas a la magnitud del impacto y su efecto
acoplado con los cambios climaacuteticos actuales (ie megasequiacutea disminucioacuten de las
precipitaciones) y acoplamientos con otros ciclos biogeoquiacutemicos (eg ciclo del
Carbono) Este impacto ha cambiado la disponibilidad de N en los ecosistemas
terrestres y acuaacuteticos y sus enlaces Los sedimentos lacustres contienen
informacioacuten de las condiciones paleoambientales del lago y su cuenca en el
momento en que se depositaron y el anaacutelisis de los isoacutetopos estables de N (δ15N)
en los sedimentos permiten reconstruir los cambios en la disponibilidad de N a
traveacutes del tiempo En este trabajo usamos una aproximacioacuten multiproxy que incluye
10
anaacutelisis sedimentoloacutegicos geoquiacutemicos e isotoacutepicos (δ15N δ13C) de los sedimentos
lacustres columna de agua y suelo-vegetacioacuten de la cuenca y la reconstruccioacuten de
los LUCC entre 1975 y 2016 a partir de imaacutegenes satelitales El objetivo de esta
tesis fue evaluar el rol de LUCC como principal control del ciclo del N en el sistema
cuenca-lago durante las uacuteltimas centurias en Chile central Nuestros principales
resultados muestran que valores maacutes positivos de δ15N en los sedimentos lacustres
estaacuten relacionados con grandes aportes de N de la cuenca que a su vez son
mayores cuanto mayor la superficie agriacutecola yo los pastizales mientras que tanto
las plantaciones forestales como los bosques favorecen la retencioacuten de nutrientes
en la cuenca (lo que se ve reflejado en un δ15N maacutes negativo) Esta tesis plantea
un cambio de estado en la dinaacutemica del N asociado a la introduccioacuten y expansioacuten
de las plantaciones forestales o bosques los que retienen el Nitroacutegeno en las
cuencas mientras que el clima juega un rol secundario
11
ABSTRACT
The Anthropocene is characterized by human disturbances at the global
scale For example changes in land use are known to disturb the N cycle since the
industrial revolution but especially since the Great Acceleration (1950 CE) onwards
This impact has changed N availability in both terrestrial and aquatic ecosystems
However there are some important uncertainties associated with the extent of this
impact and how it is coupled to ongoing climate change (ie megadroughts rainfall
variability and shifting stationarity) or to other nutrient cycles (e g carbon cycle)
Lake sediments contain paleoenvironmental information regarding the conditions of
the watershed and associated lakes and which the respective sediments are
deposited Nitrogen stable isotope analysis (δ15N) on lake sediments allows us to
reconstruct the changes in N availability through time Here we used a multiproxy
approach that uses sedimentological geochemical and isotopic analyses on
lacustrine sediments water column and soilvegetation from the watershed as well
12
as land use and cover change (LUCC) from 1975 to 2016 obtained from satellite
images The goal of this thesis is to evaluate the role of LUCC as the main driver for
N cycling in a coastal watershed system of central Chile over the last centuries Our
main results show that more positive δ15N values in lake sediments are related to
higher N contributions from the watershed which in turn increase with increased
agricultural andor pasture cover whereas either forest plantations or native forests
can favor nutrient retention in the watershed (δ15N more negative) This thesis
proposes that N dynamics are mainly driven by the introduction and expansion of
forest or tree plantations that retain nitrogen in the watershed whereas climate plays
a secondary role
13
INTRODUCCIOacuteN
El N es un elemento esencial para la vida y limita la productividad en
ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al 1997) Las actividades
humanas han tenido un profundo impacto sobre el ciclo del N global principalmente
a partir la revolucioacuten industrial (IPCC 2013 2007) Las concentraciones de N se
han incrementado por la fijacioacuten industrial de N atmosfeacuterico (viacutea el proceso Haber-
Bosch Erisman et al 2008) por el uso de fertilizantes sobre la base de N para
mejorar el rendimiento de los cultivos la produccioacuten ganadera y ademaacutes de los
cambios en el uso del suelo (abreviado en ingleacutes- LUCC) (Robertson y Vitousek
2009 Galloway et al 2008 Battye et al 2017 Xu et al 2019) Estas actividades
contribuyen significativamente al incremento de la disponibilidad bioloacutegica de N
cuyas consecuencias para los ecosistemas incluye la perdida de diversidad
modificacioacuten de la disponibilidad de nutrientes y acidificacioacuten de los suelos entre
otras Sin embargo se desconoce la cantidad destino y el alcance que ha tenido
14
el N movilizado entre los ecosistemas generado por la influencia de las actividades
humanas (Vitousek et al 1997)
La entrada natural de N en los ecosistemas terrestres y acuaacuteticos es viacutea
fijacioacuten atmosfeacuterica la que es llevada a cabo por microorganismos muchos de ellos
en relaciones simbioacuteticas (eg Rhizobium en leguminosas) y las algas (Vitousek et
al 1997 Battye et al 2017) Las principales salidas estaacuten relacionadas con la
desnitrificacioacuten volatilizacioacuten del amoniaco y por escurrimiento superficial y
subterraacuteneo (Galloway et al 2008 Diacuteaz et al 2016) En el caso de los sistemas
lacustres ademaacutes estaacuten la resuspencioacuten de N del fondo del lago los aportes desde
la cuenca (entradas de N) La depositacioacuten de MO en el fondo del lago que marca
la salida de N de la columna de agua Estas relaciones de intercambio de N tienen
un control geograacutefico (eg latitud exposicioacuten relieve etc) climaacutetico y biotico
(McLauchlan et al 2013) La alteracioacuten provocada por los LUCC no solo genera
las peacuterdidas de nutrientes del suelo por erosioacuten y escurrimiento superficial sino que
tambieacuten modifica la disponibilidad y transferencia de N entre los ecosistemas
terrestres y acuaacuteticos (Elbert et al 2012 McLauchlan et al 2013) Por ejemplo el
reemplazo de la vegetacioacuten nativa por la agricultura yo plantaciones forestales
altera el ciclo del N debido al despeje de los terrenos viacutea quema de la vegetacioacuten
pero tambieacuten debido al reemplazo de la vegetacioacuten preexistente por un
monocultivo (eg trigo pino) y el uso ineficiente de fertilizantes para mejorar el
rendimiento agriacutecola Estas actividades favorecen un incremento de los aportes de
N (y otros nutrientes) viacutea sistemas de drenajes a lagos los que actuacutean como
sumideros El incremento del N derivado de las actividades humanas tanto en los
ecosistemas terrestres como acuaacuteticos genera incertezas relacionados con la
trayectoria y la magnitud de la disponibilidad de N en ambos ecosistemas (Batye et
15
al 2017 Mclauchlan et al 2017) Por lo tanto se requiere de una liacutenea base de
largo plazo para saber coacutemo estas perturbaciones impactan la disponibilidad de N
en las uacuteltimas deacutecadassiglos en los ecosistemas lacustres y por lo tanto el alcance
real que los LUCC han tenido en el ciclo del N
Los ecosistemas mediterraacuteneos y el ciclo del N
Los ecosistemas mediterraacuteneos son especialmente sensibles a los LUCC
pues estos se encuentran sometidos a un estreacutes hiacutedrico durante largas temporadas
estivales y las precipitaciones se concentran en eventos puntuales y a veces con
altos montos pluviomeacutetricos Las precipitaciones tienen un alto potencial de arrastre
de sedimentos y nutrientes los que actuacutean como fertilizantes al llegar a los
ecosistemas lacustres A su vez un incremento en la disponibilidad de N puede
generar un desbalance con otros nutrientes (eg Fosforo) con efectos sobre la
productividad y diversidad de los lagos (Pentildeuelas et al 2012 Sardans et al 2012
McLauchlan 2013) En este sentido en Chile central se observa una disminucioacuten
de las precipitaciones especialmente fuerte en las uacuteltimas deacutecadas por lo que se ha
denoacuteminado como Megasequiacutea (Garreaud et al 2017) y el efecto de las
precipitaciones esporaacutedicas pueden generar un arrastre de nutrientes que no ha
sido evaluado
Los ecosistemas mediterraacuteneos concentran el 20 de la biodiversidad global
(Myers et al 2000) pero existe una escasez de conocimiento respecto a los
efectos del incremento de N en los cuerpos de agua como consecuencia de las
actividades humanas en el pasado histoacuterico (Ochoa-Hueso et al 2011) y coacutemo la
disponibilidad de N ha variado en el tiempo Los LUCC han aumentado el flujo de
N en las cuencas hidrograacuteficas viacutea escurrimiento superficial y lixiviacioacuten
16
favoreciendo la peacuterdida de sedimentos y nutrientes del suelo en el mundo entero
(Elser et al 2011 McLauchlan et al 2013 Xu et al 2017 y 2019) Ello ha
contribuido a la acidificacioacuten y eutrofizacioacuten generalizada de riacuteos y lagos
(McLauchlan et al 2013 Schindler et al 2008)
El ecosistema mediterraacuteneo de Chile central es una regioacuten ampliamente
intervenida desde al menos la colonizacioacuten espantildeola (1600-1800 CE) Los LUCC
han tenido efectos negativos en la disponibilidad de agua especialmente
observados con el desarrollo de las actividades minera (Prieto et al 2019) Aunque
se sabe de los impactos en los ecosistemas de bosque en teacuterminos de cobertura
debidos al desarrollo silvo-agropecuarias (Armesto et al 2010) se desconoce el
impacto de estas actividades sobre el ciclo del N en los cuerpos de agua o en queacute
momento cobran fuerza suficiente para alterar el flujo de N hacia los lagos de Chile
Central Por lo tanto en esta tesis nos centramos en entender coacutemo los LUCC han
afectado la disponibilidad de N en los lagos costeros de Laguna Matanzas y Lago
Vichuqueacuten desde la llegada de los espantildeoles a Chile hasta el presente
Los lagos como sensores ambientales
Los sedimentos lacustres son buenos sensores de cambios en los aportes
de nutrientes contaminacioacuten y eutroficacioacuten de los cuerpos de agua ya que son
capaces de preservar sentildeales quiacutemicas y fiacutesicas al momento de su formacioacuten y
ademaacutes con alta resolucioacuten y a largo plazo (Leng et al 2006) Por lo tanto
constituyen una herramienta uacutetil para reconstruir el flujo de N entre ecosistemas
terrestres y acuaacuteticos en el pasado sus consecuencias (eg incremento de la
productividad lacustre) y el rol del clima y los LUCC en el pasado (McLauchlan et
al 2013 von Gunten et al 2009) Por ejemplo el cultivo de maiacutez por parte de los
17
nativos iroqueses en Canadaacute provocoacute la eutrofizacioacuten del Lago Crawford (43degN)
durante los siglos XII y XV Entre las consecuencias registradas se documentoacute un
claro incremento de la productividad primaria y cambios en la estructura comunitaria
de diatomeas foacutesiles (incremento de Stephanodiscus complex y disminucioacuten de
Cyclotella bodanica en Ekdahl et al 2007) Otro ejemplo demuestra como las
actividades agriacutecolas en la cuenca del riacuteo Mississippi modificaron los patrones de
sedimentacioacuten y aporte de nutrientes al lago Pepin EE UU (44degN) a partir del
asentamiento europeo en 1840 CE y principalmente desde 1940 CE (Engstrom et
al 2009) Para Chile von Gunten et al (2009) a partir de indicadores
limnogeoloacutegicos evidencioacute la contaminacioacuten y eutroficacioacuten de cinco lagos cercanos
a la ciudad de Santiago (33ordmS) como consecuencia de la deposicioacuten atmosfeacuterica
de metales pesados (especialmente Cu) quema de combustible foacutesil y flujo de
nutrientes durante los uacuteltimos 200 antildeos
Caracteriacutesticas limnoloacutegicas de los lagos
Los procesos limnoloacutegicos afectan la distribucioacuten y abundancia de los
organismos en los lagos Estaacuten influenciados por forzamientos externos por
ejemplo la precipitacioacuten o la penetracioacuten de la luz y la morfometriacutea del lago En este
sentido el nivel del lago dependeraacute del balance entre las entradas y salidas de agua
(precipitaciones escurrimiento superficial y subterraacuteneo y la evaporacioacuten) y la forma
de la cuenca (profundidad pendiente aacuterea del espejo de agua)
En funcioacuten de la profundidad de penetracioacuten de la luz es posible diferenciar
dos aacutereas que determinan la distribucioacuten de los organismos la zona foacutetica (donde
penetra la luz y permite la fotosiacutentesis) y la zona afoacutetica (subyacente a la zona
foacutetica) Al depender de la radiacioacuten la zona foacutetica variacutea estacionalmente Ademaacutes
18
puede variar por el grado de turbidez la morfometriacutea del lago y la produccioacuten de
materia orgaacutenica en la columna de agua
Otro factor que influye en la productividad es el reacutegimen de mezcla de la
columna de agua pues favorece la oxigenacioacuten y el reciclado de nutrientes La
mezcla ocurre en condiciones de gran inestabilidad usualmente relacionado con el
reacutegimen de viento Por el contrario un lago estratificado resulta de grandes
diferencias en temperatura o salinidad entre la superficie (epilimnion) y el fondo del
lago (hipolimnion) que separa las masas de agua superficial y de fondo por una
termoclina (si la estratificacioacuten estaacute relacionada con diferencias de temperatura de
las masas de agua) o haloclina (diferencias de salinidad) En funcioacuten del reacutegimen
de mezcla los lagos se pueden clasificar en (Lewis 1983)
1 Amiacutecticos no hay mezcla vertical de la columna de agua
2 Monomiacutecticos friacuteos y caacutelidos se mezcla una vez al antildeo
3 Dimiacutecticos la columna de agua se mezcla dos veces al antildeo
4 Oligomiacutecticos Lagos estratificados la mayor parte del tiempo pero se mezclan a
intervalos irregulares mayores a 1 antildeo
5 Polimiacutecticos friacuteos y caacutelidos la columna de agua se mezcla varias veces al antildeo
El ciclo del N en lagos
Al estar presente en aacutecidos nucleicos aminoaacutecidos y proteiacutenas el N es un
nutriente esencial de los organismos Por lo tanto su disponibilidad en la columna
de agua es clave para la produccioacuten composicioacuten y acumulacioacuten de la MO a traveacutes
del tiempo (Olson 1998 Talbot 2002) Aunque el principal reservorio de N estaacute en
19
la atmosfera (N2) solo unos pocos organismos (eg cianobacterias) pueden fijarlo
directamente Asiacute el Nitroacutegeno Inorgaacutenico Disuelto (DIN) representa la principal
fuente de N disponible para la produccioacuten primaria en los ecosistemas acuaacuteticos
(Talbot et al 2002) El DIN estaacute compuesto por nitrato (NO3-) nitrito (N02
-) y amonio
(NH4+) y los cambios en su disponibilidad condicionan el tipo de produccioacuten primaria
(eg fijadores vs asimiladores de N ver Scott y Grant 2013 Scott 2008)
La Figura 1 resume los principales componentes en lagos del ciclo del N y
sus interacciones La principal entrada natural de N es viacutea fijacioacuten de N atmosfeacuterico
y es llevado a cabo principalmente por bacterias y algas verde-azules capaces de
romper el triple enlace entre los dos aacutetomos de N a un alto costo energeacutetico (Torres
et al 2012) El N fijado es reducido a NH3 Tras la muerte de los organismos el N
es liberado de sus tejidos por hongos y bacterias implicados en la descomposicioacuten
de la MO Una parte se deposita en el fondo del lago y la otra queda disponible para
ser asimilada por el fitoplancton como amonio mediante el proceso de
amonificacioacuten (Talbot 2002) La amonificacioacuten resulta de la reduccioacuten bacteriana
del amoniaco y se da preferentemente en condiciones anoacutexicas La volatizacioacuten del
amoniaco es un proceso por medio este se incorpora a la atmoacutesfera Este proceso
se da preferentemente en lagos aacutecidos (pH gt 85) El nitrato es otra forma de N
bioloacutegicamente disponible para el fitoplancton En los lagos el NO3 del DIN estaacute
compuesto por los aportes desde la cuenca hidrograacutefica en la que se circunscriben
por la resuspensioacuten desde el fondo del lago favorecido por procesos de mezcla
(descrito en seccioacuten anterior) y por los nitratos de la columna de agua Estos uacuteltimos
son el resultado de la nitrificacioacuten proceso realizado por bacterias aeroacutebicas
mediante el cual el amonio se oxida para formar nitrito y nitrato El ciclo se completa
20
con la desnitrificacioacuten que es la reduccioacuten bacteriana de nitrato al N atmosfeacuterico
Este proceso se da preferentemente en condiciones anoacutexicas
Figura 1 Materia Orgaacutenica sedimentaria y su relacioacuten con el ciclo del N y las
variaciones de δ15N (elaborado a partir de Talbot et al 2002) En la figura se
representan los factores clave en la acumulacioacuten de la MO sedimentaria y su
relacioacuten con el ciclo del N El δ15N de la MO es resultado de los aportes de MO
desde la cuenca MO de macroacutefitas y de la productividad bioloacutegica La productividad
en ecosistemas lacustres estaacute condicionada por DIN y la fijacioacuten de N atmosfeacuterico
El δ15N del pool del DIN disponible se va enriqueciendo por la asimilacioacuten
preferencial de 14N por parte del fitoplancton Ademaacutes el DIN remanente se va
enriqueciendo por accioacuten microbiana (amonificacioacuten nitrificacioacuten desnitrificacioacuten)
Reconstruyendo el ciclo del N a partir de variaciones en δ15N
La sentildeal isotoacutepica de N (δ15N) en los sedimentos lacustres puede ser usada
para reconstruir los cambios pasados del ciclo N la transferencia de N entre
ecosistemas terrestres y acuaacuteticos y el estado troacutefico del lago (Bruland y Mackenzie
2015 McLauchlan et al 2013 Woodward et al 2012 von Gunten et al 2009
Leng et al 2006 Meyers y Teranes 2001) La Figura 1 resume los principales
procesos y fuentes que afectan la sentildeal isotoacutepica δ15N (total o ldquobulkrdquo) en la MO de
21
los sedimentos lacustres Entre los que destacan las fuentes de MO (aloacutectona vs
autoacutectona) la bioproductividad lacustre y las entradas de N (ie fijacioacuten bioloacutegica
de N2) (Torres et al 2012) Estos procesos conducen a un fraccionamiento
isotoacutepico cineacutetico que favorece la disociacioacuten entre sus isotopos ldquopesadosrdquo (15N) y
ldquoligerosrdquo (14N) Asiacute por ejemplo los valores de δ15N de la MO se enriquecen en 15N
en la medida que los sedimentos lacustres estaacuten sometidos a perdidas de 14N viacutea
desnitrificacioacuten (Bruland y Mackenzie 2015 Diaz et al 2016 Talbot et al 2002)
Por otra parte el fitoplancton en la medida que aumenta su biomasa (eg
durante la estacioacuten caacutelida) puede llegar a asimilar el DIN hasta agotarse En este
caso el N remanente se va empobreciendo en 14N si no hay reposicioacuten de N (eg
aporte de NO3 de la cuenca) y la MO queda enriquecida en 15N Esta disminucioacuten
induce a una limitacioacuten de fitoplancton por N y los organismos diztroacuteficos pueden
verse estimulados por lo que se incrementa la entrada de N viacutea fijacioacuten de N2 (Scott
y Grantsz 2013) como resultado la MO oscila en valores maacutes bajos (en torno a 0permil)
La cantidad de MO que se deposita en el fondo del lago depende del
predominio de contribuciones provenientes desde la cuenca (aloacutectona) versus las
producidas dentro del mismo lago (autoacutectona) (Torres et al 2012) Aunque en
general los lagos reciben permanentemente aportes de sedimentos y MO desde
su cuenca los lagos pequentildeos tienden a reflejar maacutes los procesos que ocurren
solamente en su cuenca directa y reflejan los valores isotoacutepicos de la misma (Gu et
al 2006) Woodward et al (2012) realizaron un estudio en 50 lagos en Irlanda que
les permitioacute correlacionar diferentes patrones de LUCC (bosques de coniacuteferas
agricultura ganaderiacutea entre otros) con los valores de δ15N (y δ13C) en los
sedimentos superficiales de los lagos Encontraron que los valores de δ15N son maacutes
negativos en cuencas poco impactadas y maacutes positivos en aquellas con alto
22
impacto humano (eg usos de suelo agriacutecola y ganadero) McLauchlan et al (2007)
encontraron valores maacutes positivos de δ15N en los sedimentos del Mirror Lake New
Hampshire (29ordm N) a partir de la desforestacioacuten de su cuenca en 1790 CE (inicio
del asentamiento europeo en el lago) para el desarrollo de la actividad agriacutecola
Estos valores se volvieron maacutes negativos hacia valores similares al pre-
asentamiento europeo despueacutes del cese de la agricultura en la cuenca y la
recuperacioacuten del bosque a partir de 1929 CE
El anaacutelisis de δ15N en los sedimentos lacustres es una herramienta casi sin
explorar en Chile central Proponemos reconstruir la dinaacutemica de transferencia de
N cuenca-lago como consecuencia de los LUCC desde la colonizacioacuten espantildeola en
los lagos costeros de Laguna Matanzas (34ordmS) y Lago Vichuqueacuten (36ordmS) Como son
muacuteltiples los procesos que pueden afectar la sentildeal isotoacutepica de N (medido como
δ15N) en los sedimentos lacustres existen muchos problemas para su
interpretacioacuten paleoambiental (Leng et al 2006) Por ello en esta tesis utilizamos
un enfoque multiproxy que consiste en integrar el anaacutelisis limnoloacutegico y geoquiacutemico
de los sedimentos lacustres con los valores isotoacutepicos modernos de la columna de
agua (mediante monitoreo del POM) la relacioacuten vegetacioacuten-suelo de la cuenca y la
reconstruccioacuten de los principales usos de suelo de las cuencas desde 1975 CE
mediante el uso de imaacutegenes satelitales Considerando estas muacuteltiples liacuteneas de
evidencia nos planteamos utilizar las variaciones del δ15N para reconstruir los
cambios en la disponibilidad de N en los lagos costeros de Chile central y establecer
coacutemo y desde cuaacutendo las actividades humanas en la cuenca son lo suficientemente
importantes como para modificar el ciclo del N en los lagos desde la colonizacioacuten
espantildeola (siglo XVII)
23
Como hipoacutetesis planteamos que la dinaacutemica de transferencia de sedimentos
y nutrientes entre la cuenca y el lago tiene controles geoloacutegicos climaacuteticos y
bioloacutegicos que interactuacutean en el tiempo registraacutendose en la acumulacioacuten de MO de
los sedimentos lacustres (Fig1) Bajo este marco los LUCC modifican esta
dinaacutemica alterando la transferencia de N a los lagos ya que las actividades agriacutecolas
y ganaderas tienden a aportar maacutes N (aumentando δ15N) que la actividad forestal
(la que disminuye δ15N)
En el primer capiacutetulo titulado ldquoA combined approach to establishing the timing
and magnitude of anthropogenic nutrient alteration in a mediterranean coastal lake-
watershed systemrdquo se evaluoacute el rol de los LUCC en el control de la descarga de N
y sedimentos a Laguna Matanzas en un sistema cuenca-lago desde el siglo XVII
Para ello se realizoacute un anaacutelisis de la dinaacutemica sedimentaria lacustre mediante el
anaacutelisis de muacuteltiples proxys sedimentoloacutegicos (ie minerales y textura)
geoquiacutemicos (eg geoquiacutemica orgaacutenica e isotoacutepica) y bioloacutegicos (ie diatomeas) de
Laguna Matanzas que cubren los uacuteltimos 200 antildeos Ademaacutes se realizoacute una
reconstruccioacuten de los usos de suelo entre 1975 y 2016 a partir de imaacutegenes de
sateacutelites y se colectaron muestras de suelo de las principales coberturas de la
cuenca a los cuales se midioacute el δ15N
Entre los principales resultados obtenidos se destaca la influencia de la
ganaderiacutea y la denitrificacioacuten lo que explica los altos valores de δ15N evidenciados
por el registro lacustre durante la colonizacioacuten espantildeola e inicios de la repuacuteblica A
partir de la Gran Aceleracioacuten (1950 CE) la desforestacioacuten y el reemplazo de la
ganaderiacutea por plantaciones forestales tienen un correlato en el registro
sedimentario con valores de δ15N maacutes bajos Estos resultados demuestran que los
LUCC son el factor de primer orden para explicar los cambios observados en
24
nuestro registro de δ15N y a su vez implica un rol secundario del clima como posible
control de la disponibilidad de N en Laguna Matanzas Este capiacutetulo fue sometido
a la revista Scientific Reports y estaacute en revisioacuten desde el 26 de marzo de 2019 En
la elaboracioacuten del mauscrito han colaborado Claudio Latorre Blas Valero-Garceacutes
Matiacuteas Frugone Pablo Sarricolea Santiago Giralt Manuel Contreras-Lopez
Ricardo Prego y Patricia Bernardez
El segundo capiacutetulo se titula ldquoStable isotopes track land use and cover
changes in a mediterranean lake in central Chile over the last 600 yearsrdquo Aquiacute
evaluamos la dinaacutemica del N actual en el sistema cuenca-lago considerando los
valores isotoacutepicos modernos tanto de la vegetacioacuten como del suelo ademaacutes de los
cambios estacionales del POM de la columna de agua Esta informacioacuten se utiliza
como un anaacutelogo moderno para conocer el rol de los LUCC en la disponibilidad de
N en los uacuteltimos 600 antildeos Para ello se colectaron muestras filtradas de la columna
de agua a 2 5 y 20 m dos veces por antildeo entre noviembre de 2015 y julio de 2018
y se obtuvo el δsup1⁵N del POM Ademaacutes se colectoacute muestras de vegetacioacuten y suelo
de la cuenca diferenciando entre especies nativas plantaciones forestales y
vegetacioacuten herbaacutecea Esta informacioacuten se analizoacute en conjunto con la reconstruccioacuten
de los LUCC entre 1975 y 2014 y el anaacutelisis limnoloacutegico del lago Ello nos permitioacute
evaluar coacutemo el δ15N de los sedimentos lacustres refleja los valores δ15N de la
cuenca en el Lago Vichuqueacuten y el control que ejerce el clima en la sentildeal isotoacutepica
de POM Este capiacutetulo seraacute sometido a la revista Science of total environmet
proximamente En la elaboracioacuten del mauscrito han colaborado Claudio Latorre
Blas Valero-Garceacutes Matiacuteas Frugone Pablo Sarricolea Leonardo Villacis y M Laura
Carrevedo
25
Entre los principales resultados encontramos que el δ15N en los sedimentos
lacustres es maacutes positivo con presencia de agricultura en la cuenca y maacutes negativo
cuando esta es reemplazada por cobertura arboacuterea bien sea plantaciones
forestales bien sea bosque nativo A su vez durante la estacioacuten seca y caacutelida la
mayor entrada al lago es viacutea fijacioacuten de N atmosfeacuterico (bajos valores de δ15NPOM)
Durante la estacioacuten huacutemeda en cambio prevalecen los aportes de la cuenca con
altos valores de δ15NPOM Ello estariacutea evidenciando un cambio estacional en la
composicioacuten de especies en verano ligado a especies fijadoras de N y en invierno
las algas y microorganismos que consumen el DIN de la columna de agua
Referencias
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea AM Marquet PA 2010 From the Holocene to the Anthropocene A historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the
next carbon Earth rsquo s Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409 httpsdoiorg102134jeq20090005
Diacuteaz FP Frugone M Gutieacuterrez RA Latorre C 2016 Nitrogen cycling in an
extreme hyperarid environment inferred from δ15 N analyses of plants soils and herbivore diet Sci Rep 6 1ndash11 httpsdoiorg101038srep22226
Ekdahl EJ Teranes JL Wittkop CA Stoermer EF Reavie ED Smol JP
2007 Diatom assemblage response to Iroquoian and Euro-Canadian eutrophication of Crawford Lake Ontario Canada J Paleolimnol 37 233ndash246 httpsdoiorg101007s10933-006-9016-7
Elbert W Weber B Burrows S Steinkamp J Buumldel B Andreae MO
Poumlschl U 2012 Contribution of cryptogamic covers to the global cycles of carbon and nitrogen Nat Geosci 5 459ndash462
26
httpsdoiorg101038ngeo1486 Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506
httpsdoiorg101126science1215567 ARTICLE Engstrom DR Almendinger JE Wolin JA 2009 Historical changes in
sediment and phosphorus loading to the upper Mississippi River Mass-balance reconstructions from the sediments of Lake Pepin J Paleolimnol 41 563ndash588 httpsdoiorg101007s10933-008-9292-5
Erisman J W Sutton M A Galloway J Klimont Z amp Winiwarter W (2008)
How a century of ammonia synthesis changed the world Nature Geoscience 1(10) 636
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892
httpsdoiorg101126science1136674 Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J Christie
D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592 Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
IPCC 2013 Climate Change 2013 The Physical Science BasisContribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change Cambridge University Press Cambridge United Kingdom and New York NY USA
httpsdoiorg101017CBO9781107415324 Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007 Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32iumliquestfrac12S 3050iumliquestfrac12miumliquestfrac12asl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470
27
httpsdoiorg101073pnas0701779104 McLauchlan KK Gerhart LM Battles JJ Craine JM Elmore AJ Higuera
PE Mack MC McNeil BE Nelson DM Pederson N Perakis SS 2017 Centennial-scale reductions in nitrogen availability in temperate forests of the United States Sci Rep 7 1ndash7 httpsdoiorg101038s41598-017-08170-z
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Myers N Mittermeler RA Mittermeler CG Da Fonseca GAB Kent J
2000 Biodiversity hotspots for conservation priorities Nature httpsdoiorg10103835002501
Pentildeuelas J Poulter B Sardans J Ciais P Van Der Velde M Bopp L
Boucher O Godderis Y Hinsinger P Llusia J Nardin E Vicca S Obersteiner M Janssens IA 2013 Human-induced nitrogen-phosphorus imbalances alter natural and managed ecosystems across the globe Nat Commun 4 httpsdoiorg101038ncomms3934
Prieto M Salazar D Valenzuela MJ 2019 The dispossession of the San
Pedro de Inacaliri river Political Ecology extractivism and archaeology Extr Ind Soc httpsdoiorg101016jexis201902004
Robertson GP Vitousek P 2010 Nitrogen in Agriculture Balancing the Cost of
an Essential Resource Ssrn 34 97ndash125 httpsdoiorg101146annurevenviron032108105046
Sardans J Rivas-Ubach A Pentildeuelas J 2012 The CNP stoichiometry of
organisms and ecosystems in a changing world A review and perspectives Perspect Plant Ecol Evol Syst 14 33ndash47 httpsdoiorg101016jppees201108002
Schindler DW Howarth RW Vitousek PM Tilman DG Schlesinger WH
Matson PA Likens GE Aber JD 2007 Human Alteration of the Global Nitrogen Cycle Sources and Consequences Ecol Appl 7 737ndash750 httpsdoiorg1018901051-0761(1997)007[0737haotgn]20co2
Scott E and EM Grantzs 2013 N2 fixation exceeds internal nitrogen loading as
a phytoplankton nutrient source in perpetually nitrogen-limited reservoirs Freshwater Science 2013 32(3)849ndash861 10189912-1901
28
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Talbot M R (2002) Nitrogen isotopes in palaeolimnology In Tracking
environmental change using lake sediments (pp 401-439) Springer Dordrecht
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable
isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K
Yeager KM 2016 Different responses of sedimentary δ15N to climatic changes and anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
29
CAPIacuteTULO 1 A COMBINED APPROACH TO ESTABLISHING THE TIMING
AND MAGNITUDE OF ANTHROPOGENIC NUTRIENT ALTERATION IN A
MEDITERRANEAN COASTAL LAKE- WATERSHED SYSTEM
30
A combined approach to establishing the timing and magnitude of anthropogenic
nutrient alteration in a mediterranean coastal lake- watershed system
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugoneabc Pablo
Sarricolead Santiago Giralte Manuel Contreras-Lopezf Ricardo Prego g Patricia
Bernaacuterdez g Blas Valero-Garceacutesch
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Institute of Earth Sciences Jaume Almera (ICTJA-CSIC) CLluis Solegrave Sabaris sn Barcelona E-
08028 Spain
f Facultad de Ingenieriacutea y Centro de Estudios Avanzados Universidad de Playa Ancha Traslavintildea
450 Vintildea del Mar Chile
g Instituto de Investigaciones Marinas (CSIC) CEduardo Cabello 6 36208 Vigo Spain
h Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding author
E-mail address
clatorrebiopuccl magdalenafuentealbagmailcom
Abstract
Since the industrial revolution and especially during the Great Acceleration (1950
CE) human activities have profoundly altered the global nutrient cycle through land
use and cover changes (LUCC) However the timing and intensity of recent N
variability together with the extent of its impact in terrestrial and aquatic ecosystems
and coupled effects of regional LUCC and climate are not well understood Here
we used a multiproxy approach (sedimentological geochemical and isotopic
31
analyses historical records climate data and satellite images) to evaluate the role
of LUCC as the main control for N cycling in a coastal watershed system of central
Chile during the last few centuries The largest changes in N dynamics occurred in
the mid-1970s associated with the replacement of native forests and grasslands for
livestock grazing by plantations of introduced Monterey pine (Pinus radiata) and
eucalyptus (Eucalyptus globulus) Lake productivity increased and is evidenced by
an increase trend in δ15N values Our study shows that anthropogenic land
usecover changes are key in controlling nutrient supply and N availability in
Mediterranean watershed ndash lake systems and that large-scale forestry
developments during the mid-1970s likely caused the largest changes in central
Chile
Keywords Anthropocene Organic geochemistry watershedndashlake system Stable
Isotope Analyses Land usecover change Nitrogen cycle Mediterranean
ecosystems central Chile
1 INTRODUCTION
Human activities have become the most important driver of the nutrient cycles in
terrestrial and aquatic ecosystems since the industrial revolution (Gruber and
Galloway 2008 Galloway et al 2008 Holtgrieve et al 2011 Fowler et al 2013
Goyette et al 2016) Among these N is a common nutrient that limits productivity
in terrestrial and aquatic ecosystems (Vitousek and Howarth 1991 McLauchlan et
al 2013) With the advent of the Haber-Bosch industrial N fixation process in the
early 20th century total N fluxes have surpassed previous planetary boundaries
32
(Galloway et al 2008 Elser 2011) reaching unprecedented values (ie tipping
points) in the Earth system especially during what is now termed the Great
Acceleration (which began in the 1950s) which intensified in the 1970s (Howarth
2004 Steffen et al 2015) Yet dramatic changes in LUCC have occurred in the last
few centuries mainly linked to forestry and agro-pastoral activities (Poraj-Goacuterska et
al 2017 Ge et al 2019 Cheng et al 2019) These have had large impacts on N
(and Carbon -C-) availability in terrestrial and aquatic ecosystems but the synergic
effect with climate change and global N dynamics has not been established
(Vitousek et al 1997 Mclauchlan et al 2007 2013 Pongratz et al 2010
Woodward et al 2012 Mclauchlan et al 2017)
The onset of the Anthropocene poses significant challenges in mediterranean
regions that have a strong seasonality of hydrological regimes and an annual water
deficit (Stocker et al 2013) Mediterranean climates occur in all continents
(California central Chile Australia South Africa circum-Mediterranean regions)
providing a unique opportunity to investigate global change processes during the
Anthropocene in similar climate settings but with variable geographic and cultural
contexts The effects of global change in mediterranean watersheds have been
analyzed from different perspectives hydrology (Giorgi 2006 Arnell and Gosling
2013 Prudhomme et al 2014) vegetation dynamics (Lenihan et al 2003 Vicente-
Serrano et al 2013 Matesanz and Valladares 2014) sediment dynamics (Garciacutea-
Ruiz et al 2013 Garciacutea-Ruiz 2010 Syvitski et al 2005) changes in
biogeochemical cycles (Thomas et al 2013 Fowler et al 2013 Frank et al 2015)
carbon storage (Muntildeoz-Rojas et al 2015) and biodiversity (Hooper et al 2012) A
recent review showed an extraordinarily high variability of erosion rates in
mediterranean watersheds positive relationships with slope and annual
33
precipitation and the paramount effect of land use (Garciacutea-Ruiz et al 2015)
However the temporal context and effect of LUCC on nutrient supply to
mediterranean lakes has not been analyzed in much detail
Major LUCC in central Chile occurred during the Spanish Colonial period
(17th-18th centuries) (Armesto 1994 Gurnell et al 2001 Figueroa et al 2004
Martiacuten-Foreacutes et al 2012 Contreras-Loacutepez et al 2014) with the onset of
industrialization and mostly during the mid to late 20th century (von Gunten et al
2009 Gayo et al 2012) Recent copper pollution caused by 20th century mining
and industrial smelters has been documented in cores throughout the Andes
(Laguna el Ocho and Laguna Ensuentildeo von Gunten et al 2009) and also from our
surveys in the central valley (Batuco wetland) coastal range (Cordillera de Name)
and along the coast itself (Bucalemito and Colejuda) (Valero-Garceacutes et al 2010
unpublished data)
Paleolimnological studies have shown how these systems respond to
climate LUCC and anthropogenic impacts during the last millennia (Jenny et al
2002 2003 Villa Martiacutenez et al 2002 Frugone-Aacutelvarez et al 2017 Contreras et
al 2018) Furthermore changes in sediment and nutrient cycles have also been
identified in associated terrestrial ecosystems dating as far back as the Spanish
Conquest and related to fire clearance and wood extraction practices of the native
forests (Armesto 1994 Contreras-Loacutepez et al 2014) Nevertheless pollen and
limnological evidence argue for a more recent timing of the largest anthropogenic
impacts in central Chile For example paleo records show that during the mid-20th
century increased soil erosion followed replacement of native forest by Pinus
radiata and Eucalyptus globulus plantations at Laguna Matanzas Aculeo and
34
Vichuqueacuten lakes (Jenny et al 2002 Villa-Martiacutenez et al 2002 2003 Frugone-
Aacutelvarez et al 2017)
Lakes are a central component of the global carbon cycle Lakes act as a
sink of the carbon cycle both by mineralizing terrestrially derived organic matter and
by storing substantial amounts of organic carbon (OC) in their sediments (Anderson
et al 2009) Paleolimnological studies have shown a large increase in OC burial
rates during the last century (Heathcote et al 2015) however the rates and
controls on OC burial by lakes remain uncertain as do the possible effects of future
global change and the coupled effect with the N cycle LUCC intensification of
agriculture and associated nutrient loading together with atmospheric N-deposition
are expected to enhance OC sequestration by lakes Climate change has been
mainly responsible for the increased algal productivity since the end of the 19th
century and during the late 20th century in lakes from both the northern (Ruumlhland et
al2015) and southern hemispheres (Michelutti et al 2015 Carrevedo et al 2015)
but many studies suggest a complex interaction of global warming and
anthropogenic influences and it remains to be proven if climate is indeed the only
factor controlling these transitions (Catalaacuten et al 2013) Alternative causes for
recent N (Galloway et al 2008) increases in high altitude lakes such as catchment
mediated processes cannot be ruled out (Camarero and Catalan 2012 Fritz and
Anderson 2013) Few lake-watershed systems have robust enough chronologies of
recent changes to compare variations in C and N with regional and local processes
and even fewer of these are from the southern hemisphere (McLauchlan et al
2007 Holtgrieve et al 2011)
In this paper we present a multiproxy lake-watershed study including N and
C stable isotope analyses on a series of short cores from Laguna Matanzas in
35
central Chile focused in the last 200 years We complemented our record with land
use surveys satellite and aerial photograph studies Our major objectives are 1) to
reconstruct the dynamics among climate human activities and changes in the N
cycle over the last two centuries 2) to evaluate how human activities have altered
the N cycle during the Great Acceleration (since the mid-20th century)
2 STUDY SITE
Laguna Matanzas (33ordm45rsquoS 71ordm 40acuteW 7 m asl Fig 1) is a coastal lake located
in central Chile near to a large populated area (Santiago gt6106 inhabitants) The
lake has a surface area of 15 km2 with a max depth of 3 m and a watershed of 30
km2 The lake basin is emplaced over Pleistocene-Holocene aeolian and alluvial fan
deposits (SERNAGEOMIN 2003) A recent phase of dune activity occurred from the
mid to late Holocene which mostly sealed off the basin from the ocean (Villa-
Martinez 2002) Climate in Laguna Matanzas is characterized by cool-wet winters
and hot-dry summers with annual precipitation of ~510 mm and a mean annual
temperature of 12ordmC Central Chile is in the transition zone between the southern
hemisphere mid-latitude westerlies belt and the South Pacific Anticyclone (or SPA)
(Garreaud et al 2009) In winter precipitation is modulated by the north-west
displacement of the SPA the northward shift of the westerlies wind belt and an
increased frequency of storm fronts stemming off the Southern Hemisphere
Westerly Winds (SWW) (Frugone-Aacutelvarez et al 2017) Austral summers are
typically dry and warm as a strong SPA blocks the northward migration of storm
tracks stemming off the SWW
36
Historic land cover changes started after the Spanish conquest with a Jesuit
settlement in 1627 CE near El Convento village and the development of a livestock
ranch that included the Matanzas watershed After the Jesuits were expelled from
South America in 1778 CE the farm was bought by Pedro Balmaceda and had more
than 40000 head of cattle around 1800 CE (Contreras-Lopez et al 2014) The first
Pinus radiata and Eucalyptus globulus trees were planted during the second half of
the 19th century and were mostly used for dune stabilization (Albert 1900 Gibson
1972) However the main plantation phase occurred 60 years ago (Villa-Martinez
2002) as a response to the application of Chilean Forestry Laws promulgated in
1931 and 1974 and associated state subsidies
Major land cover changes occurred recently from 1975 to 2008 as shrublands
were replaced by more intensive land uses practices such as farmland and tree
plantations (Schulz et al 2010) Laguna Matanzas is part of the Reserva Nacional
Humedal El Yali protected area (Ramsar Ndeg 878) Despite this protected status the
lake and its watershed have been heavily affected by intense agricultural and
farming activities during the last decades The main inlet ldquoLas Rosasrdquo has been
diverted for crop irrigation causing a significant loss of water input to the lake
Consequently the flooded area of the lake has greatly decreased in the last couple
of decades (Fig 1b) Exotic tree species cover a large surface area of the
watershed Recently other activities such as farms for intensive chicken production
have been emplaced in the watershed
37
Figure 1 Laguna Matanzas a) Digital Elevation Model showing the watershed and
the surface hydrologic connection with Colejuda and Cabildo lakes b) Climograph
depicting the warm dry season in austral summer c) Annual precipitation from
1965-2015 (arrow shows start of the most recent ldquomega-droughtrdquo- see Garreaud et
al 2017) d) A decline of lake area occurs between 2007 and 2018 Lake surface
area decreased first along the western sector (in 2007) followed by more inland
areas (in 2018)
38
3 RESULTS
31 Age Model
The age model for the Matanzas sequence was developed using Bacon software
to establish the deposition rates and age uncertainty (Blaauw and Christen 2011)
It is based on 210Pb and 137Cs dates and two 14C dates (Table 2) According to this
age model the lake sequence spans the last 1000 years (Fig 2) A major breccia
layer (unit 3b) was deposited during the early 18th century which agrees with
historic documents indicating that a tsunami impacted Laguna Matanzas and its
watershed in 1730 CE (Contreras-Loacutepez et al 2014) Here we focus in the last 200
years were the most important changes occurred in terms of LUCC (after the
sedimentary hiatus caused by the tsunami) The Spanish colonial period (17th -18th
century) brought new forms of territorial management along with an intensification
of watershed use which remained relatively unchanged until the 1900s
39
Figure 2 a) Age-depth model obtained for the Laguna de Matanzas sedimentary
sequence A hiatus occurs at 80 cm depth (unit 3b) The section of core used for our
analysis is highlighted in a red rectangle b) Close up of the age model used for
analysis of recent anthropogenic influences on the N cycle c) Information regarding
the 14C dates used to construct age model
Lab code Sample ID
Depth (cm) Material Fraction of modern C
Radiocarbon age
Pmc Error BP Error
D-AMS 021579
MAT11-6A 104-105 Bulk Sediment
8843 041 988 37
D-AMS 001132
MAT11-6A 1345-1355
Bulk Sediment
8482 024 1268 21
POZ-57285
MAT13-12 DIC Water column 10454 035 Modern
Table 2 Laguna Matanzas radiocarbon dates
32 The sediment sequence
Laguna Matanzas sediments consist of massive to banded mud with some silt
intercalations They are composed of silicate minerals (plagioclase quartz and clay
minerals) with relatively high TOC content (Fig 3) Pyrite is a common mineral
indicating dominant anoxic conditions in the lake sediments whereas aragonite
occurs only in the uppermost section Mineralogical analyses visual descriptions
texture and geochemical composition were used to characterize five main facies
(Fig 3) F1 (organic-rich mud) represents baseline sedimentation in a shallow well-
mixed brackish highly productive lake and F1acute (dark orange) is a less organic facies
than F1 (more details see table in the supplementary material) F2 (massive to
banded silty mud) indicates periods of higher clastic input into the lake but finer
(mostly clay minerals) likely from suspension deposition associated with flooding
40
events Aragonite (up to 15 ) occurs in both facies but only in samples from the
uppermost 30 cm and it is interpreted as endogenic linked to higher MgFe waters
and elevated biologic productivity
Figure 3 Sedimentary facies and units mineralogy grain size elemental geochemical
and isotopic composition of Laguna Matanzas core Values highlighted in gray indicate
that these are above average
The banded to laminated fining upward silty clay layers (F3) reflect
deposition by high energy turbidity currents The presence of aragonite suggests
that littoral sediments were incorporated by these currents Non-graded laminated
coarse silt layers (F4) do not have aragonite indicating a dominant watershed
41
sediment source Both facies are interpreted as more energetic flood deposits but
with different sediment sources A unique breccia layer with coarse silt matrix and
cm-long soft clasts (F5) is interpreted as a high-energy event (ie a tsunami)
capable of eroding the littoral zone and depositing coarse clastic material in the
distal zone of the lake Similar coarse breccia layers have been found at several
coastal sites in Chile and interpreted as tsunami-related deposits (Vargas et al
2005 Le Roux et al 2008)
33 Sedimentary units
Three main units and six subunits have been defined (Fig 3) based on
sedimentary facies and sediment composition We use ZrTi as an indicator of the
mineral fraction transported from the watershed (Marzecoacuteva et al 2011) as higher
ZrTi (F3 and F4) is commonly associated with coarser sediments (Cuven et al
2010) A high correlation among Br BrTi and TOC (r = 046-087 p value=0)
supports the use of BrTi as an indicator of lake productivity (Marzecoacuteva et al 2011
Frugone-Aacutelvarez et al 2017) The MnFe ratio is indicative of lake bottom
oxygenation (Naecher et al 2013) as under reducing conditions Mn mobilizes more
than Fe leading to a decreased MnFe ratio (Marzecoacuteva et al 2011) SrTi indicates
periods of increased aragonite formation as Sr is preferentially included in the
aragonite mineral structure (Veizer et al 1971) (See supplementary material)
The basal unit (3c 129 to 99 cm) is relatively organic-rich (TOC mean=26
BioSi mean=5) and composed by F1 (without aragonite) with some coarser F4
flood layers (ZrTi mean=025) Unit 3b (98 to 80 cm) is interpreted as a tsunami or
storm surge deposit (breccia F5 grading into massive to banded silt F2) Unit 3a
(79 to 73 cm) is characterized by relatively low productivity (TOC=2 BrTi=002
42
BioSi=4) under anoxic conditions (MnFe=001) Unit 2a (72 to 31cm) has
relatively less organic content and more intercalated clastic facies F3 and F4 The
top of this unit (43-30 cm) has elevated TS values The Subunit 1b (30-20 cm)
shows increasing TOC BioSi and BrTi values (TOC mean = 29 BioSi mean =
54 BrTi mean=004) and the upper subunit 1a (19-0 cm) has the highest TOC
(mean = 64) and BioSi (mean = 56) high BrTi mean = 010 and the presence
of aragonite More frequent anoxic conditions (MnFe lower than 001) during units
3 and 2 shifted towards more oxic episodes (MnFe = 003) in unit 1 (Fig 3)
34 Isotopic signatures
Figure 4 shows the isotopic signature from soil samples of the major land
usescover present in the Laguna Matanzas used as an end member in comparison
with the lacustrine sedimentary units δ15N from cropland samples exhibit the
highest values whereas grassland and soil samples from lake shore areas have
intermediate values (Fig 4) Tree plantations and native forests have similarly low
δ15N values (+11 permil SD=24) All samples (except those from the lake shore)
exhibit low δ13C values (from ndash285permil to ndash298permil) CNmolar from agriculture land
lakeshore area and non-vegetation areas samples display the lowest values (about
18) CNmolar from tree plantations and native forest have the highest values (383
and 267 respectively)
43
Figure 4 C-N stable isotope plot showing a comparison of lake sediments grouped
by sedimentary units (MAT11-6A) with the soil end members of present-day (lake
shore and land usecover) from Laguna Matanzas
The δ15N values from sediment samples (MAT11-6A) range from ndash15 and
+53permil (mean= +35 SD=05) δ13C values range from ndash266permil to ndash202permil (mean=
ndash240 SD=14) In unit 3c δ15N and δ13C show relatively high values (mean=
+41permil and ndash233permil respectively) δ15N and δ13C from unit 3b and 3a fluctuate at
slightly lower values than in 3c (mean δ15N from 3a= +38permil and 3b mean= +39permil
mean δ13C from 3a= ndash242permil and 3b mean= ndash244permil) In unit 2a δ15N values are
relatively high (mean= +38permil) but show a slightly decreasing trend (from +52 to
+24permil at 66 cm and 36 cm respectively) δ13C also decreases (mean= ndash242permil)
reaching minimum values at 45 cm (ndash266permil) with a sharp increase towards the top
of this unit with maximum values ca ndash210permil Unit 1 exhibits the lowest δ15N values
(1a mean= +11permil 1b mean= +28permil) with negative values in the uppermost
44
sediments (ndash04permil at 14 cm) δ13C values show a decreasing trend over most of
subunit 1b and increase only near the very top of this unit
35 Recent land use changes in the Laguna Matanzas watershed
Major LUCC from 1975 (unit 1b) to 2016 CE in the Laguna Matanzasacutes
watershed is summarized in Figure 5 The watershed has a surface area of 30 km2
of which native forest (36) and grassland areas (44) represented 80 of the
total surface in 1975 The area occupied by agriculture was only 02 and tree
plantations were absent Isolated burned areas (33) were located mostly in the
northern part of the watershed By 1989 tree plantations surface area had increased
to 5 burned areas to 17 and agricultural fields to 9 (a 45-fold increase) and
native forest and grassland sectors decreased to 23 and 27 respectively By
2016 agricultural land and tree plantations have increased to 17 of the total area
whereas native forests decreased to 21
45
Figure 5 Land uses and cover changes from 1975 to 2016 in Laguna Matanzas
watershed from natural cover and areas for livestock grazing (grassland) to the
expansion of agriculture and forest plantation
4 DISCUSSION
41 N and C dynamics in Laguna Matanzas
Small lakes with relatively large watersheds such as Laguna Matanzas would
be expected to have relatively high contributions of allochthonous C to the sediment
OM (Gu et al 2006) Terrestrial C3 plants (δ13C =ndash26 to ndash28permil Meyers and Terranes
2001 Ku et al 2007) are dominant in the Laguna Matanzas watershed Likewise
our soil samples ranged across similar although slightly more negative values
46
(δ13C = ndash30 to ndash28permil Fig 4) to those proposed by Meyers and Terranes (2001)
and are used here as terrestrial end members oil samples were taken from the lake
shore (δ13C = ndash22permil) and POM from surface water (δ13C = ndash24permil) were more
positive than the terrestrial end member and are used as lacustrine end members
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from terrestrial vegetation and more positive δ13C values have increased
aquatic OM (Gu et al 2006 Torres et al 2012) Phytoplankton preferentially uptake
12C leaving the DIC pool enriched in 13C (Gu et al 2006) especially when there are
no important external sources of C (eg decreased C input from the watershed)
Therefore during events of elevated primary productivity the phytoplankton uptakes
12C until its depletion and are then obligated to use the heavier isotope resulting in
an increase in δ13C Changes in lake productivity thus greatly affect the C isotope
signal (Torres et al 2012) with high productivity leading to elevated δ13C values
(Torres et al 2012 Gu et al 2006)
In a similar fashion the N isotope signatures in Laguna Matanzas reflect a
combination of factors including different N sources (autochthonousallochthonous)
and lake processes such as productivity isotope fractionation in the water column
and sediment denitrification Elevated δ15N values from a POM sample (+22permil) and
average values from the lake shore (mean=+34permil SD=028) are used as aquatic
end members whereas terrestrial samples have values from +10 +24 (tree species)
to +77permil +35 (agriculture) which we use as terrestrial end members (Fig 4)
Autochthonous OM in aquatic ecosystems typically displays low δ15N values
when the OM comes from N-fixing species Atmospheric fixation of N2 by
cyanobacteria ends in OM with δ15N values close to 0permil (Leng et al 2006)
Phytoplankton preferentially uptake 14N from Dissolved Inorganic Nitrogen (DIN) in
47
the water column and derived OM typically have δ15N values lower than DIN values
When productivity increases the remaining DIN becomes depleted in 14N which in
turn increases the δ15N values of phytoplankton over time especially if the N not
replenished (Torres et al 2012) Thus high POM δ15N values from Laguna
Matanzas reflect elevated phytoplankton productivity with a 14N- depleted DIN In
addition N-watershed inputs also contribute to high δ15N values Heavily impacted
watersheds by human activities are often reflected in isotope values due to land use
changes and associated modified N fluxes For example the input of N runoff
derived from the use of inorganic fertilizers leads to the presence of elevated δ15N
(between ndash4 to +4permil) values in the water bodies (Elliott and Brush 2006 Diebel and
Vander Zanden 2009) Widory et al (2004) reported a direct relationship between
elevated δ15N values and increased nitrate concentration from manure in the
groundwater of Brittany (France) Terranes and Bernasconi (2000) found a good
correlation between augmented nutrient loading and a progressive increase in δ15N
values of sedimentary OM related to agricultural land use
Post-depositional diagenetic processes can further affect C and N isotope
signatures Several studies have shown a decrease in δ13C values of OM in anoxic
environments particularly during the first years of burial related to the selective
preservation of OM depleted in 13C (Hollander and Smith 2001 Lehmann et al
2002 Gӓlman et al 2009 Torres et al 2012) Diagenetic processes can also lead
to post-burial N isotope enrichment of the sediments Indeed 14N is consumed more
rapidly than 15N by denitrifying bacteria which intensifies under anoxic conditions
(Diebel and Vander Zanden 2009) Thus the remaining OM pool will be enriched
in 15N leaving to elevated δ15N values (Granger et al 2008 Menzel et al 2013)
48
In summary the relatively high δ15N values in sediments of Laguna Matanzas
reflect N input from an agriculturegrassland watershed with positive synergetic
effects from increased lake productivity enrichment of DIN in the water column and
most likely denitrification The increase of algal productivity associated with
increased N terrestrial input andor recycling of lake nutrients (and lesser extent
fixing atmospheric N) and denitrification under anoxic conditions can all increase
δ15N values (Fig 3) In addition elevated lake productivity without C replenishing
(eg by terrestrial C input) produces shifts towards positive δ13C values whereas C
input from the watershed generates more negative δ13C values
42 Recent evolution of the Laguna Matanzas watershed
Sedimentological compositional and geochemical indicators show three
depositional phases in the lake evolution under the human influence in the Laguna
Matanzas over the last two hundred years Although the record is longer (around
1000 years) we analyzed the last two centuries (unit 2 and 1) to provide a recent
historical context for the large changes detected during the 20th century
The first phase lasted from the beginning of the 19th century until ca 1940
(unit 2a Fig 3 4) and was characterized by moderate productivity with elevated
sediment input from the watershed as indicated by our geochemical proxies (BrTi
= 004 AlTi gt 01 Fig 6) Higher lake levels and dominant anoxic bottom conditions
(MnFe = 001 Fig3) can be related to increased precipitation (mean= 557 mmyr)
and lower temperatures (summer annual temperature lt19ordmC) During the Spanish
colonial period the Laguna Matanzas watershed was used as a livestock farm
(Jesuits 1627-1780 unit 3 followed by the Pedro Balmaceda era 1780-1810- unit
2a) (Contreras-Loacutepez et al 2014) The main buildings and farm facilities at the El
49
Convento village During this period livestock grazing and lumber extraction for
mining would have involved extensive deforestation and loss of native vegetation
(eg Armesto et al 1994 2010) However the Matanzas pollen record does not
show any significant regional deforestation during this period (Villa Martiacutenez 2002)
suggesting that the impact may have been highly localized
Lake productivity sediment input and elevated precipitation (Fig 6) all
suggest that N availability was related to this increased input from the watershed
The N from cow manure and soil particles would have led to higher δ15N values
(Elliot and Brush 2016) In addition anoxic lake bottom conditions would lead to
even further enrichment of buried sediment N The δ13C values lend further support
to our interpretation of increased sediment input -and N- from the watershed
Decreased δ13C values reach their lowest values in the entire record (ndash270permil) at
ca 1910 CE (Fig 4 6)
During most of the 19th century human activities in Laguna Matanzas were
similar to those during the Spanish Colonial period However the appearance of
Pinus radiata and Eucalyptus globulus pollen ca 1890 CE (Gibson 1972) the dune
stabilization-afforestation program which began ca 1900 CE (Albert 1900) and the
application of the Chilean forestry law of 1931 (DFL nordm265) contributed to an
increased capacity of the surrounding vegetation to retain nutrients and sediments
The law subsidized forest plantations in areas devoid of vegetation and prohibited
the cutting of forest on slopes greater than 45ordm These land use changes were coeval
with decreased sediment inputs (AlTi trend) from the watershed slightly increased
lake productivity (BrTi from 001 to 003) and a decrease in annual precipitation
(Fig 6) N isotope values become more negative during this period although they
remained high (from +49permil to +37permil) whereas the δ13C trend towards more
50
positive values reflects changes in the N source from watershed to in-lake dynamics
(e g increased endogenic productivity)
The second phase started after 1940 and is clearly marked by an abrupt
change in the general trend of δ15N that oscillates around +41permil (SD = 04permil) during
the first phase to +24permil (SD = 14permil) during the second These δ15N trends reflect
the lowest watershed nutrient and sediment inputs (based on the AlTi record)
decreased precipitation (mean = 318 mm year) and a slight increase in lake
productivity (increased BrTI) Depositional dynamics in the lake likely crossed a
threshold as human activity intensified throughout the watershed and lake levels
decreased
During the Great Acceleration δ15N values shifted towards higher values to
ca 3permil with an increase in δ13C values that are not reflected either in lake
productivity or lake level As the sediment input from the watershed increased and
precipitation remained as low as the previous decade δ15N values during this period
are likely related to watershed clearance which would have increased both nutrient
and sediment input into the lake
The δ13C trend to more positive values reaching the peaks in the 1960s (ndash
212permil) at the same time as the peaks in temperature (196 ordmC mean annual) with a
downward trend in precipitation A shift in OM origin from macrophytes and
watershed input influences to increased lake productivity could explain this trend
(Fig 4 1b)
In the 1970s the Laguna Matanzasacute watershed was mostly covered by native
forest (36) and grassland areas were intended for livestock grazing (44 Fig 5)
Both soils have δ15NSoil lt 5permil (Fig 4) Farming fields occupied a very small area and
tree plantations were almost nonexistent The decreasing trend in δ15N values seen
51
in our record is interrupted by several large peaks that occurred between ca 1975
and ca 1989 when the native forest and grassland areas fell by 23 and 27
respectively largely due to fires affecting 17 of the forests Agriculture fields
increased by 4 and forest plantations increased by 9 (unit 1a) Concomitantly
sediment ndash and likely N - inputs from the watershed decreased (as indicated by the
trend in AlTi) the precipitation was still relatively low (Fig 6) These changes are
likely related to the increase of vegetation cover especially of tree plantations (which
have more negative δ15N values) The small increase in productivity in the lake could
have been favored by increased temperature (von Gunten et al 2009) After 1989
the increase in agriculture land (17 in 2016) correlates with increasing δ15N δ13C
and TOC trends in spite of declining rainfall The increase of forest plantations was
mostly in response to the implementation of the Law Decree of Forestry
Development (DL 701 of 1974) that subsidized forest plantation After 1989 the
increase in agricultural land (17 in 2016) is synchronous with increasing δ15N
δ13C TOC and MnFe trends despite the decline in rainfall and overall lower lake
levels as more water is used for irrigation
The third phase started c 1990 CE (unit 1a) when OM accumulation rates
increase and δ13C δ15N decreased reaching their lowest values in the sequence
around 2000 CE Afterward during the 21st century δ13C and δ15N values again
began to increase The onset of unit 1 is marked by increased lake productivity and
decreased sediment input (AlTi lt 02) synchronous with intensive farming replacing
forestry and extensive agriculture (Fig 5 6)
A change in the general trend of δ15N values which decreased until 1990
(unit 1a) as the native forest and grassland area fell to 23 and 27 respectively
is most likely due to deforestation and fires Agriculture surface increased to 4 and
52
forest plantations increased by 9 (Fig 5) Concomitantly sediment ndash and likely N
ndash inputs from the watershed decreased probably related to the low precipitation (Fig
1b) and the increase of vegetation cover in the watershed in particularly by tree
plantations (with more negative δ15N Fig 4)
At present agriculture and tree plantations occupy around 34 of the
watershed surface whereas native forests and grassland cover 21 and 25
respectively Lake productivity as indicated by BrTi (Fig6) is higher and generates
OM with lower δ15N and higher δ13C values (about +2permil and ndash20permil ca 2000 CE
respectively) due to in-lake processes (ie biological N fixation and nutrient
recycling) and driven by changes in the arboreal cover which diminishes nutrient
flux into the lake (Fig6)
53
Figure 6 Anthropogenic and climatic forcing and lake dynamics response
(productivity sediment input N and C cycles) at Matanzas Lake over the last two
54
centuries Mean annual precipitation reconstructed and temperatures (von Gunten
et al 2009) Vertical gray bars indicate mega-droughts
5 CONCLUSIONS
Human activities have been the main factor controlling the N and C cycle in
the Laguna Matanzas during the last two centuries The N isotope signature in the
lake sediments reflects changes in the watershed fluxes to the lake but also in-lake
processes such as productivity and post-depositional changes Denitrification could
have been a dominant process during periods of increased anoxic conditions which
were more frequent prior to Great Acceleration (1950 CE) Higher δ15N and lower
δ13C values are associated with increased nutrient input from the watershed due to
increased livestock grazing and agriculture pressure (phase 1 Fig 7a) whereas
lower isotope values occurred during periods of increased forest plantations (phase
3 Fig 7c) During periods of increased lake productivity - such as in the last few
decades - δ15N values increased significantly
The most important change in C and N dynamics in the lake occurred after the
1950s (Phase 2 and 3) as tree plantations dominated the watershed Recent
changes in N dynamics can be explained by the higher nutrient contribution
associated with intensive agriculture (i e fertilizers) since the 1990s Although the
replacement of livestock activities with forestry and farming seems to have reduced
nutrient and soil export from the watershed to the lake the inefficient use of fertilizer
(by agriculture) can be the ultimate responsible for lake productivity increase during
the last decades
55
Figure 7 Schematic diagrams illustrating the main factors controlling the
isotope N signal in sediment OM of Laguna Matanzas N input from watershed
depends on human activities and land cover type Agriculture practices and cattle
(grassland development) contribute more N to the lake than native forest and
plantations Periods of higher productivity tend to deplete the dissolved inorganic N
in 14N resulting in higher δ15N (OM) The denitrification processes are more effective
in anoxic conditions associated with higher lake levels
6 METHODS
Short sediment cores were recovered from Laguna Matanzas using an Uwitec
gravity piston corer in 2011 and 2013 (Fig 1a) Sediment cores (MAT11-6A 129 cm
MAT13-2A 149 cm MAT13-3A 33 cm MAT13-4A 99 cm) were split
photographed sub-sampled and stored at the Pyrenean Institute of Ecology (IPE-
CSIC Spain) Core MAT11-6A was obtained from the central sector of the lake and
56
was selected for detailed multiproxy analyses (including elemental geochemistry C
and N isotope analyses XRF and 14C dating)
The isotope analyses (δ13C and δ15N) were performed at the Laboratory of
Biogeochemistry and Applied Stable Isotopes (LABASI-PUC Chile) using a Delta
V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via a
Conflo IV interface Isotope results are expressed in standard delta notation (δ) in
per mil (permil) relative to the standards Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Sediment samples
for δ13Corg were pre-treated with 50 ml of HCl and 50 ml of deionized water and
dried at 60ordmC for 4 hr to remove carbonates (Harris et al 2011)
Total Carbon (TC) Total Organic Carbon (TOC) Total Inorganic Carbon (TIC)
and Total Sulphur (TS) were measured every cm with a LECO SC 144 DR at IPE-
CSIC XRF measurements were carried out every 4 mm in MAT11-1A-1G core using
an AVAATECH X-ray Fluorescence II core scanner at the University of Barcelona
(Spain) Results are expressed as element intensities in counts per second (cps)
Tube voltage was operated at 30 kV and 10 kV to obtain the abundances of 15
elements (Al Si S Cl K Ca Ti V Mn Fe Br Rb Sr Y Zr) with an average at
least of 1600 cps (less for Br=1000)
Biogenic silica content mineralogy and grain size were measured every 4
cm Biogenic silica was measured following Mortlock and Froelich (1989) and
Bernaacuterdez et al (2005) using an Auto Analyzer Technicon AAII for dissolved silicate
analysis Mineralogy was analyzed with a Siemens D-500 X-ray diffractometer (Cu
kα 40 kV 30 mA graphite monochromator) at the ICTJA-CSIC (Spain) Grain size
analyses were performed in a Beckmann Coulter LS 13 320 Particle Size Analyzer
57
at the IPE-CSIC The samples were classified according to textural classes as
follows clay (lt2 μm) silt (20-2 microm) and sand (gt2 microm) fractions
The age-depth model for the Laguna Matanzas sedimentary sequence was
constructed using 210Pb and 137Cs dating techniques (MAT13-4A) as well as two 14C
AMS dates on bulk sediment samples (MAT11-6A fig 2) We dated the dissolved
inorganic carbon (DIC) in the water column and no significant reservoir effect is
present in the modern-day water column (10454 + 035 pcmc Table 2) An age-
depth model was obtained with the Bacon R package to estimate the deposition
rates and associated age uncertainties along the core (Blaauw and Christen 2011)
To estimate LUCC of Laguna Matanzasacutes watershed we use satellite images
Landsat MSS for 1975 Landsat TM for 1989 and Landsat OLI for 2016 all taken in
summer or autumn (Table 1) We performed supervised classification of land uses
(maximum likelihood algorithm) for each year (1975 1989 and 2016) and the results
were mapped using software ArcGIS 102 in 2017
Satellite Images Acquisition Date
Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat OLI 20160404 30 m
Table 1 Landsat imagery
Surface water samples were filtered for obtained particulate organic matter In
addition soil samples from the main land usecover present in the Laguna Matanzas
watershed were collected Elemental C N and their corresponding isotopes from
POM and soil were obtained at the LABASI and used here as end members
Daily precipitation at Santo Domingo (33ordm39`S 71ordm3 6`W)- the nearest weather
station to Laguna Matanzasndash was compiled using the redPrec R package (Fig1 d
Serrano-Notivoli et al 2017 Sarricolea et al 2019) To extend our precipitation
58
reconstruction back to 1824 we correlated this dataset with that available for
Santiago The Santiago data was compiled from data published in the Anales of
Universidad of Chile (1850) for the years 1824 to 1849 Almeyda (1940) for the years
1849 to 1864 and the Quinta Normal series from 1866 to the present (Direccioacuten
Meteoroloacutegica de Chile) We generated a linear regression model between the
presentday Santo Domingo station and the compiled Santiago data with a Pearson
coefficient of 087 and p-valuelt 001
Acknowledgments This research was funded by grants CONICYT AFB170008
to the Institute of Ecology and Biodiversity (IEB) and FONDECYT 1150763 (to CL)
Doctoral grant Becas Chile 21150224 MEDLANT (Spanish Ministry of Economy
and Competitiveness grant CGL2016-76215-R) Additional funding was provided
by the Laboratorio Internacional de Cambio Global (LINCGlobal PUC-CSIC) We
thank R Lopez E Royo and M Gallegos for help with sample analyses We thank
the Laboratory of Biogeochemistry and Applied Stable Isotopes (LABASI) of the
Department of Ecology (PUC) for sample analyses
References
Albert F 1900 Las dunas o sean arenas volantes voladeros arenas muertas invasion de las arenas playas I meacutedanos del centro de Chile comprendiendo el litoral desde el liacutemite norte de la provincia de Aconcagua hasta el liacutemite sur de la de Arauco Memorias cientiacuteficas I Lit
Anderson NJ W D Fritz SC 2009 Holocene carbon burial by lakes in SW
Greenland Glob Chang Biol httpsdoiorg101111j1365-2486200901942x
Armesto J Villagraacuten C Donoso C 1994 La historia del bosque templado
chileno Ambient y Desarro 66 66ndash72 httpsdoiorg101007BF00385244
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A
59
historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Arnell NW Gosling SN 2013 The impacts of climate change on river flow
regimes at the global scale J Hydrol 486 351ndash364 httpsdoiorg101016JJHYDROL201302010
Bernaacuterdez P Prego R Franceacutes G Gonzaacutelez-Aacutelvarez R 2005 Opal content in
the Riacutea de Vigo and Galician continental shelf Biogenic silica in the muddy fraction as an accurate paleoproductivity proxy Cont Shelf Res httpsdoiorg101016jcsr200412009
Blaauw M Christen JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474 httpsdoiorg10121411-BA618
Brush GS 2009 Historical land use nitrogen and coastal eutrophication A
paleoecological perspective Estuaries and Coasts 32 18ndash28 httpsdoiorg101007s12237-008-9106-z
Camarero L Catalan J 2012 Atmospheric phosphorus deposition may cause
lakes to revert from phosphorus limitation back to nitrogen limitation Nat Commun httpsdoiorg101038ncomms2125
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego
R Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and environmental change from a high Andean lake Laguna del Maule central Chile (36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Catalan J Pla-Rabeacutes S Wolfe AP Smol JP Ruumlhland KM Anderson NJ
Kopaacuteček J Stuchliacutek E Schmidt R Koinig KA Camarero L Flower RJ Heiri O Kamenik C Korhola A Leavitt PR Psenner R Renberg I 2013 Global change revealed by palaeolimnological records from remote lakes A review J Paleolimnol httpsdoiorg101007s10933-013-9681-2
Chen Y Y Huang W Wang W H Juang J Y Hong J S Kato T amp
Luyssaert S (2019) Reconstructing Taiwanrsquos land cover changes between 1904 and 2015 from historical maps and satellite images Scientific Reports 9(1) 3643 httpsdoiorg101038s41598-019-40063-1
Contreras S Werne J P Araneda A Urrutia R amp Conejero C A (2018)
Organic matter geochemical signatures (TOC TN CN ratio δ 13 C and δ 15 N) of surface sediment from lakes distributed along a climatological gradient on the western side of the southern Andes Science of The Total Environment 630 878-888 httpsdoiorg101016jscitotenv201802225
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia UC
60
de V 2014 ELEMENTOS DE LA HISTORIA NATURAL DEL An Mus Hist Natulas Vaplaraiso 27 51ndash67
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-010-9453-1
Diebel M Vander Zanden MJ 201209 Nitrogen stable isotopes in streams
stable isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19 1127ndash1134 httpsdoiorg10189008-03271
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506 Espizua LE Pitte P 2009 The Little Ice Age glacier advance in the Central
Andes (35degS) Argentina Palaeogeogr Palaeoclimatol Palaeoecol 281 345ndash350 httpsdoiorg101016JPALAEO200810032
Figueroa JA Castro S A Marquet P A Jaksic FM 2004 Exotic plant
invasions to the mediterranean region of Chile causes history and impacts Invasioacuten de plantas exoacuteticas en la regioacuten mediterraacutenea de Chile causas historia e impactos Rev Chil Hist Nat 465ndash483 httpsdoiorg104067S0716-078X2004000300006
Fowler D Coyle M Skiba U Sutton MA Cape JN Reis S Sheppard
LJ Jenkins A Grizzetti B Galloway N Vitousek P Leach A Bouwman AF Butterbach-bahl K Dentener F Stevenson D Amann M Voss M 2013 The global nitrogen cycle in the twenty- first century Philisophical Trans R Soc B httpsdoiorghttpdxdoiorg101098rstb20130164
Frank D Reichstein M Bahn M Thonicke K Frank D Mahecha MD
Smith P van der Velde M Vicca S Babst F Beer C Buchmann N Canadell JG Ciais P Cramer W Ibrom A Miglietta F Poulter B Rammig A Seneviratne SI Walz A Wattenbach M Zavala MA Zscheischler J 2015 Effects of climate extremes on the terrestrial carbon cycle Concepts processes and potential future impacts Glob Chang Biol httpsdoiorg101111gcb12916
Fritz SC Anderson NJ 2013 The relative influences of climate and catchment
processes on Holocene lake development in glaciated regions J Paleolimnol httpsdoiorg101007s10933-013-9684-z
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
61
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS) implications for past sea level and environmental variability J Quat Sci 32 830ndash844 httpsdoiorg101002jqs2936
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892 httpsdoiorg101126science1136674
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924 httpsdoiorg104319lo20095430917
Garciacutea-Ruiz JM 2010 The effects of land uses on soil erosion in Spain A
review Catena httpsdoiorg101016jcatena201001001
Garciacutea-Ruiz JM Loacutepez-Moreno JI Lasanta T Vicente-Serrano SM
Gonzaacutelez-Sampeacuteriz P Valero-Garceacutes BL Sanjuaacuten Y Begueriacutea S Nadal-Romero E Lana-Renault N Goacutemez-Villar A 2015 Los efectos geoecoloacutegicos del cambio global en el Pirineo Central espantildeol una revisioacuten a distintas escalas espaciales y temporales Pirineos httpsdoiorg103989Pirineos2015170005
Garciacutea-Ruiz JM Nadal-Romero E Lana-Renault N Begueriacutea S 2013
Erosion in Mediterranean landscapes Changes and future challenges Geomorphology 198 20ndash36 httpsdoiorg101016jgeomorph201305023
Garreaud RD Vuille M Compagnucci R Marengo J 2009 Present-day
South American climate Palaeogeogr Palaeoclimatol Palaeoecol httpsdoiorg101016jpalaeo200710032
Gayo EM Latorre C Jordan TE Nester PL Estay SA Ojeda KF
Santoro CM 2012 Late Quaternary hydrological and ecological changes in the hyperarid core of the northern Atacama Desert (~21degS) Earth-Science Rev httpsdoiorg101016jearscirev201204003
Ge Y Zhang K amp Yang X (2019) A 110-year pollen record of land use and land
cover changes in an anthropogenic watershed landscape eastern China Understanding past human-environment interactions Science of The Total Environment 650 2906-2918 httpsdoiorg101016jscitotenv201810058
Goyette J Bennett EM Howarth RW Maranger R 2016 Global
Biogeochemical Cycles 1000ndash1014 httpsdoiorg1010022016GB005384Received
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and
oxygen isotope fractionation during dissimilatory nitrate reduction by
62
denitrifying bacteria 53 2533ndash2545 httpsdoiorg10230740058342
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
Gurnell AM Petts GE Hannah DM Smith BPG Edwards PJ Kollmann
J Ward J V Tockner K 2001 Effects of historical land use on sediment yield from a lacustrine watershed in central Chile Earth Surf Process Landforms 26 63ndash76 httpsdoiorg1010021096-9837(200101)261lt63AID-ESP157gt30CO2-J
Heathcote A J et al Large increases in carbon burial in northern lakes during the
Anthropocene Nat Commun 610016 doi 101038ncomms10016 (2015) Hollander DJ Smith MA 2001 Microbially mediated carbon cycling as a
control on the δ13C of sedimentary carbon in eutrophic Lake Mendota (USA) New models for interpreting isotopic excursions in the sedimentary record Geochim Cosmochim Acta httpsdoiorg101016S0016-7037(00)00506-8
Holtgrieve GW Schindler DE Hobbs WO Leavitt PR Ward EJ Bunting
L Chen G Finney BP Gregory-Eaves I Holmgren S Lisac MJ Lisi PJ Nydick K Rogers LA Saros JE Selbie DT Shapley MD Walsh PB Wolfe AP 2011 A coherent signature of anthropogenic nitrogen deposition to remote watersheds of the Northern Hemisphere Science (80) 334 1545ndash1548 httpsdoiorg101126science1212267
Hooper DU Adair EC Cardinale BJ Byrnes JEK Hungate BA Matulich
KL Gonzalez A Duffy JE Gamfeldt L Connor MI 2012 A global synthesis reveals biodiversity loss as a major driver of ecosystem change Nature httpsdoiorg101038nature11118
Horvatinčić N Sironić A Barešić J Sondi I Krajcar Bronić I Borković D
2016 Mineralogical organic and isotopic composition as palaeoenvironmental records in the lake sediments of two lakes the Plitvice Lakes Croatia Quat Int httpsdoiorg101016jquaint201701022
Howarth RW 2004 Human acceleration of the nitrogen cycle Drivers
consequences and steps toward solutions Water Sci Technol httpsdoiorg1010382Fscientificamerican0490-56
Jenny B Valero-Garceacutes BL Urrutia R Kelts K Veit H Appleby PG Geyh
M 2002 Moisture changes and fluctuations of the Westerlies in Mediterranean Central Chile during the last 2000 years The Laguna Aculeo record (33deg50primeS) Quat Int 87 3ndash18 httpsdoiorg101016S1040-
63
6182(01)00058-1
Jenny B Wilhelm D Valero-Garceacutes BL 2003 The Southern Westerlies in
Central Chile Holocene precipitation estimates based on a water balance model for Laguna Aculeo (33deg50primeS) Clim Dyn 20 269ndash280 httpsdoiorg101007s00382-002-0267-3
Leng M J Lamb A L Heaton T H Marshall J D Wolfe B B Jones M D
amp Arrowsmith C 2006 Isotopes in lake sediments In Isotopes in palaeoenvironmental research (pp 147-184) Springer Dordrecht
Le Roux JP Nielsen SN Kemnitz H Henriquez Aacute 2008 A Pliocene mega-
tsunami deposit and associated features in the Ranquil Formation southern Chile Sediment Geol 203 164ndash180 httpsdoiorg101016jsedgeo200712002
Lehmann M Bernasconi S Barbieri a McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66 3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Lenihan JM Drapek R Bachelet D Neilson RP 2003 Climate change
effects on vegetation distribution carbon and fire in California Ecol Appl httpsdoiorg101890025295
McLauchlan K K Gerhart L M Battles J J Craine J M Elmore A J
Higuera P E amp Perakis S S (2017) Centennial-scale reductions in nitrogen availability in temperate forests of the United States Scientific reports 7(1) 7856 httpsdoi101038s41598-017-08170-z
Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32ordmS 3050 masl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
Martiacuten-Foreacutes I Casado MA Castro I Ovalle C Pozo A Del Acosta-Gallo
B Saacutenchez-Jardoacuten L Miguel JM De 2012 Flora of the mediterranean basin in the chilean ESPINALES Evidence of colonisation Pastos 42 137ndash160
Marzecovaacute A Mikomaumlgi a Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54 httpsdoiorg103176eco2011105
Matesanz S Valladares F 2014 Ecological and evolutionary responses of
Mediterranean plants to global change Environ Exp Bot httpsdoiorg101016jenvexpbot201309004
64
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Menzel P Gaye B Wiesner MG Prasad S Stebich M Das BK Anoop A
Riedel N Basavaiah N 2013 Influence of bottom water anoxia on nitrogen isotopic ratios and amino acid contributions of recent sediments from small eutrophic Lonar Lake central India Limnol Oceanogr 58 1061ndash1074 httpsdoiorg104319lo20135831061
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Michelutti N Wolfe AP Cooke CA Hobbs WO Vuille M Smol JP 2015
Climate change forces new ecological states in tropical Andean lakes PLoS One httpsdoiorg101371journalpone0115338
Muntildeoz-Rojas M De la Rosa D Zavala LM Jordaacuten A Anaya-Romero M
2011 Changes in land cover and vegetation carbon stocks in Andalusia Southern Spain (1956-2007) Sci Total Environ httpsdoiorg101016jscitotenv201104009
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Poraj-Goacuterska A I Żarczyński M J Ahrens A Enters D Weisbrodt D amp
Tylmann W (2017) Impact of historical land use changes on lacustrine sedimentation recorded in varved sediments of Lake Jaczno northeastern Poland Catena 153 182-193 httpdxdoiorg101016jcatena201702007
Pongratz J Reick CH Raddatz T Claussen M 2010 Biogeophysical versus
biogeochemical climate response to historical anthropogenic land cover change Geophys Res Lett 37 1ndash5 httpsdoiorg1010292010GL043010
Prudhomme C Giuntoli I Robinson EL Clark DB Arnell NW Dankers R
Fekete BM Franssen W Gerten D Gosling SN Hagemann S Hannah DM Kim H Masaki Y Satoh Y Stacke T Wada Y Wisser D 2014 Hydrological droughts in the 21st century hotspots and uncertainties from a global multimodel ensemble experiment Proc Natl Acad Sci httpsdoiorg101073pnas1222473110
65
Ruumlhland KM Paterson AM Smol JP 2015 Lake diatom responses to
warming reviewing the evidence J Paleolimnol httpsdoiorg101007s10933-
015-9837-3
Sarricolea P Meseguer-Ruiz Oacute Serrano-Notivoli R Soto M V amp Martin-Vide
J (2019) Trends of daily precipitation concentration in Central-Southern Chile Atmospheric research 215 85-98 httpsdoiorg101016jatmosres201809005
Sernageomin 2003 Mapa geologico de chile version digital Publ Geol Digit Serrano-Notivoli R de Luis M Begueriumlia S 2017 An R package for daily
precipitation climate series reconstruction Environ Model Softw httpsdoiorg101016jenvsoft201611005
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Stine S 1994 Extreme and persistent drought in California and Patagonia during
mediaeval time Nature 369 546ndash549 httpsdoiorg101038369546a0
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL
Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Syvitski JPM Voumlroumlsmarty CJ Kettner AJ Green P 2005 Impact of humans
on the flux of terrestrial sediment to the global coastal ocean Science 308(5720) 376-380 httpsdoiorg101126science1109454
Thomas RQ Zaehle S Templer PH Goodale CL 2013 Global patterns of
nitrogen limitation Confronting two global biogeochemical models with observations Glob Chang Biol httpsdoiorg101111gcb12281
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Urbina Carrasco MX Gorigoitiacutea Abbott N Cisternas Vega M 2016 Aportes a
la historia siacutesmica de Chile el caso del gran terremoto de 1730 Anuario de Estudios Americanos httpsdoiorg103989aeamer2016211
Vargas G Ortlieb L Chapron E Valdes J Marquardt C 2005 Paleoseismic
inferences from a high-resolution marine sedimentary record in northern Chile
66
(23degS) Tectonophysics httpsdoiorg101016jtecto200412031 399(1-4) 381-398httpsdoiorg101016jtecto200412031
Valero-Garceacutes BL Latorre C Morelliacuten M Corella P Maldonado A Frugone M Moreno A 2010 Recent Climate variability and human impact from lacustrine cores in central Chile 2nd International LOTRED-South America Symposium Reconstructing Climate Variations in South America and the Antarctic Peninsula over the last 2000 years Valdivia p 76 (httpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-yearshttpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-years
Vicente-Serrano SM Gouveia C Camarero JJ Begueria S Trigo R
Lopez-Moreno JI Azorin-Molina C Pasho E Lorenzo-Lacruz J Revuelto J Moran-Tejeda E Sanchez-Lorenzo A 2013 Response of vegetation to drought time-scales across global land biomes Proc Natl Acad Sci httpsdoiorg101073pnas1207068110
Villa Martiacutenez R P 2002 Historia del clima y la vegetacioacuten de Chile Central
durante el Holoceno Una reconstruccioacuten basada en el anaacutelisis de polen sedimentos microalgas y carboacuten PhD
Villa-Martiacutenez R Villagraacuten C Jenny B 2003 Pollen evidence for late -
Holocene climatic variability at laguna de Aculeo Central Chile (lat 34o S) The Holocene 3 361ndash367
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Human alteration of the global nitrogen cycle sources and consequences Ecological applications 7(3) 737-750 httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Rein B Urrutia R amp Appleby P (2009) A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 The Holocene 19(6) 873-881 httpsdoiorg1011770959683609336573
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Widory D Kloppmann W Chery L Bonnin J Rochdi H Guinamant JL
2004 Nitrate in groundwater An isotopic multi-tracer approach J Contam Hydrol 72 165ndash188 httpsdoiorg101016jjconhyd200310010
67
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
68
Supplementary material
Facie Name Description Depositional Environment
F1 Organic-rich
mud
Massive to banded black
organic - rich (TOC up to 14 )
mud with aragonite in dm - thick
layers Slightly banded intervals
contain less OM (TOClt4) and
aragonite than massive
intervals High MnFe (oxic
bottom conditions) High CaTi
BrTi and BioSi (up to 5)
Distal low energy environment
high productivity well oxygenated
and brackish waters and relative
low lake level
F2 Massive to
banded silty clay
to fine silt
cm-thick layers mostly
composed by silicates
(plagioclase quartz cristobalite
up to 65 TOC mean=23)
Some layers have relatively high
pyrite content (up to 25) No
carbonates CaTi BrTi and
BioSi (mean=48) are lower
than F1 higher ZrTi (coarser
grain size)
Deposition during periods of
higher sediment input from the
watershed
69
F3 Banded to
laminated light
brown silty clay
cm-thick layers mostly
composed of clay minerals
quartz and plagioclase (up to
42) low organic matter
(TOC mean=13) low pyrite
and BioSi content
(mean=46) and some
aragonite
Flooding events reworking
coastal deposits
F4 Laminated
coarse silts
Thin massive layers (lt2mm)
dominated by silicates Low
TOC (mean=214 ) BrTi
(mean=002) MnFe (lt02)
TIC (lt034) BioSi
(mean=46) and TS values
(lt064) and high ZrTi
Rapid flooding events
transporting material mostly
from within the watershed
F5 Breccia with
coarse silt
matrix
A 17 cm thick (80-97 cm
depth) layer composed by
irregular mm to cm-long ldquosoft-
clastsrdquo of silty sediment
fragments in a coarse silt
matrix Low CaTi BrTi and
MnFe ratios and BioSi
Rapid high energy flood
events
70
(mean=43) and high ZrTi
(gt018)
Table Sedimentological and compositional characteristics of Laguna Matanzas
facies
71
CAPIacuteTULO 2 STABLE ISOTOPES TRACK LAND USE AND COVER
CHANGES IN A MEDITERRANEAN LAKE IN CENTRAL CHILE OVER THE
LAST 600 YEARS
72
Stable isotopes track land use and cover changes in a mediterranean lake in
central Chile over the last 600 years
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugone-Aacutelvarezabc Pablo
Sarricolead Leonardo Villaciacutese M Laura Carrevedob Blas Valero-Garceacutescg
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Departamento de Ecologiacutea Universidad de ChileLas Palmeras 3425 Nuntildeoa Santiago Chile
f Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding authors E-mail addresses clatorrebiopuccl magdalenafuentealbagmailcom
Key words Anthropocene lake sediment δsup1⁵N δsup1sup3C Great Acceleration organic
geochemistry watershedndashlake system Stable Isotope Analyses land usecover
change Nitrogen cycle mediterranean ecosystems central Chile
73
Abstract
Nutrient fluxes in many aquatic ecosystems are currently being overridden by
anthropic controls especially since the industrial revolution (mid-1800s) and the
Great Acceleration (mid-1900s) Land use and cover changes (LUCC) alter the
availability and fluxes of nutrients such as nitrogen that are transferred via runoff
and groundwater into lakes By altering lake productivity and trophic status these
changes are often preserved in the sedimentary record Here we use
geolimnological proxies and stable isotope analyses (δsup1⁵N δsup13C) on lake sediments
to reconstruct the lake-watershed nutrient transfer in a central Chilean lake (Lago
Vichuqueacuten) over the past 600 years We compare our multiproxy record from recent
lake sediments to the soilvegetation relationship across the watershed as well as
land usecover changes from 1975 to 2014 derived from satellite images Our results
show that lake sediment δsup1⁵N values increased with meadow cover but decreased
with tree plantations suggesting increased nitrogen retention when trees dominate
the watershed Our δsup1⁵N record from central Chile can thus be interpreted as a proxy
for nutrient availability over the last 600 years mainly derived from land use changes
coupled with climate drivers Although variable sources of organic matter and in situ
fractionation often hinder straightforward environmental interpretations of stable N
isotope signatures our study shows that lake sediment δsup1⁵N can be a useful tool for
assessing the contribution of past human activities in nutrient and nitrogen cycling
1 Introduction
Nitrogen (N) is a limiting nutrient in terrestrial and aquatic ecosystems (Vitousek
et al 1997) Changes in its availability can drive eutrophication and increase
pollution in these ecosystems (McLauchlan et al 2013) Although recent human
74
impacts on the global N cycle have been significant the consequences of increased
anthropogenic N on these ecosystems are unclear (Gerhart and Mclauchlan 2014
Battye et al 2017) Isotopic signatures are key to understand N dynamics in lakes
nevertheless in situ andor diagenetic fractionation along with multiple sources of
organic matter (OM) often hinder straightforward environmental interpretations from
isotopes Monitoring δ15N and δ13C values as components of the N cycle
specifically those related to the link between terrestrial and aquatic ecosystems can
help differentiate between effects from processes versus sources in stable isotope
values (eg from Particulate Organic Matter -POM- soil and vegetation) and
improve how we interpret variations in δ15N (and δ13C) values at longer temporal
scales
The main processes controlling stable N isotopes in bulk lake OM are source
lake productivity diagenesis and denitrification (Talbot 2001 Leng et al 2006
Fariacuteas et al 2009 Torres et al 2012 Brahney et al 2014) N sources depend on
contributions from the watershed (ie soil and biomass) the transfer of atmospheric
N to the lake (ie atmospheric N2 fixation) and nutrient recycling (Leng et al 2006)
Atmospheric N2 (isotopically defined as δ15N =0) is fixed by cyanobacteria with
minimal fractionation resulting in δ15NOM values that fluctuate from -2 to +2permil (Fogel
and Cifuentes 1993 Talbot 2001 Elliot and Brush 2006) Besides N2 fixation by
cyanobacteria phytoplankton species assimilate dissolved inorganic nitrogen (DIN)
and values typically range from +7 to +10permil (Xu et al 2016 Meyers et al 2003) In
addition seasonal changes in POM occur in the lake water column Gu et al (2006)
sampled the water column of Lake Wauberg (Florida USA) during a hydrologic year
and found a higher development of N fixing species during the summer A major
factor behind this increase are human activities in the watershed which control the
75
inputs of the sediment and nutrients to the lake (Woodward et al 2011) Some
studies have shown higher δ15N values in lake sediments from watersheds that are
highly affected by human activities (Bruland and Mackenzie 2010 Woodward et al
2011) Syntenic fertilizers displayed δ15N values between -4 to +4permil cattle manure
around +8 to +22permil and surface and groundwater OM between 0 to +10permil (Elliott
and Brush 2006 Leng et al 2006) Although relatively low δ15N values from
fertilizers constitute major N input to human-altered watersheds the elevated loss
of 14N via volatilization of ammonia and denitrification leaves the remaining total N
input enriched in 15N (Bruland and Mackenzie 2010)
In addition to the different sources and variations in lake productivity early
diagenesis at the sedimentndashwater interface in the sediment can further alter
sediment OM isotope values (Brahney et al 2014 Galman et al 2009) During
diagenetic loss of N in lacustrine systems kinetic fractionation occurs and the
remaining N is enriched in 15N leading to higher δ15N values (Galman et al 2006
Talbot 2001) Denitrification tends to enrich lake sediments in 15N due to the
assimilation of nitrates by denitrifying bacteria (Granger et al 2008) and is more
prevalent in anoxic environments (Brahney et al 2016 Lehman et al 2002)
Carbon isotopes in lake sediments can also provide useful information about
paleoenvironmental changes OM origin and depositional processes (Meyers et al
2003) Allochthonous organic sources (high CN ratios) produce isotope values
similar to values from catchment vegetation Autochthonous organic matter (low CN
ratio) is influenced by fractionation both in the lake and the watershed leading up to
carbon assimilation by lacustrine biota (Woodward et al 2011) Changes in
productivity greatly affect δ13COM values (Torres et al 2012) as during C uptake
plankton preferentially assimilate 12C leaving the dissolved inorganic carbon (DIC)
76
pool enriched in 13C (Gu et al 2006) with phytoplankton averaging ca 20 permil lower
than the respective δ13CDIC values (Leng et al 2006) Under conditions of low to
moderate primary productivity plankton preferentially uptake the lighter 12C
resulting in low δ13C values displayed by the OM (Torres et al 2012) Conversely
during high primary productivity phytoplankton will uptake 12C until its depletion and
is then forced to assimilate the heavier isotope resulting in an increase in δ13C
values Higher productivity in C-limited lakes due to slow water-atmosphere
exchange of CO2 also results in high δ13C values (Galman et al 2009) In these
cases algae are forced to uptake dissolved bicarbonate with δ13C values between
7 to 10permil higher than those of the atmospheric CO2 dissolved in water (Sun et al
2016 Torres et al 2012 Galman et al 2009)
Stable isotope analyses from lake sediments are thus useful tools to
reconstruct shifts in lake-watershed dynamics caused by changes in limnological
parameters and LUCC Our knowledge of the current processes that can affect
stable isotope signals in a watershed-lake system is limited however as monitoring
studies are scarce Besides in order to use stable isotope signatures to reconstruct
past environmental changes we require a multiproxy approach to understand the
role of the different variables in controlling these values Hence in this study we
carried out a detailed survey of current N dynamics in a coastal central Chilean lake
(Lago Vichuqueacuten) and a multiproxy study of a sediment record spanning the last
600 years The characterization of the recent changes in the watershed since 1970s
is based on satellite images to compare recent changes in the lake and assess how
these are related with climate variability and an ever increasing human footprint
(Lara et al 2012 Gallardo et al 2015 Steffen et al 2015) Our main goal is to
investigate how stable isotope values from lake sediment reflect changes in the lake
77
ndash watershed system during periods of high watershed disruption (eg Spanish
Conquest late XIX century Great Acceleration) and recent climate change (eg
Little Ice Age and current global warming)
2 Study Site
Lago Vichuqueacuten (34ordm49`S 72ordm04W 4 m asl 30 m of depth Fig 1) is a
mesotrophic to eutrophic lake with dominant anoxic bottom conditions that is
stratified year around (Frugone-Aacutelvarez et al 2017) The main inlets are the
Vichuqueacuten and Huintildee rivers and the only outlet is the Llico estuary that drains into
the Pacific Ocean High tides can sporadically shift the flow direction of the Llico
estuary which increases the marine influence in the lake Dune accretion gradually
limited ocean-lake connectivity until the estuary was almost completely closed off
by ca 10 ky BP and the current lake regime formed (Frugone-Aacutelvarez et al 2017)
The area is characterized by a mediterranean climate with cold-wet winters and
hot-dry summers and an annual precipitation of ~650 mm and a mean annual
temperature of 15ordmC During the austral winter months (June - August) precipitation
is modulated by north-west shifts of the South Pacific Anticyclone (SPA) followed by
an increased frequency of storm fronts stemming off the South Westerly Winds
(SWW) A strengthened SPA during austral summers (December - March) which
are typically dry and warm blocks the northward migration of storm tracks stemming
off the SWW (Fig1 Frugone-Aacutelvarez et al 2017)
78
Figure 1 a) The location of the Lago Vichuqueacuten in central Chile b) Map of land
uses and cover in 2016 c) Ombrothermic diagram mediterranean climates are
characterized by cold-wet winters with surplus moisture from June to August and
hot-dry summers d) Lake bathymetry showing location of cores and water sampling
sites used in this study
Although major land cover changes in the area have occurred since 1975 to the
present as the native forests were replaced by tree (Monterey pine and eucalyptus)
plantations the region was settled before the Spanish conquest (Frugone-Alvarez
et al 2018) Historical documents indicate that Vichuqueacuten was occupied by a
Mitimaes colony as part of the Inka empire (Odone 1998) Unlike other Andean
areas during the Inka period the introduction of corn and quinoa in the Vichuqueacuten
watershed do not seem to have intensified land use The Spanish colonial period in
Chile lasted from 1542 CE to the independence in 1810 CE The first historical
document (1550 CE) shows that the areas around Vichuqueacuten were settled by the
Spanish early on as the Vichuqueacuten watershed was given as an ldquoencomiendardquo
system to Juan Cuevas The ldquoencomiendardquo system provided the owners with land
79
and indigenous people to work but also the introduction of wheat wine cattle
grazing and logging of native forests for lumber extraction and increasing land for
agricultural activities (Odone et al 1998 Armesto et al 2010) During the 19th
century (the Republic) the export of wheat to Australia and Canada generated
intensive changes in land cover use The town of Vichuqueacuten became the regional
capital and plans to build a maritime port in the bay of Vichuqueacuten were drawn
However the fall of international markets in 1880 paralyzed these plans During the
20th century the plantation of trees (Pinus radiata and Eucalyptus globulus) in areas
cleared of native forests was subsidized by two national laws DFL nordm265 (1931) and
DFL nordm 701 (1974) both of which provided funds for such plantations During the last
decades the urbanization with summer vacation homes along the shorelines of
Lago Vichuqueacuten has greatly increased Extensive lake eutrophication is currently a
large environmental problem (EULA 2008)
3 Methods
Coring campaigns were organized from 2011 to 2015 (Fig 1d) We recovered
12 short cores and 4 long cores (up to 14 m long) at several sites using a hammer-
modified UWITEC gravity corer Sediment cores (VIC11-2A 170 cm VIC13-2B 170
cm VIC13-2D 1200 cm and VIC15-1A 119 cm) were split photographed sub-
sampled and stored in the Pyrenean Institute of Ecology (IPE-CSIC Spain) Core
VIC13-2B was selected for detailed multiproxy analyses (including elemental
geochemistry C and N isotopes XRF and 14C dating) Stable isotope analyses
(δ13C δ15N) were performed at the Laboratory of Biogeochemistry and Applied
Stable Isotopes (LABASI-PUC Chile) Sediment samples for δ13Corg were pre-
treated with 50 ml of HCl and 50 ml of deionized water and dried at 60ordmC for 4 hr to
remove carbonates (Harris et al 2011) Isotope analyses were conducted using a
80
Delta V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via
a Conflo IV interface Isotope results are expressed in standard delta notation (δ)
and expressed in per mil (permil) relative to the Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Total carbon (TC)
Total Organic Carbon (TOC) Total Inorganic Carbon (TIC) and Total Sulfur (TS)
were measured every 2 cm with a LECO SC 144 DR at the IPE-CSIC
An AVAATECH X-Ray Fluorescence II core scanner with a Rh X-ray tube from
the University of Barcelona was used to obtain XRF logs every 4 mm of resolution
Results are expressed as element intensities in counts per second (cps) Tube
voltage was operated at 30 kV and 10 kV to obtain the abundances of 18 elements
(Pb Al Si S Cl K Ca Ti V Mn Fe Co Zn Br Rb Sr Y Zr) with an average of
at least of 1800 cps (less for Pb=437 and Zn=934 cps) Pb was included due to
similar behavior with Co and Fe Element ratios were calculated to describe changes
in redox conditions (MnFe) endogenic productivity (BrTi) carbonate formation
(SrCa and CaTi) and sediment input (ZrTi) (Barreiro-Lostres et al 2014
Carrevedo et al 2015 Frugone-Aacutelvarez et al 2017 Kylander et al 2011 Moreno
et al 2007a)
Several campaigns were carried out to sample the POM from the water column
two per hydrologic year from November 2015 to August 2018 A liter of water was
recovered in three sites through to the lake two are from the shallower areas (with
samples taken at 2 and 5 m depth at each site) and one in the deeper central portion
(at 2 5 and 20 m depth) of the lake (Fig 1d) The samples were filtered using glass
fiber filters (25 m) and then frozen to obtain the stable carbon and nitrogen isotope
signal of lacustrine POM Additionally soil and vegetation samples from the
following communities native species meadow hydrophytic vegetation and
81
Monterrey pine (Pinus radiata) were taken for stable isotope analyses (see in
supplementary material)
The age model for the complete Lago Vichuqueacuten sedimentary sequence is
based on Frugone-Aacutelvarez et al (2017) with the last 1000 years based on
210Pb137Cs dating techniques in VIC15-1A and 14C AMS dates on bulk sediment
samples (Supplementary Table S1) The 14C measurements of lake water DIC show
a moderate reservoir effect of 180 plusmn 25 14C yr) The revised age-depth model used
here includes three more 14C AMS dates performed with the program Bacon to
establish the deposition rates along the core (Blaauw and Christen 2011 Fig 2)
The age-depth model indicates that average resolution between 0 to 87 cm is lt2
cm per year and from 88 to 170 cm it is lt47 cm per year
82
Figure 2 Chronological age-depth model for the Lago Vichuqueacuten sedimentary
sequence spanning the last 1000 yr (see also Frugone-Alvarez et al 2017)
To estimate land use changes in the watershed we use Landsat MSS images
for 1975 and Landsat TM for 1989 and 2014 all taken in either summer or autumn
(Table 1) We performed supervised classification of land uses (maximum likelihood
83
algorithm) for each year (1975 1989 and 2014) and results were mapped using
ArcGIS 102
Table 1 Images using for LUCC reconstruction
Source of LUCC
Acquisition
Date Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat TM 19991226 30 m
CONAF 2009 30 m
Land cover Chile 2014 30 m
CONAF 2016 30 m
Previous Work on Lago Vichuqueacuten sedimentary sequence
The sediments are organic-poor dark brown to brown laminated silt with some
intercalated thin coarser clastic layers Lacustrine facies have been classified
according to elemental composition (TOC TS TIC and TN) grain size and
sedimentary textures (Frugone-Aacutelvarez et al 2017) Four main types of lacustrine
facies were identified in this short core Facies L1 is a laminated (1cm) black to dark
brown diatom-poor silt with relatively higher TOC percentages (TOC about 185)
TS (about of 18) and TOCTN (about 92) Facies L2 is characterized by a
homogeneous black organic-poor silty clay with low TOC (mean= 13) TS (mean=
13) TOCTN (mean= 88) values Facies L3 is a laminated (1cm) black organic-
poor (TOC mean 14) silts with higher TS (mean= 21) and TOCTN ratios
(mean= 88) Facies L3 is interpreted as ldquobaselinerdquo deposition on the distal areas
84
of Vichuqueacuten lake Mineralogy is dominated by mica (muscovite) clay minerals
(chlorite) and plagioclase (albite) Quartz and calcite occur at the top and pyrite
occurs in the lower part of the sequence Facies T is composed by massive banded
sediments 4 cm thick and coarse grain size It is interpreted as an instantaneous
depositional event (for more detail see Frugone-Aacutelvarez et al 2017) In this work
we identified four subunits based on geochemical and stable isotope signals
4 Results
41 Geochemistry and PCA analysis
High positive correlations exist between Al Si K and Ti (r = 078 ndash 096
supplementary Fig S1) and between Zn Zr Rb and Y (r = 072 - 096) which reflect
the silicate content of the watershed-derived sediments (Marzecoacuteva et al 2011) Zr
is commonly associated with minerals more abundant in coarser deposits Thus the
ZrTi profile could reflect grain size (Cuven et al 2010) showing higher variability
in the upper part of the Lake Vichuqueacuten sequence and in the alternation between
facies L1 and L2 Another group of elements made up of Co Fe and Pb displayed
positive correlations (r = 067ndash 097) and represents the input of heavy metals Br
Cl and Ca show a positive correlation (r = 049 ndash 06 p-value = 00001) BrTi ratio
is interpreted as a productivity indicator due to Br having a strong affinity with humic
and organic compounds (Marzecoacuteva et al 2011 Frugone-Aacutelvarez et al 2017) In
our Lago Vichuqueacuten sequence BrTi values are high from 170 to 132 cm and from
36 cm to the top of core where it shows several sharps peaks (Fig 3) The MnFe
ratio is indicative of lake bottom oxygenation (Naecher et al 2013) as under
reducing conditions Mn tends to become more mobile than Fe leading to a
decreased MnFe (Marzecoacuteva et al 2011) Several sharp peaks in MnFe occurred
85
from 127 to 120 cm and from 57 to 30 cm (MgFegt03) The silicate group and the
Br Cl Ca Mn group are negatively correlated (r= -012 and -066)
Principal Component Analysis (PCA) was undertaken on the XRF
geochemical data to investigate the main factors controlling sediment deposition in
Lago Vichuqueacuten (Fig 3) The four main eigenvectors explain 801 of the variance
(supplementary material Table S2) The principal component (PC1) explain 437
of the total of variance and grouped elements are associated with terrigenous input
to the lake Positive values of the biplot have been attributed to higher heavy metals
deposition (Pb Co and Fe) and negative values to higher silicate input (Al K Ti and
Si) in a high energy catchment environment (Zr) The PC2 explains 198 of the
total of variance and highlights the endogenic productivity in the lake The positive
loading is compound by Br Cl Ca and Sr and to a lesser extent by Y Zn Rb and
Mn The PC2 may explain both the precipitation of carbonates (Ca Sr) and biological
production (Br)
86
Figure 3 Principal Component Analysis of XRF geochemical measurements in
VIC13-2B Lago Vichuqueacuten lake sediments
42 Sedimentary units
Based on geochemical and stable isotope analysis we identified four
lithological subunits in the short core sedimentary sequence Our PCA analyses and
Pearson correlations pointed out which variables were better for characterizing the
subunits (Fig 4) in terms of endogenic productivity heavy metals and terrestrial
input The unit 2c (170-131 cm) is characterized by the occurrence of some clastic
layers (L2 from 160 and 150 cm) high δsup1⁵N values (mean= +59 plusmn 04permil) with
Magnetic Susceptibility (MS) BrTi ZrTi and SrCa values increase towards the top
Also it has relatively low δsup1sup3C values (mean= -265 plusmn 11permil) and CNmolar ratios
(mean=78 plusmn 04) The ZrTi show upwards increasing values reaching the highest
values of the sequence at the top of this unit suggesting a coarsening upward trend
and relatively higher depositional energy The MS trend also indicates higher
erosion in the watershed and enhanced delivery of ferromagnetic minerals likely
from soil particles particularly from 150 to 160 cm (Frugone-Aacutelvarez at al 2017)
The subunit 2b (130-118 cm) is also composed of black silts but it has the
lowest MS values of the whole sequence and its onset is marked by a sharp
decrease in ZrTi This subunit is marked by sharp peaks of MnFe (at 126 and 120
cmgt027) SrCa (at 125 and 120 cmgt033) CaTi (at 124 and 120 cmgt199) TOC
(about 26 at 130 124 122 and 118 cm) and δsup1⁵N (gt+67permil at 126 and 120 cm)
BrTi oscillates around low values (about 01) whereas the δsup1sup3C values range
between -262 and -282permil
87
The unit 2a (58-117 cm) shows increasing and then decreasing MS values
and displayed sharps peaks of δsup1⁵N (gt64 at 102 to 99 87 and 75 cm) and CN
(about 10 at 86 74 66 and 60 cm) SrCa (mean = 04 plusmn 013) BrTi (mean = 008
plusmn 002) MnFe (mean = 002 plusmn 001) and δsup1sup3C (mean = -260 plusmn 07permil) oscillate in
low values The upper part of this unit (72 ndash 57 cm) shows an increase of SrCa
(from 03 to 05)
The upper unit 1a (0 - 57 cm) starts with a fine clastic layer (T at 57 to 54
cm) interpreted as deposition during a high-energy event It is characterized by
lower BrTi (mean = 003 plusmn 002) TOC (mean = 12 plusmn 03) and δsup1sup3C (mean= -
266 plusmn 06permil) higher δsup1⁵N (mean = +55 plusmn 08 permil) This unit is made of alternating
fine (lt 1cm) black and dark brown laminae with carbonate (L1 facies) Recently
deposited sediments show decreasing MS values increasing TOC (mean= 15 plusmn
04 ) sharp peaks of BrTi (gt015 at 33 21 5 and 1 cm) and the lowest δsup1⁵N values
of the entire record (mean= 45 plusmn 06 permil) and δsup1sup3C values (mean= -277 plusmn 09 permil)
Heavy metal content increases abruptly in the upper section of the core (2 - 20 cm)
(peaks of FeTi CoTi and PbTi)
88
Figure 4 Sedimentary units of Lago Vichuqueacuten sedimentary sequence Selected
variables illustrate watershed input (Magnetic Susceptibility ZrTi CoTi and MnFe)
endogenic productivity (SrCa CaTi TS) organic matter source (BrTi TOC
CNmolar and stable isotope records (δ13Corg and δ15Nbulk)
43 Recent seasonal changes of particulate organic matter on water column
The average of δ15NPOM from the three sites across to the lake was +86 plusmn 58
permil oscillating widely from -14 to +193permil (Fig 5) Important seasonal differences
occur at 2 and 5 m depth with more negative values during summer (~53 plusmn 52 permil)
than winter (~122 plusmn 40permil) At 20 m depth δ15NPOM values exhibit small seasonal
ranges (mean = +10permil for winter and +87permil for summer) The δ13CPOM average was
-296 plusmn 33permil with slightly seasonal and water column depth differences However
more positive δ13CPOM values occurred during winter (mean=-290 plusmn 41permil) than in
summer (mean= -301 plusmn 19 permil) Like the δ15NPOM values CNPOM (mean= 83 plusmn 17)
displayed important seasonal and water depth differences Lower CNPOM ratios
89
occurred during summer at 2 and 5 m depth (mean= 95 plusmn 17) and higher and more
constant values during winter (mean = 73 plusmn 09) at the same depth At 20 m CNPOM
shows similar values in both winter (70) and summer (74)
Figure 5 Seasonal POM averages from samples taken of the Lago Vichuqueacuten
water column Samples were taken from 2015 to 2018 at 2 (n=16) 5 (n=14) and 20
(n=8) meters depth
44 Stable isotope values across the Lake Vichuqueacuten watershed
Figure 6 shows modern vegetation soil and sediment isotope values found for
the watershed Huintildee tributary and Vichuqueacuten lake The δsup1⁵NFoliar values from
meadow plantations and macrophytes have similar range values with a mean of
+89 plusmn50 permil (n=18) +74 plusmn 75 permil (n=5) +82 plusmn 36 permil (n=6) respectively Native
vegetation has overall lower δsup1⁵NFoliar values with a mean of 14 plusmn 21 permil (n=28 see
Table 2 in supplementary material) The δsup1sup3CFoliar from terrestrial samples exhibit
similar values across the different plant communities (tree plantation mean=-274 plusmn
13permil meadow mean = -275 plusmn 23 permil native species mean=-272 plusmn 14permil) whereas
macrophytes display slightly more negative values with a mean of -287 plusmn 23permil
Macrophytes substrate displayed the highest δsup1⁵N values with a mean of +60 plusmn
14permil (n=3) Similar and relatively high δsup1⁵N values for soil of plantation (mean= +54
plusmn 13permil n=4) meadow (mean= +49 plusmn 04permil n=8) and surface river sediment
90
(mean= +52 plusmn 17permil n=10) Native forest soils oscillate around slightly more
negative values with a mean of +45 plusmn 12permil (n=4) Slightly more positive soil δsup1sup3C
values occur both underneath native forests and in tree plantations with means of -
284 plusmn 23permil and -298 plusmn 13permil respectively compared to either meadow soils
(mean= -302 plusmn 11permil) aquatic sediments underneath macrophytes (-311 plusmn 10permil)
or from surface river sediments (mean= -312 plusmn 10permil)
Figure 6 Stable isotope content of modern soil river sediment and foliar vegetation
used as end members in the sedimentary sequence of Lago Vichuqueacuten a)
Scatterplot of foliar δsup1⁵N and δsup13C values for vegetation from Lago Vichuqueacuten
watershed (plantation meadow and native species) and macrophytes on Lake
Vichuqueacuten See supplementary material for more detail of vegetation types b)
Scatterplot of soil δsup1⁵N and δsup13C values across the watershed the sediments of the
Huintildee and Vichuqueacuten rivers and the fluvial sediment corresponding to the
macrophyte vegetation
45 Land use and cover change from 1975 to 2014
Major land use changes between 1975 CE and 2016 CE in the Lago
Vichuqueacuten watershed are summarized in Figure 7 The watershed has a surface
area of 535329 km2 of which native vegetation (26) and shrublands (53)
represent 79 of the total surface in 1975 Meadows are confined to the valley and
91
represent 17 of watershed surface Tree plantations initially occupied 1 of the
watershed and were first located along the lake periphery By 1989 the areas of
native forests shrublands and meadows had decreased to 22 31 and 14
respectively whereas tree plantations had expanded to 30 These trends
continued almost invariably until 2016 when shrublands and meadows reached 17
and 5 of the total areas while tree plantations increased to 66 Native forests
had practically disappeared by 1989 and then increased up to 7 of the total area
in 2016 (Fig 6) preferentially occupying the southeastern sector of the watershed
Figure 7 Changes in land use and cover between 1975 and 2016 in the Lago
Vichuqueacuten watershed as measured from satellite images The major change is
represented by the replacement of native forest shrubland and meadows by
plantations of Monterrey pine (Pinus radiata)
Figure 8 shows correlations between lake sediment stable isotope values and
changes in the soil cover from 1975 to 2013 Positive relationships occurred
between sediment δsup1⁵N and δsup13C values from core VIC13-2B (unit 1) and the
92
percentage of native forest (r=089 for δsup1⁵N 082 for δsup13C) shrublands (r = 071 for
δsup1⁵N 057 for δsup13C) and meadow cover (r = 088 for δsup1⁵N 081 for δsup13C) All of these
correlations are significant (p value lt 0001) In contrast significant negative
correlations (p lt0001) occurred between tree plantation cover and lake sediment
stable isotope values (δsup1⁵N r = -082 and δsup13C r = -067) native forest (r = -097)
meadows (r = -086) and shrubland (r =-093)
Figure 8 Correlation plots of land use and cover change versus lake sediment
stable isotope values The δsup1⁵N values are positively correlated with native forests
agricultural fields and meadow cover across the watershed Total Plantation area
increases are negatively correlated with native forest meadow and shrubland total
area Significance levels are indicated by the symbols p-values (0 0001 001
005 01 1) lt=gt symbols ( )
93
5 Discussion
51 Seasonal variability of POM in the water column
The stable isotope values of POM can vary during the annual cycle due to
climate and biologic controls namely temperature and length of the photoperiod
which affect phytoplankton growth rates and isotope fractionation in the water
column (Gu et al 2006 Leng et al 2006) POM from Lago Vichuqueacuten surface
samples (2 and 5 m depth) displayed lower δ13CPOM values during the summer than
in winter During C uptake phytoplankton preferentially utilize 12C leaving the
DICpool enriched in 13C Therefore as temperature increases during the summer
phytoplankton growth generates OM enriched in 12C until this becomes depleted
and then the biomas come to enriched u At the onset of winter the DICpool is now
enriched in 13C and despite an overall decrease in phytoplankton production the
OM produced will also be enriched in 13C In contrast δ13CPOM values at 20 m depth
did not reflect these seasonal differences probably due to water-column
stratification that maintains similar temperatures and biological activity throughout
the year
Lake N availability depends on N sources including inputs from the
watershed and the atmosphere (ie deposition of N compounds and fixation of
atmospheric N2) which varies during the hydrologic year The fixation of atmospheric
N2 is an important natural source of N to the lake occurring mainly during the
summer season associated with higher temperature and light (Gu et al 2006)
Because the δ15NPOM of N2 fixers is close to 0permil with almost no overall isotope
fractionation (Leng et al 2006) when N2 is the major N source δ15NPOM values are
typically low However when DIN concentrations are high or alternatively when little
94
N2 fixation occurs δ15NPOM values will be higher (Gu et al 2006) δ15NPOM values
from summer Lago Vichuqueacuten samples were lower than those from winter with large
differences between the seasons (~5permil Fig 5 and 9) Furthermore δ15NPOM values
were high when monthly average temperature was low and monthly precipitation
was high (Fig 9) This pattern is consistent with the dominance of high N2 fixation
by cyanobacteria associated with increased summer temperatures This correlation
of δ15NPOM values with temperature further suggests a functional group shift i e
from N fixers to phytoplankton that uptake DIN The correlation between wetter
months and higher δ15NPOM values could be caused by increased N input from the
watershed due to increased runoff during the winter season The lack of data of the
δ15NPOM between the 30 mm and the 160 mm of precipitation is due to the
mediterranean-type climate that concentrates precipitations in the winter months
Finally Lago Vichuqueacuten is characterized by pronounced seasonality leading to
higher phytoplankton biomass in summer characterized by low δ15NPOM In winter
low biomass production and increased input from watershed is associated to high
δ15NPOM
95
Figure 9 Seasonal changes can drive δ15NPOM values in Lago Vichuqueacuten Data
correspond to average monthly temperature and total monthly precipitation for the
months when the water samples were taken (years 2015 - 2018) P-valuelt005
52 Stable isotope signatures in the Lake Vichuqueacuten watershed
The natural abundance of 15N14N isotopes of soil and vegetation samples
from the Lago Vichuqueacuten watershed appear to result from a combination of factors
isotope fractionation different N sources for plants and soil microorganisms (eg N2
fixation Nitrates) depth in the soil where N uptake by vegetation occurs and N loss
mechanisms (ie denitrification leaching and ammonia volatilization Hogberg
1997) The lowest δsup1⁵Nfoliar values are associated with native species and are
probably due to the contribution of N-fixing species (eg Sophora) (Fig6 and for
more detail see Table S3 in supplementary material) The number native N-fixers
species present in the Chilean mediterranean vegetation are not well known
however so there could be other N-fixing species in this group Alternatively δsup1⁵Nfoliar
values reflect soil N uptake (Kahmen et al 2008) In environments limited by N
plants have δsup1⁵N values similar to the soil δsup1⁵N values however the denitrification
and volatilization of ammonia can lead to the remain N of soil to come enriched in
15N (Cifuentes et al 1989 Craine et al 2009 Bruland and Mckenzie 2010) Soil N
isotope samples from native species communities tends to display relatively high
δsup1⁵N values respect to foliar samples due to loss of N-soil
The higher foliar and soil δsup1⁵N values obtained from samples of meadows
aquatic macrophytes and tree plantations can be attributed to the presence of
greater amounts of nitrate available for plant uptake (Fig 6) Houmlgberg (1997)
suggests that the availability of different N sources in soils (ie nitrates versus
96
ammonia) with different residence times can also explain these δsup1⁵NFoliar values
Indeed Feigin et al (1974) described differences of up to 20permil between ammonia
and nitrates sources Denitrification and nitrification discriminate much more against
15N than N2 fixation does (Craine et al 2009) leading to soils (and plants after
uptake) enriched in 14N
In general multiple processes that affect the isotopic signal result in similar
δsup1⁵N values between the soil of the watershed and the sediments of the river
However POM isotope fluctuations allow to say that more negative δsup1⁵N values are
associated to lake productivity while more positive δsup1⁵N values are associated with
N input from the watershed
δ13C values in Lago Vichuqueacuten watershed reflect CO2 gas exchange between
C3 plants and algae with the atmosphere During photosynthesis plants
discriminate against the heavier stable isotope of carbon (13C) in favor of the lighter
isotope 12C which results in δ13CFoliar values between -220permil and -320permil (Browman
and Cook 2002 Liu et al 2007) Likewise the δ13CFoliar values from Lago Vichuqueacuten
oscillate around -27permil (Fig 6) Plants transform atmospheric CO2 into soil organic
carbon (C) which in turn reflects this initial discrimination against 13C during C
uptake and post photosynthetic fractionation (Farquhar et al 1982 1989 Badeck
et al 2005 Pausch and Kuzyakov 2017) Slightly more positive δ13CFoliar values
(about 15permil) were measured in comparison with their δ13CSoil values This may be
reflecting the C transference from plants to the soil but also a soil-atmosphere
interchange The preferential assimilation of the light isotopes (12C) during soil
respiration carried by the roots and the microbial biomass that is associated with the
decomposition of litter roots and soil organic matter explain this differential
(Berhardt et al 2006 Blagodatskaya et al 2010 Midwood and Millard 2011)
97
In general the δ13CSoil values from the Lago Vichuqueacuten watershed oscillated
around -290permil and did not vary with our plant classification types Here we use
these values as terrestrial-end members to track changes in source OM (Fig 6)
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from the terrestrial watershed By the other hand more positive δ13C
values most likely reflect an increased aquatic OM component as indicated by POM
isotope fluctuations (Fig 9)
53 Recently land use and cover change and its influences on N inputs to the lake
Tree plantations in the Lago Vichuqueacuten watershed have increased greatly in
the last few decades (from 1 in 1975 to 66 in 2016 Fig 7) replacing previous
native forests (from 26 to 7) meadows (17 to 5) and shrublands (53 to
17) In 1975 tree plantations were confined to the lake perimeter with discrete
patches distributed throughout the watershed The ldquoForest Lawrdquo (DFL 701- passed
in 1974) allocated state funding to afforestation efforts and management of tree
plantations which greatly favored the replacement native forests by introduced trees
This increase is marked by a sharp and steady decrease in lake sediment δ15N and
δ13C values because tree plantations function as a nutrient sink whereas other land
uses such as meadows favor nutrient loss from the watershed (Fig8) Bruland and
Mackenzie (2014) noted a decrease in wetland δ15N values when watershed
forested cover increased and concluded that N inputs to the wetlands are lower from
the forested areas as they generally do not export as much N as agricultural lands
A positive correlation between native vegetation and δ15Ncore values can be
explained by the relatively scarce arboreal cover in the watershed in 1975 when
native forest occupied just 26 of the watershed surface whereas shrublands and
98
meadows occupied more than the 70 of the surface of the watershed with the
concomitant elevated loss of N (Fig 7 and Fig 8)
54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years
Sedimentological compositional and geochemical indicators all show changes in
the N cycle associated to LUCC lake dynamics and climate changes (Fig 10) From
the pre-Columbian indigenous settlement including the Spanish colonial period up
to the start of the Republic (1300 - 1800 CE) the introduction of crops such as
quinoa and wheat but also the clearing of land for extensive agriculture would have
favored the entry of N into the lake Conversely major changes observed during the
last century were characterized by a sharp decrease of N input that were coeval
with the increase of tree plantations
From 1365 until 1550 CE (unit 2c 2b and half of 2a Fig 3 and Fig 10) pre-
Columbian agricultural societies planted mostly crops of corn and quinoa (Ramirez
and Vidal 1985) This period is characterized by large peaks seen in the δsup1⁵N record
(eg 1403 1452 1476 and 1505) with overall high δsup1⁵N values (up to 5permil) indicating
that N input from watershed was elevated and oscillating to the beat of the NT These
positive δsup1⁵N peaks could be due to several causes including a) the clearing of land
for farming b) N loss via denitrification which would be generally augmented in
anoxic environments (Diebel et al 2009) such as Lago Vichuqueacuten (low MnFe
values about 0001 Fig 4) or c) climate especially cold-wet winters or hot-dry
summers can also exert control on the δsup1⁵N record Indeed tree-ring records and
summer temperature reconstructions show overall wetcold conditions during this
period (Christie et al 2009 Von Gunten et al 2009 Fig 10) Increased
precipitation would bring more sediment (and nutrients) from the watershed into the
99
lake and increase lake productivity which is also detected by the geochemical
proxies for productivity (peaks of BrTi SrCa and TOC Fig 4 and Fig 10) (see also
Frugone-Alvarez et al 2017)
Figure 10 Changes in the N availability during the last six centuries in Lago
Vichuqueacuten a) lake sediment δsup1⁵N values displayed more positive values during the
prehistoric period Spanish Colony and the starting 19th century which is associated
with enhanced N input from the watershed by extensive clearing and crop
plantations The inset shows this relationship between sediment δsup1⁵N and
100
percentage of meadow cover over the last 30 years b) Summer temperature
reconstruction from central Chile (von Gunten et al 2009) showing a
correspondence with cold phases and high δsup1⁵N values at Vichuqueacuten except for the
last 100 years c) Palmer Drought Severity Index (PDSI) is an interannual moisture
variability reconstruction for late springndashearly summer during the last six centuries
(Christie et al 2009) Grey shadow indicating higher precipitation periods
From 1550 and 1810 CE (unit 2a) and during the Colonial period several peaks
of δ15N could be further related not only to high precipitation (Fig 10 grey shadow)
but also pulses of enhanced N input from the watershed linked to human land use
In 1550 CE Juan Cuevas was granted lands and indigenous workers under the
encomienda system for agricultural and mining development of the Vichuqueacuten
village (Ramiacuterez and Vidal 1985) Historical documents show that around 1579 CE
the Vichuqueacuten watershed was occupied by indigenous communities dedicated to
wheat plantations and vineyards wood extraction and gold mining (Odone 1998)
The introduction of the Spanish agricultural system implied not just a change in the
types of crops used (from quinoa to vineyards and wheat) but also a clearing of
native species for the continuous increase of agricultural surface and wood
extraction During the Colonial period wheat from Vichuqueacuten was exported to Peru
(Ramiacuterez and Vidal 1985) Armesto et al (2009) indicate that during the XVIII and
XIX centuries the extraction of wood for mining operations was important enough to
cause extensive loss of native forests The independence and instauration of the
Chilean Republic did not change this prevailing system Increases in the
contributions of N to the lake during the second half of the XIX century (peaks in
δsup1⁵N ca 1870 CE) could be due to expanding farm land dedicated for wheat
101
production and increased commercial trade with California and Canada (Ramiacuterez
and Vidal 1985)
In contrast LUCC in the last century are clearly related to the development of
large-scale tree plantations (Fig 7) The δsup1⁵N record reaches the lowest values of
the entire sequence in the last few decades (Fig 10) A marked increase in lake
productivity NT concentration and decreasing sediment input is synchronous (unit
1 Fig 4) with trees replacing meadows shrublands and areas with native forests
(Fig 7 and 8) The implementation of the lsquoForest Lawrsquo in 1974 had a stronger impact
on the landscape and lake ecosystem dynamics than the impacts of ongoing climate
change in the region which is much more recent (Garreaud et al 2018) although
the prevalence of hot dry summers seen over the last decade would also be
associated with low δsup1⁵N values and increase of NT (Fig 10) Heavy metals ratios
(FeTi CoTi and PbTi) show notable increases in lake sediments from 1992 to 2011
CE (Fig 4) Although this could be related to mining in the El Maule region the
closest mines are 60 Km away (Pencahue and Romeral) so local factors related to
shoreline urbanization for the summer homes and an increase in tourist activity
could also be a major factor
6 Conclusions
The N isotope signal in the watershed depends on the rates of exchange
between vegetation and soil cover (Fig 11) Plants preferentially uptake 14N and the
underlying soils become enriched in 15N especially when the terrestrial ecosystem
is N-limited andor significant N loss occurs (ie denitrification andor ammonia
volatilization) LUCC in the Lago Vichuqueacuten watershed have heavily modified the
links between terrestrial and aquatic ecosystems with agriculture practices
102
contributing more N to the lake than tree plantations or native forests In situ lake
processes can also fractionate N isotopes An increase of N-fixing species results
in OM depleted in 15N which results in POM with lower δsup1⁵N values during these
periods During winter phytoplankton is typically enriched in 15N due to the
decreased abundance of N-fixing species and increased N input from the
watershed This in turn generates the highest δsup1⁵NPOM values in Lago Vichuqueacuten
Periods of elevated productivity tend to deplete the dissolved inorganic N of 14N
resulting in even higher δ15N values
Our study illustrates the usefulness of δsup1⁵N as a tool for reconstructing the past
influence of LUCC on N availability in lake ecosystems To constrain the relative
roles of the diverse forcing mechanisms that can alter N cycling in mediterranean
ecosystems all main components of the N cycle should be monitored seasonally
(or monthly) including the measurements of δ15N values in land samples
(vegetation-soil) as well as POM
103
Figure 11 Summary of human and environmental factors controlling the δ15N
values of lake sediments Particulate organic matter(POM) δ15N values in
mediterranean lakes are driven by N input from the watershed that in turn depend
on land use and cover changes (ie forest plantation agriculture) andor seasonal
changes in N sources andor lake ecosystem processes (ie bioproductivity redox
condition denitrification) Arrows indicate the direction of N fluxes (inputoutput from
the N cycle) N cycle processes that deplete lake sediments of 15N are shown in
blue whereas those that enrich sediments in 15N are shown in red
104
Supplementary material
Figure S1 Pearson correlate coefficient between geochemical variables in core
VIC13-2B Positive and large correlations are in blue whereas negative and small
correlations are in red (p valuelt0001)
Figure S2 Principal Component Analysis of geochemical elements from core
VIC13-2B
105
Table S1 Lago Vichuqueacuten radiocarbon samples
RADIOCARBON
LAB CODE
SAMPLE
CODE
DEPTH
(m)
MATERIAL
DATED
14C AGE ERROR
D-AMS 029287
VIC13-2B-
1 043 Bulk 1520 24
D-AMS 029285
VIC13-2B-
2 085 Bulk 1700 22
D-AMS 029286
VIC13-2B-
2 124 Bulk 1100 29
Poz-63883 Chill-2D-1 191 Bulk 945 30
D-AMS 001133
VIC11-2A-
2 201 Bulk 1150 44
Poz-63884
Chill-2D-
1U 299 Bulk 1935 30
Poz-64089
VIC13-2D-
2U 463 Bulk 1845 30
Poz-64090
VIC13-2A-
3U 469 Bulk 1830 35
D-AMS 010068
VIC13-2D-
4U 667 Bulk 2831 25
Poz-63886
VIC13-2D-
4U 719 Bulk 3375 35
106
D-AMS 010069
VIC13-2D-
5U 775 Bulk 3143 27
Poz-64088
VIC13-2D-
5U 807 Bulk 3835 35
D-AMS-010066
VIC13-2D-
7U 1075 Bulk 6174 31
Poz-63885
VIC13-2D-
7U 1197 Bulk 6440 40
Poz-5782 VIC13-15 DIC 180 25
Table S2 Loadings of the trace chemical elements used in the PCA
Elementos PC1 PC2 PC3 PC4
Zr 0922 0025 -0108 -0007
Zn 0913 -0124 -0212 0001
Rb 0898 -0057 -0228 0016
K 0843 0459 0108 0113
Ti 0827 0497 0060 -0029
Al 0806 0467 0080 0107
Si 0803 0474 0133 0136
Y 0784 -0293 -0174 0262
V 0766 0455 0090 -0057
Br 0422 -0716 -0045 0226
Ca 0316 -0429 0577 0489
Sr 0164 -0420 0342 -0182
Cl 0151 -0781 -0397 0162
107
Mn -0121 -0091 0859 0095
S -0174 -0179 -0051 0714
Pb -0349 0414 -0282 0500
Fe -0700 0584 -0023 0280
Co -0704 0564 -0107 0250
Table S3 Stable Isotope values from vegetation of Lago Vichuqueacutenacutes watershed
Taxa Classification δsup1⁵N δsup1sup3C CN
molar
Poaceae Meadow 1216 -2589 3602
Juncacea Meadow 1404 -2450 3855
Cyperaceae Meadow 1031 -2596 1711
Taraxacum
officinale Meadow 836 -2400 2035
Poaceae Meadow 660 -2779 1583
Poaceae Meadow 453 -2813 1401
Poaceae Meadow 966 -2908 4010
Juncus Meadow 1247 -2418 3892
Poaceae Meadow 747 -3177 6992
Poaceae Meadow 942 -2764 3147
Poaceae Meadow 1479 -2634 2895
Poaceae Meadow 1113 -2776 1795
Poaceae Meadow 2215 -2737 7971
Poaceae Meadow 1121 -2944 2934
Poaceae Meadow 638 -3206 1529
108
Macrophytes Macrophytes 886 -3044 2286
Macrophytes Macrophytes 1056 -2720 2673
Macrophytes Macrophytes 769 -3297 1249
Macrophytes Macrophytes 967 -2763 1442
Macrophytes Macrophytes 959 -2670 2105
Macrophytes Macrophytes 334 -2728 1038
Acacia dealbata
Introduced
species 656 -2696 1296
Acacia dealbata
Introduced
species 487 -2941 1782
Acacia dealbata
Introduced
species 220 -2611 3888
Luma apiculata Native species 433 -2542 4135
Luma apiculata Native species 171 -2664 7634
Luma apiculata Native species -001 -2736 6283
Luma apiculata Native species 029 -2764 6425
Azara sp Native species 159 -2868 8408
Azara sp Native species 101 -2606 2885
Baccharis concava Native species 104 -2699 5779
Baccharis concava Native species 265 -2488 4325
Baccharis concava Native species 287 -2562 7802
Baccharis concava Native species 427 -2781 5204
Baccharis linearis Native species 190 -2610 4414
Baccharis linearis Native species 023 -2825 5647
109
Peumus boldus Native species 042 -2969 6327
Peumus boldus Native species 205 -2746 4110
Peumus boldus Native species 183 -2743 6293
Chusquea quila Native species 482 -2801 4275
Poaceae meadow 217 -2629 7214
Lobelia sp Native species 224 -2645 3963
Lobelia sp Native species -091 -2565 4538
Aristotelia chilensis Native species -035 -2785 5247
Aristotelia chilensis Native species -305 -2889 2305
Aristotelia chilensis Native species 093 -2836 5457
Chusquea quila Native species 173 -2754 3534
Chusquea quila Native species 045 -2950 6739
Quillaja saponaria Native species 223 -2838 9385
Scirpus meadow 018 -2820 7115
Sophora sp Native species -184 -2481 2094
Sophora sp Native species -181 -2717 1721
Pinus radiata
Introduced
trees 1581 -2602 3679
Pinus radiata
Introduced
trees 1431 -2784 4852
Pinus radiata
Introduced
trees -091 -2708 9760
Pinus radiata
Introduced
trees 153 -2568 3470
110
Salix sp
Introduced
trees 632 -2878 1921
LITERATURE CITED
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A historical
framework for land cover change in southwestern South America in the past 15000
years Land use policy 27 148ndash160
httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the next
carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014
httpsdoiorg101002eft2235
Blaauw M Christeny JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474
httpsdoiorg10121411-BA618
Bowman DMJS Cook GD 2002 Can stable carbon isotopes (δ13C) in soil
carbon be used to describe the dynamics of Eucalyptus savanna-rainforest
boundaries in the Australian monsoon tropics Austral Ecol
httpsdoiorg101046j1442-9993200201158x
Brahney J Ballantyne AP Turner BL Spaulding SA Otu M Neff JC 2014
Separating the influences of diagenesis productivity and anthropogenic nitrogen
deposition on sedimentary δ15N variations Org Geochem 75 140ndash150
httpsdoiorg101016jorggeochem201407003
111
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409
httpsdoiorg102134jeq20090005
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego R
Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and
environmental change from a high Andean lake Laguna del Maule central Chile
(36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the
Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from
tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Craine JM Elmore AJ Aidar MPM Dawson TE Hobbie EA Kahmen A
Mack MC Kendra K Michelsen A Nardoto GB Pardo LH Pentildeuelas J
Reich B Schuur EAG Stock WD Templer PH Virginia RA Welker JM
Wright IJ 2009 Global patterns of foliar nitrogen isotopes and their relationships
with cliamte mycorrhizal fungi foliar nutrient concentrations and nitrogen
availability New Phytol 183 980ndash992 httpsdoiorg101111j1469-
8137200902917x
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty
Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-
010-9453-1
Diebel M Vander Zanden MJ 2012 Nitrogen stable isotopes in streams stable
isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19
1127ndash1134 httpsdoiorg10189008-03271
112
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in freshwater
wetlands record long-term changes in watershed nitrogen source and land use SO
- Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash
2916
Fariacuteas L Castro-Gonzales M Cornejo M Charpentier J Faundez J
Boontanon N Yoshida N 2009 Denitrification and nitrous oxide cycling within the
upper oxycline of the oxygen minimum zone off the eastern tropical South Pacific
Limnol Oceanogr 54 132ndash144
Farquhar GD Ehleringer JR Hubick KT 1989 Carbon Isotope Discrimination
and Photosynthesis Annu Rev Plant Physiol Plant Mol Biol
httpsdoiorg101146annurevpp40060189002443
Farquhar GD OrsquoLeary MH Berry JA 1982 On the relationship between
carbon isotope discrimination and the intercellular carbon dioxide concentration in
leaves Aust J Plant Physiol httpsdoiorg101071PP9820121
Fogel ML Cifuentes LA 1993 Isotope fractionation during primary production
Org Geochem httpsdoiorg101007978-1-4615-2890-6_3
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A
Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-
resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS)
implications for past sea level and environmental variability J Quat Sci 32 830ndash
844 httpsdoiorg101002jqs2936
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924
httpsdoiorg104319lo20095430917
113
Gerhart LM McLauchlan KK 2014 Reconstructing terrestrial nutrient cycling
using stable nitrogen isotopes in wood Biogeochemistry 120 1ndash21
httpsdoiorg101007s10533-014-9988-8
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and oxygen
isotope fractionation during dissimilatory nitrate reduction by denitrifying bacteria 53
2533ndash2545 httpsdoiorg10230740058342
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater
eutrophic lake Limnol Oceanogr 51 2837ndash2848
httpsdoiorg104319lo20065162837
Harris D Horwaacuteth WR van Kessel C 2001 Acid fumigation of soils to remove
carbonates prior to total organic carbon or CARBON-13 isotopic analysis Soil Sci
Soc Am J 65 1853 httpsdoiorg102136sssaj20011853
Houmlgberg P 1997 Tansley review no 95 natural abundance in soil-plant systems
New Phytol httpsdoiorg101046j1469-8137199700808x
Kylander ME Ampel L Wohlfarth B Veres D 2011 High-resolution X-ray
fluorescence core scanning analysis of Les Echets (France) sedimentary sequence
New insights from chemical proxies J Quat Sci 26 109ndash117
httpsdoiorg101002jqs1438
Lara A Solari ME Prieto MDR Pentildea MP 2012 Reconstruccioacuten de la
cobertura de la vegetacioacuten y uso del suelo hacia 1550 y sus cambios a 2007 en la
ecorregioacuten de los bosques valdivianos lluviosos de Chile (35o - 43o 30acute S) Bosque
(Valdivia) 33 03ndash04 httpsdoiorg104067s0717-92002012000100002
Lehmann M Bernasconi S Barbieri A McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during
114
simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66
3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007
Liu G Li M An L 2007 The Environmental Significance Of Stable Carbon
Isotope Composition Of Modern Plant Leaves In The Northern Tibetan Plateau
China Arctic Antarct Alp Res httpsdoiorg1016571523-0430(07-505)[liu-
g]20co2
Marzecovaacute A Mikomaumlgi A Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54
httpsdoiorg103176eco2011105
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash
1643 httpsdoiorg1011770959683613496289
Meyers PA 2003 Application of organic geochemistry to paleolimnological
reconstruction a summary of examples from the Laurention Great Lakes Org
Geochem 34 261ndash289 httpsdoiorg101016S0146-6380(02)00168-7
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland
Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Odone C 1998 El Pueblo de Indios de Vichuqueacuten siglos XVI y XVII Rev Hist
Indiacutegena 3 19ndash67
Pausch J Kuzyakov Y 2018 Carbon input by roots into the soil Quantification of
rhizodeposition from root to ecosystem scale Glob Chang Biol
httpsdoiorg101111gcb13850
115
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98
httpsdoiorg1011772053019614564785
Sun W Zhang E Jones RT Liu E Shen J 2016 Biogeochemical processes
and response to climate change recorded in the isotopes of lacustrine organic
matter southeastern Qinghai-Tibetan Plateau China Palaeogeogr Palaeoclimatol
Palaeoecol httpsdoiorg101016jpalaeo201604013
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of
different trophic status J Paleolimnol 47 693ndash706
httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl
httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M 2009 High-resolution quantitative climate
reconstruction over the past 1000 years and pollution history derived from lake
sediments in Central Chile Philos Fak PhD 246
Woodward C Chang J Zawadzki A Shulmeister J Haworth R Collecutt S
Jacobsen G 2011 Evidence against early nineteenth century major European
induced environmental impacts by illegal settlers in the New England Tablelands
south eastern Australia Quat Sci Rev 30 3743ndash3747
httpsdoiorg101016jquascirev201110014
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K Yeager
KM 2016 Different responses of sedimentary δ15N to climatic changes and
116
anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau
J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
117
DISCUSION GENERAL
El Nitroacutegeno (N) es un elemento clave que controla la dinaacutemica y
funcionamiento de ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al
1997) Pese a que el N (N2) es muy abundante en la atmoacutesfera (78) es escaso
en los ecosistemas pues la mayoriacutea de la biota no puede acceder a su forma
molecular (Galloway et al 1995 Zaehle et al 2013) En este sentido la entrada
natural de N en los ecosistemas es viacutea fijacioacuten bioloacutegica del N2 por bacterias que lo
convierten en compuestos bioloacutegicamente disponibles (Battye et al 2017) Debido
a que la fijacioacuten es un proceso energeacuteticamente costoso los ecosistemas
comuacutenmente estaacuten limitados por N (Vitousek and Howarth 1991) Los LUCC
contribuyen al incremento del N disponible y son una de las principales causas de
eutroficacioacuten y contaminacioacuten de los cuerpos de agua (Woodward et al 2012)
En Chile central los LUCC principalmente relacionados con las actividades
agriacutecolas y forestales han causado grandes cambios en el ciclo del N debido al
118
reemplazo de la cobertura natural por un monocultivo y al uso de fertilizantes que
modifican los aportes de MO y N a los cuerpos de agua El programa de
estabilizacioacuten de laderas impulsada por el gobierno (Albert 1900) y la ley forestal
de 1974 han fomentado la forestacioacuten con plantaciones de Pinus radiata y
Eucalyptus globulus y en el caso de Lago Vichuqueacuten estas actualmente cubren maacutes
del 60 de su cuenca (Cap2 Fig 5) Ademaacutes en Laguna Matanzas la
sobreexplotacioacuten del agua en conjunto con el cambio climaacutetico actual el que ha
conducido a una de las megasequiacuteas maacutes severas en las uacuteltimas deacutecadas
(Garreaud et al 2017) generoacute una combinacioacuten letal que secoacute el lago en tan solo
10 antildeos (Cap1 Fig1) Las evidencias obtenidas a partir de estos dos lagos
permiten identificar las huellas del Antropoceno en Chile central basadas en el
registro sedimentario lacustre
La transferencia de N desde los ecosistemas terrestres a acuaacuteticos es un
proceso natural con controles bioacuteticos climaacuteticos y geoloacutegicos El enlace
hidroloacutegico entre los ecosistemas terrestres y acuaacuteticos hace que el estado troacutefico
de los lagos esteacute vinculado al de su cuenca (McLauchlan et al 2013) En Chile
central se sabe poco del impacto de los LUCC sobre el ciclo de nutrientes en los
ecosistemas lacustres aunque al menos desde la colonizacioacuten espantildeola se tienen
registros de influencia humana en las cuencas Durante la colonia espantildeola
Laguna Matanzas fue una hacienda ganadera que exportaba productos caacuternicos al
Peruacute (Contreras-Lopez et al 2014) A su vez en el Lago Vichuqueacuten se realizaban
extensas actividades agriacutecolas hasta comienzos del s XX como por ejemplo
cultivos de vid y trigo ademaacutes de extraccioacuten de madera y mineriacutea de oro (Odone
1997) En las uacuteltimas deacutecadas los LUCC han estado vinculados principalmente con
el desarrollo forestal al compaacutes de una estrategia gubernamental de fomentar con
119
incentivos financieros (Ley de Bosques DFL nordm265 de 1931 y DFL 701 de 1974)
esta actividad El incremento de la superficie forestal es especialmente fuerte en
ambos lagos a partir de la deacutecada de 1980 y hasta 2016 (Laguna de Matanzas 0-
17 Lago Vichuqueacuten 1-66) Este aumento habriacutea sido en detrimento del bosque
nativo y los pastizales (Cap 1 Fig 5 y Cap 2 Fig 8) El incremento de la superficie
forestal generariacutea una disminucioacuten de los aportes de sedimentos y nutrientes al lago
y en este sentido un cambio de estado en los flujos de N (e g tipping points) que
a su vez ha quedado registrado en la sentildeal isotoacutepica (δ15N) y la acumulacioacuten de
MO en los sedimentos lacustres
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central
Los sedimentos lacustres son verdaderos ldquosumiderosrdquo que pueden llegar a
registrar lo que ocurre en cuanto a la historia ambiental de su cuenca En esta tesis
se utilizan los isoacutetopos estables de N (15N y 14N) en sedimentos lacustres para
reconstruir los cambios en la disponibilidad de N en el tiempo y establecer la
magnitud de impacto generado por actividades humanas El fraccionamiento
cineacutetico en los lagos es mediado por procesos bioloacutegicos que mediante la
asimilacioacuten de N genera un producto (eg fitoplancton) maacutes ldquoligerordquo (valores maacutes
bajos de 15N) que su remanente (eg DIN) (Fogel y Cifuentes 1993) Sin embargo
en periacuteodos en que los lagos estaacuten limitados por N el fitoplancton discrimina menos
y la MO resultante queda enriquecida en 15N (Fig 1 a y b) Ademaacutes la
desnitrificacioacuten acentuada en lagos de fondo anoacutexicos (eg Laguna Matanzas
entre 1800 y 1940 Cap1 Fig 6) conduce a un enriquecimiento en 15N de los
sedimentos y de la columna de agua Esta informacioacuten queda reflejada en la MO
120
de los sedimentos lacustres lo que permite conocer la disponibilidad del N en el
tiempo a partir de las variaciones de 15N
En esta tesis analizamos la MO de los sedimentos lacustres para reconstruir
la influencia de los LUCC en los aportes de sedimentos y N al lago En teacuterminos de
asimilacioacuten de N se puede distinguir entre dos grupos principales de productores
primarios que componen el POM (Fig1)
1 Los que asimilan el DIN compuesto por amonio nitratos y nitritos Aquiacute el
δ15N resultante fluctuaraacute en torno a valores maacutes positivos en la medida que
la productividad se incrementa o no hay reposicioacuten del DIN (Fig 1)
2 Los fijadores de N Dado que es un proceso energeacuteticamente costoso en
ambientes que no estaacuten limitados por N muchas veces son excluiacutedas
competitivamente por el resto del fitoplancton Si el DIN queda agotado por
el incremento de la productividad (o bien si el ambiente estaacute limitado ya sea
por N o por otros nutrientes) los fijadores de N pueden florecer y la MO que
se genera a partir de ello tendraacute un δ15N con valores en torno a 0permil
De este modo la MO en los sedimentos lacustres dependeraacute de la
composicioacuten de productores primarios (ie fijadores vs asimiladores del DIN)
ademaacutes de los aportes desde la cuenca tanto de MO como del N de la cuenca que
pasa a formar parte del DIN (eg fertilizantes estieacutercol) (Fig 1)
La MO de los lagos estudiados en esta tesis ha sido analizada a partir de
variables geoquiacutemicas isotoacutepicas frotis y estaacute compuesta principalmente por
diatomeas y MO amorfa A partir de los datos obtenidos en esta tesis los valores
de δ15N aumentan cuando hay una mayor cobertura agriacutecola o pastizales A su vez
tiende a disminuir cuando aumenta la cobertura arboacuterea independiente si esta es
por plantaciones forestales o por bosque nativo
121
Por otra parte la dinaacutemica del lago tambieacuten imprime caracteriacutesticas
especiales en el POM observaacutendose variaciones estacionales en los valores
δ15NPOM Durante la estacioacuten caacutelida y seca se observan valores maacutes bajos que
durante la estacioacuten friacutea y huacutemeda posiblemente relacionado con el incremento de
la contribucioacuten de especies fijadoras de N (Lago Vichuqueacuten Fig 9 Cap 2) durante
el verano En la estacioacuten friacutea y huacutemeda el DIN seriacutea la principal fuente de N Las
mayores entradas de MO y N terrestre debidos a un incremento del lavado de la
cuenca como consecuencia de las precipitaciones y la mineralizacioacuten de la MO
podriacutean estimular la produccioacuten de MO lacustre por crecimiento del fitoplancton
Como consecuencia se observan tendencias decrecientes de los valores de
δ15N en la MO sedimentaria cuando la productividad en el lago estaacute relacionada
con especies fijadoras de N mientras que el δ15N muestra aumentos cuando la
productividad del lago estaacute asociada principalmente al consumo del DIN pero
tambieacuten cuando predominan procesos de perdida de N (eg desnitrificacioacuten Fig
1)
Hemos identificado tres fases en la dinaacutemica del δ15N para ambos lagos
Entre 1800 y 1940 CE la cuenca de laguna de Matanzas estaba dominada por
actividades agriacutecolas y ganaderas (δ15Nsuelo gt 4permil Cap 1 Fig 4 y 6) Las entradas
de sedimentos y MO desde la cuenca son altas (δ13C lt -209permil ZrTi ~029 AlTi
~013) y coincide con un clima maacutes huacutemedo y friacuteo (Von Gunten et al 2009
Christiansen et al 2011) El lago teniacutea un nivel de agua maacutes profundo dominado
por condiciones anoacutexicas (MnFe ~0013) que contribuiriacutean a mayores valores de
δ15N en la columna de agua y por ende de los sedimentos acumulados en el fondo
debido a la peacuterdida diferencial de 14N viacutea desnitrificacioacuten (Fig 7 Cap1) La baja
122
produccioacuten de MO lacustre (BrTi ~002 TOC~17) coincide con un bajo contenido
de NT (~478 μg) y valores maacutes positivos de δ15N (gt 4permil)
La actividad agriacutecola tambieacuten dominaba en la cuenca del Lago Vichuqueacuten
durante este periacuteodo (δ15Nsuelo gt4permil Cap2 Fig 6) El aporte sedimentario desde la
cuenca es alto (AlTi ~029 ZrTi ~032) y fue favorecido por mayores
precipitaciones a las actuales (Christiansen et al 2011) El Lago Vichuqueacuten es un
lago estratificado anoacutexico (MnFe ~002) y profundo (~30 m) por lo que la
desnitrificacioacuten juega un rol importante en el enriquecimiento de la MO
sedimentaria La baja produccioacuten de MO (BrTi ~007 TOC ~12) de origen
lacustre (δ13C ~ -268permil) coincide con un bajo contenido de NT (~309μg +6) y
valores maacutes positivos de δ15N (56permil +03)
Durante esta fase en ambos lagos los aportes de N de la cuenca parecen
ser los principales responsables en las fluctuaciones del δ15N Esta sentildeal podriacutea
estar amplificada por bajas temperaturas (que afectan la productividad lacustre) y
altas precipitaciones que favorecen el lavado de la cuenca y aumentan el aporte de
sedimentos y MO desde la cuenca predominantemente agriacutecola
Entre 1940 y 1980 CE en Laguna Matanzas se observa un incremento en
la acumulacioacuten de MO algal (TOC ~2 +1 δ13C ~-230+09permil) El ambiente
deposicional es ligeramente menos turbulento que el anterior (AlTi ~011+001
ZrTi ~03+001) y el lago estaacute en una fase de transicioacuten hacia condiciones maacutes
oacutexicas (MnFe ~002+0004) y maacutes productivas (BrTi ~004+0007) A su vez δ15N
tiende hacia valores maacutes negativos (δ15N ~30permil+03) mientras que el NT aumenta
oscilando en antifase con el δ15N
En Lago Vichuqueacuten en cambio se observa un ligero incremento en la
acumulacioacuten de MO (TOC ~13 +02) de origen terrestre (δ13C ~-27permil +04) La
123
productividad es similar al periacuteodo anterior (BrTi ~007+003) y el ambiente
deposicional es menos turbulento (AlTi ~02 +009 Zrti ~033 +007) El δ15N y el
NT oscilan en torno a valores ligeramente maacutes bajos (δ15N ~51permil +04 NT ~292μg
+6) Este periacuteodo se caracteriza por ser maacutes caacutelido que el anterior lo que
posibilitariacutea un incremento en la productividad lacustre en Laguna Matanzas pero
que no es observada en el Lago Vichuqueacuten
Entre 1980 y la actualidad en Laguna Matanzas habriacutea un incremento de la
acumulacioacuten de MO (TOC ~59 + 3) asociada a un incremento en la productividad
del lago (BrTi ~009+005) y a los aportes de MO terrestres (δ13C ~-24 + 2permil) El
lago se vuelve maacutes oacutexico (Mn ~0003 +0006) y las entradas de sedimento
disminuyen (AlTi~ 009 +002) El δ15N tiende a valores maacutes negativos (δ15N ~13permil
+1) y el NT aumenta de 20 a 55 μg (NT ~346 + 9 μg) tomando valores sin
precedentes en todo el registro En los sedimentos lacustres del Lago Vichuqueacuten
tambieacuten se observa un incremento de la acumulacioacuten de MO (TOC ~17 + 05)
asociada a un incremento en la productividad del lago (BrTi ~01 + 003) y las
entradas de sedimento desde la cuenca disminuyen (AlTi ~005 + 005) El δ15N
(~43 + 05permil) tiende a valores maacutes negativos y el NT incrementa de 20 a 55 μg (NT
~346 + 9 μg) oscilando en antifase
Durante esta fase en ambos lagos se observa un aumento en la
acumulacioacuten de MO sedimentaria un incremento en el NT y valores maacutes negativos
de δ15N que coincide con el incremento de la superficie forestal de las cuencas
(Laguna Matanzas gt17 Lago Vichuqueacuten gt60)
124
Figura 2 Comparacioacuten de los cambios del ciclo del N entre Laguna Matanzas y
Lago Vichuqueacuten con los procesos biogeoquiacutemicos contrastantes en rojo En L
Matanzas las condiciones oacutexicas podriacutean estar favoreciendo la nitrificacioacuten del
amonio En Vichuqueacuten las condiciones anoacutexicas favorecen un aumento en δ15N de
la MO en los sedimentos por desnitrificacioacuten y mineralizacioacuten
Los ambientes mediterraacuteneos en el que los lagos del presente estudio se
encuentran insertos estaacuten caracterizados por una estacioacuten seca prolongada y las
precipitaciones ocurren en eventos puntuales alcanzando altos montos
pluviomeacutetricos en poco tiempo Las lluvias favorecen un lavado de la cuenca la
perdida de N y otros nutrientes Por lo que es esperable que los sedimentos del
lago reflejen mejor los valores isotoacutepicos de los suelos de la cuenca durante los
periacuteodos con altos montos pluviomeacutetricos En el Lago Vichuqueacuten por ejemplo el
125
POM superficial (2 y 5 m de profundidad) despliega valores isotoacutepicos maacutes
positivos en invierno presumiblemente como resultado de mayores aportes de MO
y N de su cuenca (Fig 1b) En ambos lagos los valores maacutes positivos en los
sedimentos lacustres ocurren durante periacuteodos con mayores montos pluviomeacutetricos
(Cap1 Fig 6 y Cap 2 Fig 12)
Ademaacutes del efecto de la precipitacioacuten en el aporte de nutrientes al lago en
esta tesis hemos encontrado que los mayores aportes de N desde la cuenca se dan
cuando la cobertura vegetal del suelo es menor En Laguna Matanzas el desarrollo
de la actividad ganadera desde la llegada de los espantildeoles a la cuenca habriacutea
incrementado los aportes de N al lago Los valores de δ15N en los sedimentos
lacustres depositados durante este periacuteodo son los maacutes positivos de todo el registro
(~4permil) A su vez en Lago Vichuqueacuten los valores isotoacutepicos maacutes positivos se
registran entre 1300 y 1900 CE que coinciden con el mayor desarrollo de
actividades agriacutecola y de extraccioacuten de maderas de la cuenca (Fig 10 Cap 2)
Los recientes LUCC (a partir de 1975) en las cuencas de Laguna Matanzas
y Lago Vichuqueacuten estaacuten caracterizados por un incremento de la superficie forestal
y agriacutecola en reemplazo de la cobertura de pastizal (Laguna Matanzas) y bosque
nativo (Lago Vichuqueacuten) Los valores de δ15N en los testigos sedimentarios fueron
maacutes positivos cuanto mayor la superficie agriacutecola de la cuenca y maacutes negativos
cuanto mayor la superficie forestal Aunque por los alcances de esta tesis no
podemos ser concluyentes respecto del origen del NT (aloacutectonoautoacutectono) parece
ser que esta maacutes ligado a los aportes de la cuenca pese a la disminucioacuten del aporte
sedimentario observado en ambos lagos Las plantaciones forestales a diferencia
del bosque nativo son fertilizadas con una mezcla de N P y K (Toral et al 2005)
Por lo que pese a la disminucioacuten del aporte sedimentario la concentracioacuten de
126
nutrientes en el aporte al lago puede verse incrementada por el tipo de manejo
forestal con respecto al bosque nativo
Los resultados del primer capiacutetulo demuestran que 1) las plantaciones
forestales yo bosque nativo retiene maacutes sedimentos y MO mientras que el uso de
suelo agriacutecola y los pastizales destinados al desarrollo de la actividad ganadera lo
libera 2) Al parecer la desnitrificacioacuten es el proceso natural maacutes importante de
perdida de N en lagos costeros y favorece el enriquecimiento de 15N tanto en la
columna de agua (ie 15N-DIN) como en los sedimentos una vez enterrados y 3) la
desnitrificacioacuten depende de los cambios en las condiciones REDOX en el lago La
oxigenacioacuten en lagos poco profundos -como Laguna Matanzas- es altamente
fluctuante a escalas decadal-centenal depende de la profundidad de la laacutemina de
agua y de la variabilidad de la precipitacioacuten En este sentido en Laguna Matanzas
habriacutea condiciones anoacutexicas en ambientes cuando el lago alcanza niveles maacutes
altos Estas condiciones se observan entre 1820 y 1940 CE y son coetaacuteneas con
episodios maacutes huacutemedos y friacuteos (von Gunten et al 2009 Christie et al 2009) pero
tambieacuten con una fuerte actividad ganadera en la cuenca
Las fuentes variables de N y su fluctuacioacuten en el tiempo hacen necesario
contar con un anaacutelogo moderno que permite usar al δ15N de los sedimentos
lacustres como un indicador indirecto de los cambios en la disponibilidad de N en
el tiempo Por ello en el segundo capiacutetulo integramos muestras de suelo-
vegetacioacuten y un monitoreo de POM de la columna de agua en verano e invierno La
composicioacuten isotoacutepica de POM estariacutea reflejando cambios de la fuente de N (fijacioacuten
vs consumo del DIN) que variacutea durante el antildeo hidroloacutegico Durante el verano la
mayor temperatura y horas-luz favoreceriacutean las entradas de N al lago viacutea fijacioacuten
bioloacutegica de N2 atmosfeacuterico resultando en un incremento de la MO con un valor
127
isotoacutepico bajo (POM verano~5permil Fig 1b) El N2 (δ15N = 0permil) es fijado praacutecticamente
sin fraccionamiento isotoacutepico generando un δ15N de la OM cercano a 0permil (Leng et
al 2006) En invierno las precipitaciones pueden estar favoreciendo un incremento
en la entrada de nutrientes al lago El DIN de la columna de agua llega a contener
valores de δ15N maacutes altos reflejando estos mayores aportes de la cuenca y el POM
del Lago Vichuqueacuten queda enriquecido en 15N (δ15N ~11permil Cap 2 Fig 9) Estas
variaciones estacionales pueden estar evidenciando un cambio en la composicioacuten
de especies de POM desde especies fijadoras a especies que consumen el N de la
columna de agua Sin embargo para corroborar esta informacioacuten seriacutea deseable
contar con el anaacutelisis de la composicioacuten de especies de las muestras de agua
extraidas
Entre los resultados maacutes importantes estaacuten los valores isotoacutepicos del suelo
y la biomasa representativa de la cuenca que incluye un listado de las especies
nativas de la zona que hasta ahora no se contaba (Cap 2 Tabla en material
suplementario) Las especies colectadas despliegan valores de δ15Nfoliar maacutes
positivos (plantaciones forestales y pastizal ~89permil) que el suelo (35permil) salvo por
las especies nativas las que oscilan en torno a valores cercanos al atmosfeacuterico
(~14permil) La razoacuten isotoacutepica 15N14N refleja la dinaacutemica de intercambio de N entre la
vegetacioacuten y el suelo (Koba et al 2002) La discriminacioacuten es positiva en la mayoriacutea
de los sistemas bioloacutegicos por lo que las plantas tienen valores de δ15N maacutes bajos
que el suelo (Evans et al 2001) como sucede con las especies nativas en Lago
Vichuqueacuten En consecuencia las plantas quedan empobrecidas en 15N y conducen
a un enriquecimiento en 15N del suelo sobretodo si no hay reposicioacuten de N por otras
viacuteas (eg fijacioacuten de N) Por lo tanto los valores maacutes negativos de δsup1⁵Nfoliar de las
especies nativas pueden estar relacionados con el consumo preferencial de 14N del
128
suelo donde la discriminacioacuten isotoacutepica de la vegetacioacuten contra el 15N conduce a
valores del δsup1⁵Nsuelo maacutes altos (Houmlgberg 1997) Por el contrario los valores maacutes
positivos de δsup1⁵Nfoliar de las plantaciones forestales y pastos con respecto al suelo
puede deberse por una parte que el suelo no cuenta con mecanismos naturales de
reposicioacuten de N salvo por la descomposicioacuten de la hojarasca que puede ser maacutes
lenta en Pinus radiata que en bosque nativo (Lusk et al 2001) y por otra por el alto
impacto de los aportes de N (y otros nutrientes) derivado de las actividades
humanas (eg uso de fertilizantes) en el suelo
El alcance maacutes significativo de esta tesis se relaciona con un cambio en la
tendencia maacutes negativa de δsup1⁵N y el incremento del NT a partir de 1940 y a partir
de 1970 CE en ambos lagos La causa maacutes probable de esta tendencia es el
reemplazo de la cobertura natural o preexistente a 1940 CE por plantaciones
forestales
En la figura 2 se observa una siacutentesis de los principales procesos que
afectan la sentildeal isotoacutepica de la MO en los sedimentos lacustres de L Matanzas y
L Vichuqueacuten La entrada de N y MO desde la cuenca depende del tipo de cobertura
Las plantaciones forestales y el bosque nativo retienen maacutes nutrientes y sedimentos
en la cuenca mientras que la agricultura los favorece Sin embargo las praacutecticas
de manejo que incluyen la aplicacioacuten de N (y otros nutrientes) pueden aportar maacutes
nutrientes al lago que la cobertra de bosque nativo Cuando las actividades
forestales cubren una mayor superficie de la cuenca (1940 en adelante) δsup1⁵N oscila
en valores maacutes negativos y el NT tiende a incrementarse en los sedimentos
lacustres Cabe destacar que los valores de δsup1⁵N observados en los testigos
sedimentarios a fines del s XX no tienen precedente durante la conquista y colonia
espantildeola o durante el resto del periodo de la Repuacuteblica
129
Figura 2 Modelo de transferencia de N entre los ecosistemas terrestres y
acuaacuteticos El δ15N de los sedimentos lacustres depende de la interaccioacuten entre los
aportes de la cuenca y porcesos ldquoin-lakerdquo Flechas indican la direccioacuten del flujo de
N En azul las salidasentradas de 14N en rojo las salidasentradas de 15N δ15N de
la interaccioacuten plantas-suelo depende del tipo de cobertura de suelo
130
CONCLUSIONES GENERALES
La transferencia de N entre cuencas y lagos es un factor de control del ciclo
del N en los lagos y queda reflejado en la sentildeal isotoacutepica de los sedimentos
lacustres (Leng et al 2006 McLauchlan et al 2007) En Laguna Matanzas el
suelo de las especies nativas y las plantaciones forestales despliegan valores de
δ15N maacutes bajos (~1permil) que los de Lago Vichuqueacuten (~35permil) Coincidentemente los
sedimentos lacustres de Laguna Matanzas oscilan con valores de δ15N maacutes bajos
(-15 a +45permil Fig 1) que los de Lago Vichuqueacuten (+3 a +8permil)
Por otra parte el estudio de la MO de los sedimentos lacustres ha permitido
reconstruir los cambios en los aportes de MO y N desde la cuenca Aunque no es
posible afirmar queacute tipo de N estaacute entrando al lago (eg Nitroacutegeno orgaacutenico e
inorganico) si podemos decir que los cambios en las fluctuaciones de δ15N son
coincentes con el crecimiento de la actividad forestal (1980 - hasta 2016 L
Matanzas 0-17 y L Vichuqueacuten 1-66) El δ15N fluctuacutea hacia valores maacutes
negativos Ademaacutes se observa un incremento del NT en los sedimentos lacustres
cuanto mayor es la superficie forestal
Entre 1800-1940 las cuencas eran fundamentalmente agriacutecolas y
ganaderas (Cap1 Fig 7 y Cap 2 Fig 12) el δ15N de los sedimentos lacustres
oscila con valores maacutes positivos (L Matanzas +40 plusmn 05permil y L Vichuqueacuten 56 plusmn
033permil Fig 1) hay una baja bioproductividad en el lago (BrTi ~002 TOC~17)
lo que coincide con un bajo contenido de NT (~478 μg) Este es un periacuteodo de altas
precipitaciones (Christie et al 2009) que tienden a favorecer el lavado de la cuenca
131
y con ello el aporte de MO sedimentos y nutrientes desde la cuenca tiende a verse
favorecido Aunque las principales actividades humanas en estas cuencas son
diferentes (ganaderiacutea en Laguna Matanzas Contreras-Lopez et al 2014
agricultura en Lago Vichuqueacuten Odone 1997) en ambos casos hubo un reemplazo
de la cobertura natural de bosques nativo que favorecioacute mayores aportes de N y
sedimentos desde la cuenca en un efecto sumado con el aumento de las
precipitaciones
A partir de 1980 CE en ambos lagos se observa una disminucioacuten en los
valores de δ15N y un incremento del NT que no tiene precedentes en todo el registro
y pese a que ambos lagos son limnologicamente muy diferentes En Lago
Vichuqueacuten el δ15N tiende a oscilar en antifase con el NT (Fig 1) En el caso de
Laguna Matanzas se observa una tendencia hacia valores maacutes negativos y a partir
de 1990s a valores maacutes positivos mientras que el NT tiende a incrementarse de
manera sostenida Ello podriacutea estar relacionado con el crecimiento de la actividad
forestal habriacutea una mayor retencioacuten de N y sedimentos en la cuenca ligado al
incremento de la superficie forestal Ademaacutes en el caso de L Matanzas el
incremento de la actividad agriacutecola (sumado al de las plantaciones forestales)
podriacutea expicar el cambio en la tendencia en la deacutecada de los 1990s
En el contexto de Antropoceno esta tesis nos permite identificar un gran
impacto de las actividades humanas en Chile central a partir de la deacutecada de 1940
y una Gran Aceleracoacuten a partir de la deacutecada de 1970s En el registro sedimentario
de los lagos en estudio se observa una gran acumulacioacuten de MO el δ15N oscila
hacia valores maacutes negativos y un gran incremento del NT y el crecimiento de la
actividad forestal parece ser la principal responsable Ademaacutes la sobreexplotacioacuten
132
del agua junto al cambio climaacutetico global ha generado una combinacioacuten letal para
los lagos costeros de Chile central
Figura 1 Comparacioacuten de las fluctuaciones de δ15N durante los uacuteltimos 300
antildeos en Laguna Matanzas y Lago Vichuqueacuten
133
Referencias
Battye W Aneja VP Schlesinger WH 2017 Earthrsquos Future Is nitrogen the next carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia
UC de V 2014 Elementos de la historia natural del An Mus Hist Natulas Vaplaraiso 27 51ndash67
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Evans RD Evans RD 2001 Physiological mechanisms influencing
plant nitrogen isotope composition 6 121ndash126 Fogel ML Cifuentes LA 1993 Isotope fractionation during primary
production Org Geochem 73ndash98 httpsdoiorg101007978-1-4615-2890-6_3 Galloway JN Schlesinger WH Ii HL Schnoor L Tg N 1995
Nitrogen fixation Anthropogenic enhancement-environmental response reactive N Global Biogeochem Cycles 9 235ndash252
Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J
Christie D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Koba K Koyama LA Kohzu A Nadelhoffer KJ 2003 Natural 15 N
Abundance of Plants Coniferous Forest httpsdoiorg101007s10021-002-0132-6
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos
Transactions American Geophysical Union httpsdoiorg1010292007EO070007 McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE
2007 Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
134
Phytologist N Oct N 2006 Tansley Review No 95 15N Natural Abundance in Soil-Plant Systems Peter Hogberg 137 179ndash203
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Turner BL Kasperson RE Matson PA McCarthy JJ Corell RW
Christensen L Eckley N Kasperson JX Luers A Martello ML Polsky C Pulsipher A Schiller A 2003 A framework for vulnerability analysis in sustainability science Proc Natl Acad Sci U S A httpsdoiorg101073pnas1231335100
Vitousek PM Aber JD Howarth RW Likens GE Matson PA
Schindler DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
Vitousek PM Howarth RW 1991 Nitrogen limitation on land and in the
sea How can it occur Biogeochemistry httpsdoiorg101007BF00002772 von Gunten L Grosjean M Rein B Urrutia R Appleby P 2009 A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 Holocene 19 873ndash881 httpsdoiorg1011770959683609336573
Xu R Tian H Pan S Dangal SRS Chen J Chang J Lu Y Maria
Skiba U Tubiello FN Zhang B 2019 Increased nitrogen enrichment and shifted patterns in the worldrsquos grassland 1860-2016 Earth Syst Sci Data httpsdoiorg105194essd-11-175-2019
Zaehle S 2013 Terrestrial nitrogen-carbon cycle interactions at the global
scale Philos Trans R Soc B Biol Sci 368 httpsdoiorg101098rstb20130125
10
anaacutelisis sedimentoloacutegicos geoquiacutemicos e isotoacutepicos (δ15N δ13C) de los sedimentos
lacustres columna de agua y suelo-vegetacioacuten de la cuenca y la reconstruccioacuten de
los LUCC entre 1975 y 2016 a partir de imaacutegenes satelitales El objetivo de esta
tesis fue evaluar el rol de LUCC como principal control del ciclo del N en el sistema
cuenca-lago durante las uacuteltimas centurias en Chile central Nuestros principales
resultados muestran que valores maacutes positivos de δ15N en los sedimentos lacustres
estaacuten relacionados con grandes aportes de N de la cuenca que a su vez son
mayores cuanto mayor la superficie agriacutecola yo los pastizales mientras que tanto
las plantaciones forestales como los bosques favorecen la retencioacuten de nutrientes
en la cuenca (lo que se ve reflejado en un δ15N maacutes negativo) Esta tesis plantea
un cambio de estado en la dinaacutemica del N asociado a la introduccioacuten y expansioacuten
de las plantaciones forestales o bosques los que retienen el Nitroacutegeno en las
cuencas mientras que el clima juega un rol secundario
11
ABSTRACT
The Anthropocene is characterized by human disturbances at the global
scale For example changes in land use are known to disturb the N cycle since the
industrial revolution but especially since the Great Acceleration (1950 CE) onwards
This impact has changed N availability in both terrestrial and aquatic ecosystems
However there are some important uncertainties associated with the extent of this
impact and how it is coupled to ongoing climate change (ie megadroughts rainfall
variability and shifting stationarity) or to other nutrient cycles (e g carbon cycle)
Lake sediments contain paleoenvironmental information regarding the conditions of
the watershed and associated lakes and which the respective sediments are
deposited Nitrogen stable isotope analysis (δ15N) on lake sediments allows us to
reconstruct the changes in N availability through time Here we used a multiproxy
approach that uses sedimentological geochemical and isotopic analyses on
lacustrine sediments water column and soilvegetation from the watershed as well
12
as land use and cover change (LUCC) from 1975 to 2016 obtained from satellite
images The goal of this thesis is to evaluate the role of LUCC as the main driver for
N cycling in a coastal watershed system of central Chile over the last centuries Our
main results show that more positive δ15N values in lake sediments are related to
higher N contributions from the watershed which in turn increase with increased
agricultural andor pasture cover whereas either forest plantations or native forests
can favor nutrient retention in the watershed (δ15N more negative) This thesis
proposes that N dynamics are mainly driven by the introduction and expansion of
forest or tree plantations that retain nitrogen in the watershed whereas climate plays
a secondary role
13
INTRODUCCIOacuteN
El N es un elemento esencial para la vida y limita la productividad en
ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al 1997) Las actividades
humanas han tenido un profundo impacto sobre el ciclo del N global principalmente
a partir la revolucioacuten industrial (IPCC 2013 2007) Las concentraciones de N se
han incrementado por la fijacioacuten industrial de N atmosfeacuterico (viacutea el proceso Haber-
Bosch Erisman et al 2008) por el uso de fertilizantes sobre la base de N para
mejorar el rendimiento de los cultivos la produccioacuten ganadera y ademaacutes de los
cambios en el uso del suelo (abreviado en ingleacutes- LUCC) (Robertson y Vitousek
2009 Galloway et al 2008 Battye et al 2017 Xu et al 2019) Estas actividades
contribuyen significativamente al incremento de la disponibilidad bioloacutegica de N
cuyas consecuencias para los ecosistemas incluye la perdida de diversidad
modificacioacuten de la disponibilidad de nutrientes y acidificacioacuten de los suelos entre
otras Sin embargo se desconoce la cantidad destino y el alcance que ha tenido
14
el N movilizado entre los ecosistemas generado por la influencia de las actividades
humanas (Vitousek et al 1997)
La entrada natural de N en los ecosistemas terrestres y acuaacuteticos es viacutea
fijacioacuten atmosfeacuterica la que es llevada a cabo por microorganismos muchos de ellos
en relaciones simbioacuteticas (eg Rhizobium en leguminosas) y las algas (Vitousek et
al 1997 Battye et al 2017) Las principales salidas estaacuten relacionadas con la
desnitrificacioacuten volatilizacioacuten del amoniaco y por escurrimiento superficial y
subterraacuteneo (Galloway et al 2008 Diacuteaz et al 2016) En el caso de los sistemas
lacustres ademaacutes estaacuten la resuspencioacuten de N del fondo del lago los aportes desde
la cuenca (entradas de N) La depositacioacuten de MO en el fondo del lago que marca
la salida de N de la columna de agua Estas relaciones de intercambio de N tienen
un control geograacutefico (eg latitud exposicioacuten relieve etc) climaacutetico y biotico
(McLauchlan et al 2013) La alteracioacuten provocada por los LUCC no solo genera
las peacuterdidas de nutrientes del suelo por erosioacuten y escurrimiento superficial sino que
tambieacuten modifica la disponibilidad y transferencia de N entre los ecosistemas
terrestres y acuaacuteticos (Elbert et al 2012 McLauchlan et al 2013) Por ejemplo el
reemplazo de la vegetacioacuten nativa por la agricultura yo plantaciones forestales
altera el ciclo del N debido al despeje de los terrenos viacutea quema de la vegetacioacuten
pero tambieacuten debido al reemplazo de la vegetacioacuten preexistente por un
monocultivo (eg trigo pino) y el uso ineficiente de fertilizantes para mejorar el
rendimiento agriacutecola Estas actividades favorecen un incremento de los aportes de
N (y otros nutrientes) viacutea sistemas de drenajes a lagos los que actuacutean como
sumideros El incremento del N derivado de las actividades humanas tanto en los
ecosistemas terrestres como acuaacuteticos genera incertezas relacionados con la
trayectoria y la magnitud de la disponibilidad de N en ambos ecosistemas (Batye et
15
al 2017 Mclauchlan et al 2017) Por lo tanto se requiere de una liacutenea base de
largo plazo para saber coacutemo estas perturbaciones impactan la disponibilidad de N
en las uacuteltimas deacutecadassiglos en los ecosistemas lacustres y por lo tanto el alcance
real que los LUCC han tenido en el ciclo del N
Los ecosistemas mediterraacuteneos y el ciclo del N
Los ecosistemas mediterraacuteneos son especialmente sensibles a los LUCC
pues estos se encuentran sometidos a un estreacutes hiacutedrico durante largas temporadas
estivales y las precipitaciones se concentran en eventos puntuales y a veces con
altos montos pluviomeacutetricos Las precipitaciones tienen un alto potencial de arrastre
de sedimentos y nutrientes los que actuacutean como fertilizantes al llegar a los
ecosistemas lacustres A su vez un incremento en la disponibilidad de N puede
generar un desbalance con otros nutrientes (eg Fosforo) con efectos sobre la
productividad y diversidad de los lagos (Pentildeuelas et al 2012 Sardans et al 2012
McLauchlan 2013) En este sentido en Chile central se observa una disminucioacuten
de las precipitaciones especialmente fuerte en las uacuteltimas deacutecadas por lo que se ha
denoacuteminado como Megasequiacutea (Garreaud et al 2017) y el efecto de las
precipitaciones esporaacutedicas pueden generar un arrastre de nutrientes que no ha
sido evaluado
Los ecosistemas mediterraacuteneos concentran el 20 de la biodiversidad global
(Myers et al 2000) pero existe una escasez de conocimiento respecto a los
efectos del incremento de N en los cuerpos de agua como consecuencia de las
actividades humanas en el pasado histoacuterico (Ochoa-Hueso et al 2011) y coacutemo la
disponibilidad de N ha variado en el tiempo Los LUCC han aumentado el flujo de
N en las cuencas hidrograacuteficas viacutea escurrimiento superficial y lixiviacioacuten
16
favoreciendo la peacuterdida de sedimentos y nutrientes del suelo en el mundo entero
(Elser et al 2011 McLauchlan et al 2013 Xu et al 2017 y 2019) Ello ha
contribuido a la acidificacioacuten y eutrofizacioacuten generalizada de riacuteos y lagos
(McLauchlan et al 2013 Schindler et al 2008)
El ecosistema mediterraacuteneo de Chile central es una regioacuten ampliamente
intervenida desde al menos la colonizacioacuten espantildeola (1600-1800 CE) Los LUCC
han tenido efectos negativos en la disponibilidad de agua especialmente
observados con el desarrollo de las actividades minera (Prieto et al 2019) Aunque
se sabe de los impactos en los ecosistemas de bosque en teacuterminos de cobertura
debidos al desarrollo silvo-agropecuarias (Armesto et al 2010) se desconoce el
impacto de estas actividades sobre el ciclo del N en los cuerpos de agua o en queacute
momento cobran fuerza suficiente para alterar el flujo de N hacia los lagos de Chile
Central Por lo tanto en esta tesis nos centramos en entender coacutemo los LUCC han
afectado la disponibilidad de N en los lagos costeros de Laguna Matanzas y Lago
Vichuqueacuten desde la llegada de los espantildeoles a Chile hasta el presente
Los lagos como sensores ambientales
Los sedimentos lacustres son buenos sensores de cambios en los aportes
de nutrientes contaminacioacuten y eutroficacioacuten de los cuerpos de agua ya que son
capaces de preservar sentildeales quiacutemicas y fiacutesicas al momento de su formacioacuten y
ademaacutes con alta resolucioacuten y a largo plazo (Leng et al 2006) Por lo tanto
constituyen una herramienta uacutetil para reconstruir el flujo de N entre ecosistemas
terrestres y acuaacuteticos en el pasado sus consecuencias (eg incremento de la
productividad lacustre) y el rol del clima y los LUCC en el pasado (McLauchlan et
al 2013 von Gunten et al 2009) Por ejemplo el cultivo de maiacutez por parte de los
17
nativos iroqueses en Canadaacute provocoacute la eutrofizacioacuten del Lago Crawford (43degN)
durante los siglos XII y XV Entre las consecuencias registradas se documentoacute un
claro incremento de la productividad primaria y cambios en la estructura comunitaria
de diatomeas foacutesiles (incremento de Stephanodiscus complex y disminucioacuten de
Cyclotella bodanica en Ekdahl et al 2007) Otro ejemplo demuestra como las
actividades agriacutecolas en la cuenca del riacuteo Mississippi modificaron los patrones de
sedimentacioacuten y aporte de nutrientes al lago Pepin EE UU (44degN) a partir del
asentamiento europeo en 1840 CE y principalmente desde 1940 CE (Engstrom et
al 2009) Para Chile von Gunten et al (2009) a partir de indicadores
limnogeoloacutegicos evidencioacute la contaminacioacuten y eutroficacioacuten de cinco lagos cercanos
a la ciudad de Santiago (33ordmS) como consecuencia de la deposicioacuten atmosfeacuterica
de metales pesados (especialmente Cu) quema de combustible foacutesil y flujo de
nutrientes durante los uacuteltimos 200 antildeos
Caracteriacutesticas limnoloacutegicas de los lagos
Los procesos limnoloacutegicos afectan la distribucioacuten y abundancia de los
organismos en los lagos Estaacuten influenciados por forzamientos externos por
ejemplo la precipitacioacuten o la penetracioacuten de la luz y la morfometriacutea del lago En este
sentido el nivel del lago dependeraacute del balance entre las entradas y salidas de agua
(precipitaciones escurrimiento superficial y subterraacuteneo y la evaporacioacuten) y la forma
de la cuenca (profundidad pendiente aacuterea del espejo de agua)
En funcioacuten de la profundidad de penetracioacuten de la luz es posible diferenciar
dos aacutereas que determinan la distribucioacuten de los organismos la zona foacutetica (donde
penetra la luz y permite la fotosiacutentesis) y la zona afoacutetica (subyacente a la zona
foacutetica) Al depender de la radiacioacuten la zona foacutetica variacutea estacionalmente Ademaacutes
18
puede variar por el grado de turbidez la morfometriacutea del lago y la produccioacuten de
materia orgaacutenica en la columna de agua
Otro factor que influye en la productividad es el reacutegimen de mezcla de la
columna de agua pues favorece la oxigenacioacuten y el reciclado de nutrientes La
mezcla ocurre en condiciones de gran inestabilidad usualmente relacionado con el
reacutegimen de viento Por el contrario un lago estratificado resulta de grandes
diferencias en temperatura o salinidad entre la superficie (epilimnion) y el fondo del
lago (hipolimnion) que separa las masas de agua superficial y de fondo por una
termoclina (si la estratificacioacuten estaacute relacionada con diferencias de temperatura de
las masas de agua) o haloclina (diferencias de salinidad) En funcioacuten del reacutegimen
de mezcla los lagos se pueden clasificar en (Lewis 1983)
1 Amiacutecticos no hay mezcla vertical de la columna de agua
2 Monomiacutecticos friacuteos y caacutelidos se mezcla una vez al antildeo
3 Dimiacutecticos la columna de agua se mezcla dos veces al antildeo
4 Oligomiacutecticos Lagos estratificados la mayor parte del tiempo pero se mezclan a
intervalos irregulares mayores a 1 antildeo
5 Polimiacutecticos friacuteos y caacutelidos la columna de agua se mezcla varias veces al antildeo
El ciclo del N en lagos
Al estar presente en aacutecidos nucleicos aminoaacutecidos y proteiacutenas el N es un
nutriente esencial de los organismos Por lo tanto su disponibilidad en la columna
de agua es clave para la produccioacuten composicioacuten y acumulacioacuten de la MO a traveacutes
del tiempo (Olson 1998 Talbot 2002) Aunque el principal reservorio de N estaacute en
19
la atmosfera (N2) solo unos pocos organismos (eg cianobacterias) pueden fijarlo
directamente Asiacute el Nitroacutegeno Inorgaacutenico Disuelto (DIN) representa la principal
fuente de N disponible para la produccioacuten primaria en los ecosistemas acuaacuteticos
(Talbot et al 2002) El DIN estaacute compuesto por nitrato (NO3-) nitrito (N02
-) y amonio
(NH4+) y los cambios en su disponibilidad condicionan el tipo de produccioacuten primaria
(eg fijadores vs asimiladores de N ver Scott y Grant 2013 Scott 2008)
La Figura 1 resume los principales componentes en lagos del ciclo del N y
sus interacciones La principal entrada natural de N es viacutea fijacioacuten de N atmosfeacuterico
y es llevado a cabo principalmente por bacterias y algas verde-azules capaces de
romper el triple enlace entre los dos aacutetomos de N a un alto costo energeacutetico (Torres
et al 2012) El N fijado es reducido a NH3 Tras la muerte de los organismos el N
es liberado de sus tejidos por hongos y bacterias implicados en la descomposicioacuten
de la MO Una parte se deposita en el fondo del lago y la otra queda disponible para
ser asimilada por el fitoplancton como amonio mediante el proceso de
amonificacioacuten (Talbot 2002) La amonificacioacuten resulta de la reduccioacuten bacteriana
del amoniaco y se da preferentemente en condiciones anoacutexicas La volatizacioacuten del
amoniaco es un proceso por medio este se incorpora a la atmoacutesfera Este proceso
se da preferentemente en lagos aacutecidos (pH gt 85) El nitrato es otra forma de N
bioloacutegicamente disponible para el fitoplancton En los lagos el NO3 del DIN estaacute
compuesto por los aportes desde la cuenca hidrograacutefica en la que se circunscriben
por la resuspensioacuten desde el fondo del lago favorecido por procesos de mezcla
(descrito en seccioacuten anterior) y por los nitratos de la columna de agua Estos uacuteltimos
son el resultado de la nitrificacioacuten proceso realizado por bacterias aeroacutebicas
mediante el cual el amonio se oxida para formar nitrito y nitrato El ciclo se completa
20
con la desnitrificacioacuten que es la reduccioacuten bacteriana de nitrato al N atmosfeacuterico
Este proceso se da preferentemente en condiciones anoacutexicas
Figura 1 Materia Orgaacutenica sedimentaria y su relacioacuten con el ciclo del N y las
variaciones de δ15N (elaborado a partir de Talbot et al 2002) En la figura se
representan los factores clave en la acumulacioacuten de la MO sedimentaria y su
relacioacuten con el ciclo del N El δ15N de la MO es resultado de los aportes de MO
desde la cuenca MO de macroacutefitas y de la productividad bioloacutegica La productividad
en ecosistemas lacustres estaacute condicionada por DIN y la fijacioacuten de N atmosfeacuterico
El δ15N del pool del DIN disponible se va enriqueciendo por la asimilacioacuten
preferencial de 14N por parte del fitoplancton Ademaacutes el DIN remanente se va
enriqueciendo por accioacuten microbiana (amonificacioacuten nitrificacioacuten desnitrificacioacuten)
Reconstruyendo el ciclo del N a partir de variaciones en δ15N
La sentildeal isotoacutepica de N (δ15N) en los sedimentos lacustres puede ser usada
para reconstruir los cambios pasados del ciclo N la transferencia de N entre
ecosistemas terrestres y acuaacuteticos y el estado troacutefico del lago (Bruland y Mackenzie
2015 McLauchlan et al 2013 Woodward et al 2012 von Gunten et al 2009
Leng et al 2006 Meyers y Teranes 2001) La Figura 1 resume los principales
procesos y fuentes que afectan la sentildeal isotoacutepica δ15N (total o ldquobulkrdquo) en la MO de
21
los sedimentos lacustres Entre los que destacan las fuentes de MO (aloacutectona vs
autoacutectona) la bioproductividad lacustre y las entradas de N (ie fijacioacuten bioloacutegica
de N2) (Torres et al 2012) Estos procesos conducen a un fraccionamiento
isotoacutepico cineacutetico que favorece la disociacioacuten entre sus isotopos ldquopesadosrdquo (15N) y
ldquoligerosrdquo (14N) Asiacute por ejemplo los valores de δ15N de la MO se enriquecen en 15N
en la medida que los sedimentos lacustres estaacuten sometidos a perdidas de 14N viacutea
desnitrificacioacuten (Bruland y Mackenzie 2015 Diaz et al 2016 Talbot et al 2002)
Por otra parte el fitoplancton en la medida que aumenta su biomasa (eg
durante la estacioacuten caacutelida) puede llegar a asimilar el DIN hasta agotarse En este
caso el N remanente se va empobreciendo en 14N si no hay reposicioacuten de N (eg
aporte de NO3 de la cuenca) y la MO queda enriquecida en 15N Esta disminucioacuten
induce a una limitacioacuten de fitoplancton por N y los organismos diztroacuteficos pueden
verse estimulados por lo que se incrementa la entrada de N viacutea fijacioacuten de N2 (Scott
y Grantsz 2013) como resultado la MO oscila en valores maacutes bajos (en torno a 0permil)
La cantidad de MO que se deposita en el fondo del lago depende del
predominio de contribuciones provenientes desde la cuenca (aloacutectona) versus las
producidas dentro del mismo lago (autoacutectona) (Torres et al 2012) Aunque en
general los lagos reciben permanentemente aportes de sedimentos y MO desde
su cuenca los lagos pequentildeos tienden a reflejar maacutes los procesos que ocurren
solamente en su cuenca directa y reflejan los valores isotoacutepicos de la misma (Gu et
al 2006) Woodward et al (2012) realizaron un estudio en 50 lagos en Irlanda que
les permitioacute correlacionar diferentes patrones de LUCC (bosques de coniacuteferas
agricultura ganaderiacutea entre otros) con los valores de δ15N (y δ13C) en los
sedimentos superficiales de los lagos Encontraron que los valores de δ15N son maacutes
negativos en cuencas poco impactadas y maacutes positivos en aquellas con alto
22
impacto humano (eg usos de suelo agriacutecola y ganadero) McLauchlan et al (2007)
encontraron valores maacutes positivos de δ15N en los sedimentos del Mirror Lake New
Hampshire (29ordm N) a partir de la desforestacioacuten de su cuenca en 1790 CE (inicio
del asentamiento europeo en el lago) para el desarrollo de la actividad agriacutecola
Estos valores se volvieron maacutes negativos hacia valores similares al pre-
asentamiento europeo despueacutes del cese de la agricultura en la cuenca y la
recuperacioacuten del bosque a partir de 1929 CE
El anaacutelisis de δ15N en los sedimentos lacustres es una herramienta casi sin
explorar en Chile central Proponemos reconstruir la dinaacutemica de transferencia de
N cuenca-lago como consecuencia de los LUCC desde la colonizacioacuten espantildeola en
los lagos costeros de Laguna Matanzas (34ordmS) y Lago Vichuqueacuten (36ordmS) Como son
muacuteltiples los procesos que pueden afectar la sentildeal isotoacutepica de N (medido como
δ15N) en los sedimentos lacustres existen muchos problemas para su
interpretacioacuten paleoambiental (Leng et al 2006) Por ello en esta tesis utilizamos
un enfoque multiproxy que consiste en integrar el anaacutelisis limnoloacutegico y geoquiacutemico
de los sedimentos lacustres con los valores isotoacutepicos modernos de la columna de
agua (mediante monitoreo del POM) la relacioacuten vegetacioacuten-suelo de la cuenca y la
reconstruccioacuten de los principales usos de suelo de las cuencas desde 1975 CE
mediante el uso de imaacutegenes satelitales Considerando estas muacuteltiples liacuteneas de
evidencia nos planteamos utilizar las variaciones del δ15N para reconstruir los
cambios en la disponibilidad de N en los lagos costeros de Chile central y establecer
coacutemo y desde cuaacutendo las actividades humanas en la cuenca son lo suficientemente
importantes como para modificar el ciclo del N en los lagos desde la colonizacioacuten
espantildeola (siglo XVII)
23
Como hipoacutetesis planteamos que la dinaacutemica de transferencia de sedimentos
y nutrientes entre la cuenca y el lago tiene controles geoloacutegicos climaacuteticos y
bioloacutegicos que interactuacutean en el tiempo registraacutendose en la acumulacioacuten de MO de
los sedimentos lacustres (Fig1) Bajo este marco los LUCC modifican esta
dinaacutemica alterando la transferencia de N a los lagos ya que las actividades agriacutecolas
y ganaderas tienden a aportar maacutes N (aumentando δ15N) que la actividad forestal
(la que disminuye δ15N)
En el primer capiacutetulo titulado ldquoA combined approach to establishing the timing
and magnitude of anthropogenic nutrient alteration in a mediterranean coastal lake-
watershed systemrdquo se evaluoacute el rol de los LUCC en el control de la descarga de N
y sedimentos a Laguna Matanzas en un sistema cuenca-lago desde el siglo XVII
Para ello se realizoacute un anaacutelisis de la dinaacutemica sedimentaria lacustre mediante el
anaacutelisis de muacuteltiples proxys sedimentoloacutegicos (ie minerales y textura)
geoquiacutemicos (eg geoquiacutemica orgaacutenica e isotoacutepica) y bioloacutegicos (ie diatomeas) de
Laguna Matanzas que cubren los uacuteltimos 200 antildeos Ademaacutes se realizoacute una
reconstruccioacuten de los usos de suelo entre 1975 y 2016 a partir de imaacutegenes de
sateacutelites y se colectaron muestras de suelo de las principales coberturas de la
cuenca a los cuales se midioacute el δ15N
Entre los principales resultados obtenidos se destaca la influencia de la
ganaderiacutea y la denitrificacioacuten lo que explica los altos valores de δ15N evidenciados
por el registro lacustre durante la colonizacioacuten espantildeola e inicios de la repuacuteblica A
partir de la Gran Aceleracioacuten (1950 CE) la desforestacioacuten y el reemplazo de la
ganaderiacutea por plantaciones forestales tienen un correlato en el registro
sedimentario con valores de δ15N maacutes bajos Estos resultados demuestran que los
LUCC son el factor de primer orden para explicar los cambios observados en
24
nuestro registro de δ15N y a su vez implica un rol secundario del clima como posible
control de la disponibilidad de N en Laguna Matanzas Este capiacutetulo fue sometido
a la revista Scientific Reports y estaacute en revisioacuten desde el 26 de marzo de 2019 En
la elaboracioacuten del mauscrito han colaborado Claudio Latorre Blas Valero-Garceacutes
Matiacuteas Frugone Pablo Sarricolea Santiago Giralt Manuel Contreras-Lopez
Ricardo Prego y Patricia Bernardez
El segundo capiacutetulo se titula ldquoStable isotopes track land use and cover
changes in a mediterranean lake in central Chile over the last 600 yearsrdquo Aquiacute
evaluamos la dinaacutemica del N actual en el sistema cuenca-lago considerando los
valores isotoacutepicos modernos tanto de la vegetacioacuten como del suelo ademaacutes de los
cambios estacionales del POM de la columna de agua Esta informacioacuten se utiliza
como un anaacutelogo moderno para conocer el rol de los LUCC en la disponibilidad de
N en los uacuteltimos 600 antildeos Para ello se colectaron muestras filtradas de la columna
de agua a 2 5 y 20 m dos veces por antildeo entre noviembre de 2015 y julio de 2018
y se obtuvo el δsup1⁵N del POM Ademaacutes se colectoacute muestras de vegetacioacuten y suelo
de la cuenca diferenciando entre especies nativas plantaciones forestales y
vegetacioacuten herbaacutecea Esta informacioacuten se analizoacute en conjunto con la reconstruccioacuten
de los LUCC entre 1975 y 2014 y el anaacutelisis limnoloacutegico del lago Ello nos permitioacute
evaluar coacutemo el δ15N de los sedimentos lacustres refleja los valores δ15N de la
cuenca en el Lago Vichuqueacuten y el control que ejerce el clima en la sentildeal isotoacutepica
de POM Este capiacutetulo seraacute sometido a la revista Science of total environmet
proximamente En la elaboracioacuten del mauscrito han colaborado Claudio Latorre
Blas Valero-Garceacutes Matiacuteas Frugone Pablo Sarricolea Leonardo Villacis y M Laura
Carrevedo
25
Entre los principales resultados encontramos que el δ15N en los sedimentos
lacustres es maacutes positivo con presencia de agricultura en la cuenca y maacutes negativo
cuando esta es reemplazada por cobertura arboacuterea bien sea plantaciones
forestales bien sea bosque nativo A su vez durante la estacioacuten seca y caacutelida la
mayor entrada al lago es viacutea fijacioacuten de N atmosfeacuterico (bajos valores de δ15NPOM)
Durante la estacioacuten huacutemeda en cambio prevalecen los aportes de la cuenca con
altos valores de δ15NPOM Ello estariacutea evidenciando un cambio estacional en la
composicioacuten de especies en verano ligado a especies fijadoras de N y en invierno
las algas y microorganismos que consumen el DIN de la columna de agua
Referencias
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea AM Marquet PA 2010 From the Holocene to the Anthropocene A historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the
next carbon Earth rsquo s Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409 httpsdoiorg102134jeq20090005
Diacuteaz FP Frugone M Gutieacuterrez RA Latorre C 2016 Nitrogen cycling in an
extreme hyperarid environment inferred from δ15 N analyses of plants soils and herbivore diet Sci Rep 6 1ndash11 httpsdoiorg101038srep22226
Ekdahl EJ Teranes JL Wittkop CA Stoermer EF Reavie ED Smol JP
2007 Diatom assemblage response to Iroquoian and Euro-Canadian eutrophication of Crawford Lake Ontario Canada J Paleolimnol 37 233ndash246 httpsdoiorg101007s10933-006-9016-7
Elbert W Weber B Burrows S Steinkamp J Buumldel B Andreae MO
Poumlschl U 2012 Contribution of cryptogamic covers to the global cycles of carbon and nitrogen Nat Geosci 5 459ndash462
26
httpsdoiorg101038ngeo1486 Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506
httpsdoiorg101126science1215567 ARTICLE Engstrom DR Almendinger JE Wolin JA 2009 Historical changes in
sediment and phosphorus loading to the upper Mississippi River Mass-balance reconstructions from the sediments of Lake Pepin J Paleolimnol 41 563ndash588 httpsdoiorg101007s10933-008-9292-5
Erisman J W Sutton M A Galloway J Klimont Z amp Winiwarter W (2008)
How a century of ammonia synthesis changed the world Nature Geoscience 1(10) 636
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892
httpsdoiorg101126science1136674 Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J Christie
D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592 Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
IPCC 2013 Climate Change 2013 The Physical Science BasisContribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change Cambridge University Press Cambridge United Kingdom and New York NY USA
httpsdoiorg101017CBO9781107415324 Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007 Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32iumliquestfrac12S 3050iumliquestfrac12miumliquestfrac12asl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470
27
httpsdoiorg101073pnas0701779104 McLauchlan KK Gerhart LM Battles JJ Craine JM Elmore AJ Higuera
PE Mack MC McNeil BE Nelson DM Pederson N Perakis SS 2017 Centennial-scale reductions in nitrogen availability in temperate forests of the United States Sci Rep 7 1ndash7 httpsdoiorg101038s41598-017-08170-z
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Myers N Mittermeler RA Mittermeler CG Da Fonseca GAB Kent J
2000 Biodiversity hotspots for conservation priorities Nature httpsdoiorg10103835002501
Pentildeuelas J Poulter B Sardans J Ciais P Van Der Velde M Bopp L
Boucher O Godderis Y Hinsinger P Llusia J Nardin E Vicca S Obersteiner M Janssens IA 2013 Human-induced nitrogen-phosphorus imbalances alter natural and managed ecosystems across the globe Nat Commun 4 httpsdoiorg101038ncomms3934
Prieto M Salazar D Valenzuela MJ 2019 The dispossession of the San
Pedro de Inacaliri river Political Ecology extractivism and archaeology Extr Ind Soc httpsdoiorg101016jexis201902004
Robertson GP Vitousek P 2010 Nitrogen in Agriculture Balancing the Cost of
an Essential Resource Ssrn 34 97ndash125 httpsdoiorg101146annurevenviron032108105046
Sardans J Rivas-Ubach A Pentildeuelas J 2012 The CNP stoichiometry of
organisms and ecosystems in a changing world A review and perspectives Perspect Plant Ecol Evol Syst 14 33ndash47 httpsdoiorg101016jppees201108002
Schindler DW Howarth RW Vitousek PM Tilman DG Schlesinger WH
Matson PA Likens GE Aber JD 2007 Human Alteration of the Global Nitrogen Cycle Sources and Consequences Ecol Appl 7 737ndash750 httpsdoiorg1018901051-0761(1997)007[0737haotgn]20co2
Scott E and EM Grantzs 2013 N2 fixation exceeds internal nitrogen loading as
a phytoplankton nutrient source in perpetually nitrogen-limited reservoirs Freshwater Science 2013 32(3)849ndash861 10189912-1901
28
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Talbot M R (2002) Nitrogen isotopes in palaeolimnology In Tracking
environmental change using lake sediments (pp 401-439) Springer Dordrecht
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable
isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K
Yeager KM 2016 Different responses of sedimentary δ15N to climatic changes and anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
29
CAPIacuteTULO 1 A COMBINED APPROACH TO ESTABLISHING THE TIMING
AND MAGNITUDE OF ANTHROPOGENIC NUTRIENT ALTERATION IN A
MEDITERRANEAN COASTAL LAKE- WATERSHED SYSTEM
30
A combined approach to establishing the timing and magnitude of anthropogenic
nutrient alteration in a mediterranean coastal lake- watershed system
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugoneabc Pablo
Sarricolead Santiago Giralte Manuel Contreras-Lopezf Ricardo Prego g Patricia
Bernaacuterdez g Blas Valero-Garceacutesch
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Institute of Earth Sciences Jaume Almera (ICTJA-CSIC) CLluis Solegrave Sabaris sn Barcelona E-
08028 Spain
f Facultad de Ingenieriacutea y Centro de Estudios Avanzados Universidad de Playa Ancha Traslavintildea
450 Vintildea del Mar Chile
g Instituto de Investigaciones Marinas (CSIC) CEduardo Cabello 6 36208 Vigo Spain
h Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding author
E-mail address
clatorrebiopuccl magdalenafuentealbagmailcom
Abstract
Since the industrial revolution and especially during the Great Acceleration (1950
CE) human activities have profoundly altered the global nutrient cycle through land
use and cover changes (LUCC) However the timing and intensity of recent N
variability together with the extent of its impact in terrestrial and aquatic ecosystems
and coupled effects of regional LUCC and climate are not well understood Here
we used a multiproxy approach (sedimentological geochemical and isotopic
31
analyses historical records climate data and satellite images) to evaluate the role
of LUCC as the main control for N cycling in a coastal watershed system of central
Chile during the last few centuries The largest changes in N dynamics occurred in
the mid-1970s associated with the replacement of native forests and grasslands for
livestock grazing by plantations of introduced Monterey pine (Pinus radiata) and
eucalyptus (Eucalyptus globulus) Lake productivity increased and is evidenced by
an increase trend in δ15N values Our study shows that anthropogenic land
usecover changes are key in controlling nutrient supply and N availability in
Mediterranean watershed ndash lake systems and that large-scale forestry
developments during the mid-1970s likely caused the largest changes in central
Chile
Keywords Anthropocene Organic geochemistry watershedndashlake system Stable
Isotope Analyses Land usecover change Nitrogen cycle Mediterranean
ecosystems central Chile
1 INTRODUCTION
Human activities have become the most important driver of the nutrient cycles in
terrestrial and aquatic ecosystems since the industrial revolution (Gruber and
Galloway 2008 Galloway et al 2008 Holtgrieve et al 2011 Fowler et al 2013
Goyette et al 2016) Among these N is a common nutrient that limits productivity
in terrestrial and aquatic ecosystems (Vitousek and Howarth 1991 McLauchlan et
al 2013) With the advent of the Haber-Bosch industrial N fixation process in the
early 20th century total N fluxes have surpassed previous planetary boundaries
32
(Galloway et al 2008 Elser 2011) reaching unprecedented values (ie tipping
points) in the Earth system especially during what is now termed the Great
Acceleration (which began in the 1950s) which intensified in the 1970s (Howarth
2004 Steffen et al 2015) Yet dramatic changes in LUCC have occurred in the last
few centuries mainly linked to forestry and agro-pastoral activities (Poraj-Goacuterska et
al 2017 Ge et al 2019 Cheng et al 2019) These have had large impacts on N
(and Carbon -C-) availability in terrestrial and aquatic ecosystems but the synergic
effect with climate change and global N dynamics has not been established
(Vitousek et al 1997 Mclauchlan et al 2007 2013 Pongratz et al 2010
Woodward et al 2012 Mclauchlan et al 2017)
The onset of the Anthropocene poses significant challenges in mediterranean
regions that have a strong seasonality of hydrological regimes and an annual water
deficit (Stocker et al 2013) Mediterranean climates occur in all continents
(California central Chile Australia South Africa circum-Mediterranean regions)
providing a unique opportunity to investigate global change processes during the
Anthropocene in similar climate settings but with variable geographic and cultural
contexts The effects of global change in mediterranean watersheds have been
analyzed from different perspectives hydrology (Giorgi 2006 Arnell and Gosling
2013 Prudhomme et al 2014) vegetation dynamics (Lenihan et al 2003 Vicente-
Serrano et al 2013 Matesanz and Valladares 2014) sediment dynamics (Garciacutea-
Ruiz et al 2013 Garciacutea-Ruiz 2010 Syvitski et al 2005) changes in
biogeochemical cycles (Thomas et al 2013 Fowler et al 2013 Frank et al 2015)
carbon storage (Muntildeoz-Rojas et al 2015) and biodiversity (Hooper et al 2012) A
recent review showed an extraordinarily high variability of erosion rates in
mediterranean watersheds positive relationships with slope and annual
33
precipitation and the paramount effect of land use (Garciacutea-Ruiz et al 2015)
However the temporal context and effect of LUCC on nutrient supply to
mediterranean lakes has not been analyzed in much detail
Major LUCC in central Chile occurred during the Spanish Colonial period
(17th-18th centuries) (Armesto 1994 Gurnell et al 2001 Figueroa et al 2004
Martiacuten-Foreacutes et al 2012 Contreras-Loacutepez et al 2014) with the onset of
industrialization and mostly during the mid to late 20th century (von Gunten et al
2009 Gayo et al 2012) Recent copper pollution caused by 20th century mining
and industrial smelters has been documented in cores throughout the Andes
(Laguna el Ocho and Laguna Ensuentildeo von Gunten et al 2009) and also from our
surveys in the central valley (Batuco wetland) coastal range (Cordillera de Name)
and along the coast itself (Bucalemito and Colejuda) (Valero-Garceacutes et al 2010
unpublished data)
Paleolimnological studies have shown how these systems respond to
climate LUCC and anthropogenic impacts during the last millennia (Jenny et al
2002 2003 Villa Martiacutenez et al 2002 Frugone-Aacutelvarez et al 2017 Contreras et
al 2018) Furthermore changes in sediment and nutrient cycles have also been
identified in associated terrestrial ecosystems dating as far back as the Spanish
Conquest and related to fire clearance and wood extraction practices of the native
forests (Armesto 1994 Contreras-Loacutepez et al 2014) Nevertheless pollen and
limnological evidence argue for a more recent timing of the largest anthropogenic
impacts in central Chile For example paleo records show that during the mid-20th
century increased soil erosion followed replacement of native forest by Pinus
radiata and Eucalyptus globulus plantations at Laguna Matanzas Aculeo and
34
Vichuqueacuten lakes (Jenny et al 2002 Villa-Martiacutenez et al 2002 2003 Frugone-
Aacutelvarez et al 2017)
Lakes are a central component of the global carbon cycle Lakes act as a
sink of the carbon cycle both by mineralizing terrestrially derived organic matter and
by storing substantial amounts of organic carbon (OC) in their sediments (Anderson
et al 2009) Paleolimnological studies have shown a large increase in OC burial
rates during the last century (Heathcote et al 2015) however the rates and
controls on OC burial by lakes remain uncertain as do the possible effects of future
global change and the coupled effect with the N cycle LUCC intensification of
agriculture and associated nutrient loading together with atmospheric N-deposition
are expected to enhance OC sequestration by lakes Climate change has been
mainly responsible for the increased algal productivity since the end of the 19th
century and during the late 20th century in lakes from both the northern (Ruumlhland et
al2015) and southern hemispheres (Michelutti et al 2015 Carrevedo et al 2015)
but many studies suggest a complex interaction of global warming and
anthropogenic influences and it remains to be proven if climate is indeed the only
factor controlling these transitions (Catalaacuten et al 2013) Alternative causes for
recent N (Galloway et al 2008) increases in high altitude lakes such as catchment
mediated processes cannot be ruled out (Camarero and Catalan 2012 Fritz and
Anderson 2013) Few lake-watershed systems have robust enough chronologies of
recent changes to compare variations in C and N with regional and local processes
and even fewer of these are from the southern hemisphere (McLauchlan et al
2007 Holtgrieve et al 2011)
In this paper we present a multiproxy lake-watershed study including N and
C stable isotope analyses on a series of short cores from Laguna Matanzas in
35
central Chile focused in the last 200 years We complemented our record with land
use surveys satellite and aerial photograph studies Our major objectives are 1) to
reconstruct the dynamics among climate human activities and changes in the N
cycle over the last two centuries 2) to evaluate how human activities have altered
the N cycle during the Great Acceleration (since the mid-20th century)
2 STUDY SITE
Laguna Matanzas (33ordm45rsquoS 71ordm 40acuteW 7 m asl Fig 1) is a coastal lake located
in central Chile near to a large populated area (Santiago gt6106 inhabitants) The
lake has a surface area of 15 km2 with a max depth of 3 m and a watershed of 30
km2 The lake basin is emplaced over Pleistocene-Holocene aeolian and alluvial fan
deposits (SERNAGEOMIN 2003) A recent phase of dune activity occurred from the
mid to late Holocene which mostly sealed off the basin from the ocean (Villa-
Martinez 2002) Climate in Laguna Matanzas is characterized by cool-wet winters
and hot-dry summers with annual precipitation of ~510 mm and a mean annual
temperature of 12ordmC Central Chile is in the transition zone between the southern
hemisphere mid-latitude westerlies belt and the South Pacific Anticyclone (or SPA)
(Garreaud et al 2009) In winter precipitation is modulated by the north-west
displacement of the SPA the northward shift of the westerlies wind belt and an
increased frequency of storm fronts stemming off the Southern Hemisphere
Westerly Winds (SWW) (Frugone-Aacutelvarez et al 2017) Austral summers are
typically dry and warm as a strong SPA blocks the northward migration of storm
tracks stemming off the SWW
36
Historic land cover changes started after the Spanish conquest with a Jesuit
settlement in 1627 CE near El Convento village and the development of a livestock
ranch that included the Matanzas watershed After the Jesuits were expelled from
South America in 1778 CE the farm was bought by Pedro Balmaceda and had more
than 40000 head of cattle around 1800 CE (Contreras-Lopez et al 2014) The first
Pinus radiata and Eucalyptus globulus trees were planted during the second half of
the 19th century and were mostly used for dune stabilization (Albert 1900 Gibson
1972) However the main plantation phase occurred 60 years ago (Villa-Martinez
2002) as a response to the application of Chilean Forestry Laws promulgated in
1931 and 1974 and associated state subsidies
Major land cover changes occurred recently from 1975 to 2008 as shrublands
were replaced by more intensive land uses practices such as farmland and tree
plantations (Schulz et al 2010) Laguna Matanzas is part of the Reserva Nacional
Humedal El Yali protected area (Ramsar Ndeg 878) Despite this protected status the
lake and its watershed have been heavily affected by intense agricultural and
farming activities during the last decades The main inlet ldquoLas Rosasrdquo has been
diverted for crop irrigation causing a significant loss of water input to the lake
Consequently the flooded area of the lake has greatly decreased in the last couple
of decades (Fig 1b) Exotic tree species cover a large surface area of the
watershed Recently other activities such as farms for intensive chicken production
have been emplaced in the watershed
37
Figure 1 Laguna Matanzas a) Digital Elevation Model showing the watershed and
the surface hydrologic connection with Colejuda and Cabildo lakes b) Climograph
depicting the warm dry season in austral summer c) Annual precipitation from
1965-2015 (arrow shows start of the most recent ldquomega-droughtrdquo- see Garreaud et
al 2017) d) A decline of lake area occurs between 2007 and 2018 Lake surface
area decreased first along the western sector (in 2007) followed by more inland
areas (in 2018)
38
3 RESULTS
31 Age Model
The age model for the Matanzas sequence was developed using Bacon software
to establish the deposition rates and age uncertainty (Blaauw and Christen 2011)
It is based on 210Pb and 137Cs dates and two 14C dates (Table 2) According to this
age model the lake sequence spans the last 1000 years (Fig 2) A major breccia
layer (unit 3b) was deposited during the early 18th century which agrees with
historic documents indicating that a tsunami impacted Laguna Matanzas and its
watershed in 1730 CE (Contreras-Loacutepez et al 2014) Here we focus in the last 200
years were the most important changes occurred in terms of LUCC (after the
sedimentary hiatus caused by the tsunami) The Spanish colonial period (17th -18th
century) brought new forms of territorial management along with an intensification
of watershed use which remained relatively unchanged until the 1900s
39
Figure 2 a) Age-depth model obtained for the Laguna de Matanzas sedimentary
sequence A hiatus occurs at 80 cm depth (unit 3b) The section of core used for our
analysis is highlighted in a red rectangle b) Close up of the age model used for
analysis of recent anthropogenic influences on the N cycle c) Information regarding
the 14C dates used to construct age model
Lab code Sample ID
Depth (cm) Material Fraction of modern C
Radiocarbon age
Pmc Error BP Error
D-AMS 021579
MAT11-6A 104-105 Bulk Sediment
8843 041 988 37
D-AMS 001132
MAT11-6A 1345-1355
Bulk Sediment
8482 024 1268 21
POZ-57285
MAT13-12 DIC Water column 10454 035 Modern
Table 2 Laguna Matanzas radiocarbon dates
32 The sediment sequence
Laguna Matanzas sediments consist of massive to banded mud with some silt
intercalations They are composed of silicate minerals (plagioclase quartz and clay
minerals) with relatively high TOC content (Fig 3) Pyrite is a common mineral
indicating dominant anoxic conditions in the lake sediments whereas aragonite
occurs only in the uppermost section Mineralogical analyses visual descriptions
texture and geochemical composition were used to characterize five main facies
(Fig 3) F1 (organic-rich mud) represents baseline sedimentation in a shallow well-
mixed brackish highly productive lake and F1acute (dark orange) is a less organic facies
than F1 (more details see table in the supplementary material) F2 (massive to
banded silty mud) indicates periods of higher clastic input into the lake but finer
(mostly clay minerals) likely from suspension deposition associated with flooding
40
events Aragonite (up to 15 ) occurs in both facies but only in samples from the
uppermost 30 cm and it is interpreted as endogenic linked to higher MgFe waters
and elevated biologic productivity
Figure 3 Sedimentary facies and units mineralogy grain size elemental geochemical
and isotopic composition of Laguna Matanzas core Values highlighted in gray indicate
that these are above average
The banded to laminated fining upward silty clay layers (F3) reflect
deposition by high energy turbidity currents The presence of aragonite suggests
that littoral sediments were incorporated by these currents Non-graded laminated
coarse silt layers (F4) do not have aragonite indicating a dominant watershed
41
sediment source Both facies are interpreted as more energetic flood deposits but
with different sediment sources A unique breccia layer with coarse silt matrix and
cm-long soft clasts (F5) is interpreted as a high-energy event (ie a tsunami)
capable of eroding the littoral zone and depositing coarse clastic material in the
distal zone of the lake Similar coarse breccia layers have been found at several
coastal sites in Chile and interpreted as tsunami-related deposits (Vargas et al
2005 Le Roux et al 2008)
33 Sedimentary units
Three main units and six subunits have been defined (Fig 3) based on
sedimentary facies and sediment composition We use ZrTi as an indicator of the
mineral fraction transported from the watershed (Marzecoacuteva et al 2011) as higher
ZrTi (F3 and F4) is commonly associated with coarser sediments (Cuven et al
2010) A high correlation among Br BrTi and TOC (r = 046-087 p value=0)
supports the use of BrTi as an indicator of lake productivity (Marzecoacuteva et al 2011
Frugone-Aacutelvarez et al 2017) The MnFe ratio is indicative of lake bottom
oxygenation (Naecher et al 2013) as under reducing conditions Mn mobilizes more
than Fe leading to a decreased MnFe ratio (Marzecoacuteva et al 2011) SrTi indicates
periods of increased aragonite formation as Sr is preferentially included in the
aragonite mineral structure (Veizer et al 1971) (See supplementary material)
The basal unit (3c 129 to 99 cm) is relatively organic-rich (TOC mean=26
BioSi mean=5) and composed by F1 (without aragonite) with some coarser F4
flood layers (ZrTi mean=025) Unit 3b (98 to 80 cm) is interpreted as a tsunami or
storm surge deposit (breccia F5 grading into massive to banded silt F2) Unit 3a
(79 to 73 cm) is characterized by relatively low productivity (TOC=2 BrTi=002
42
BioSi=4) under anoxic conditions (MnFe=001) Unit 2a (72 to 31cm) has
relatively less organic content and more intercalated clastic facies F3 and F4 The
top of this unit (43-30 cm) has elevated TS values The Subunit 1b (30-20 cm)
shows increasing TOC BioSi and BrTi values (TOC mean = 29 BioSi mean =
54 BrTi mean=004) and the upper subunit 1a (19-0 cm) has the highest TOC
(mean = 64) and BioSi (mean = 56) high BrTi mean = 010 and the presence
of aragonite More frequent anoxic conditions (MnFe lower than 001) during units
3 and 2 shifted towards more oxic episodes (MnFe = 003) in unit 1 (Fig 3)
34 Isotopic signatures
Figure 4 shows the isotopic signature from soil samples of the major land
usescover present in the Laguna Matanzas used as an end member in comparison
with the lacustrine sedimentary units δ15N from cropland samples exhibit the
highest values whereas grassland and soil samples from lake shore areas have
intermediate values (Fig 4) Tree plantations and native forests have similarly low
δ15N values (+11 permil SD=24) All samples (except those from the lake shore)
exhibit low δ13C values (from ndash285permil to ndash298permil) CNmolar from agriculture land
lakeshore area and non-vegetation areas samples display the lowest values (about
18) CNmolar from tree plantations and native forest have the highest values (383
and 267 respectively)
43
Figure 4 C-N stable isotope plot showing a comparison of lake sediments grouped
by sedimentary units (MAT11-6A) with the soil end members of present-day (lake
shore and land usecover) from Laguna Matanzas
The δ15N values from sediment samples (MAT11-6A) range from ndash15 and
+53permil (mean= +35 SD=05) δ13C values range from ndash266permil to ndash202permil (mean=
ndash240 SD=14) In unit 3c δ15N and δ13C show relatively high values (mean=
+41permil and ndash233permil respectively) δ15N and δ13C from unit 3b and 3a fluctuate at
slightly lower values than in 3c (mean δ15N from 3a= +38permil and 3b mean= +39permil
mean δ13C from 3a= ndash242permil and 3b mean= ndash244permil) In unit 2a δ15N values are
relatively high (mean= +38permil) but show a slightly decreasing trend (from +52 to
+24permil at 66 cm and 36 cm respectively) δ13C also decreases (mean= ndash242permil)
reaching minimum values at 45 cm (ndash266permil) with a sharp increase towards the top
of this unit with maximum values ca ndash210permil Unit 1 exhibits the lowest δ15N values
(1a mean= +11permil 1b mean= +28permil) with negative values in the uppermost
44
sediments (ndash04permil at 14 cm) δ13C values show a decreasing trend over most of
subunit 1b and increase only near the very top of this unit
35 Recent land use changes in the Laguna Matanzas watershed
Major LUCC from 1975 (unit 1b) to 2016 CE in the Laguna Matanzasacutes
watershed is summarized in Figure 5 The watershed has a surface area of 30 km2
of which native forest (36) and grassland areas (44) represented 80 of the
total surface in 1975 The area occupied by agriculture was only 02 and tree
plantations were absent Isolated burned areas (33) were located mostly in the
northern part of the watershed By 1989 tree plantations surface area had increased
to 5 burned areas to 17 and agricultural fields to 9 (a 45-fold increase) and
native forest and grassland sectors decreased to 23 and 27 respectively By
2016 agricultural land and tree plantations have increased to 17 of the total area
whereas native forests decreased to 21
45
Figure 5 Land uses and cover changes from 1975 to 2016 in Laguna Matanzas
watershed from natural cover and areas for livestock grazing (grassland) to the
expansion of agriculture and forest plantation
4 DISCUSSION
41 N and C dynamics in Laguna Matanzas
Small lakes with relatively large watersheds such as Laguna Matanzas would
be expected to have relatively high contributions of allochthonous C to the sediment
OM (Gu et al 2006) Terrestrial C3 plants (δ13C =ndash26 to ndash28permil Meyers and Terranes
2001 Ku et al 2007) are dominant in the Laguna Matanzas watershed Likewise
our soil samples ranged across similar although slightly more negative values
46
(δ13C = ndash30 to ndash28permil Fig 4) to those proposed by Meyers and Terranes (2001)
and are used here as terrestrial end members oil samples were taken from the lake
shore (δ13C = ndash22permil) and POM from surface water (δ13C = ndash24permil) were more
positive than the terrestrial end member and are used as lacustrine end members
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from terrestrial vegetation and more positive δ13C values have increased
aquatic OM (Gu et al 2006 Torres et al 2012) Phytoplankton preferentially uptake
12C leaving the DIC pool enriched in 13C (Gu et al 2006) especially when there are
no important external sources of C (eg decreased C input from the watershed)
Therefore during events of elevated primary productivity the phytoplankton uptakes
12C until its depletion and are then obligated to use the heavier isotope resulting in
an increase in δ13C Changes in lake productivity thus greatly affect the C isotope
signal (Torres et al 2012) with high productivity leading to elevated δ13C values
(Torres et al 2012 Gu et al 2006)
In a similar fashion the N isotope signatures in Laguna Matanzas reflect a
combination of factors including different N sources (autochthonousallochthonous)
and lake processes such as productivity isotope fractionation in the water column
and sediment denitrification Elevated δ15N values from a POM sample (+22permil) and
average values from the lake shore (mean=+34permil SD=028) are used as aquatic
end members whereas terrestrial samples have values from +10 +24 (tree species)
to +77permil +35 (agriculture) which we use as terrestrial end members (Fig 4)
Autochthonous OM in aquatic ecosystems typically displays low δ15N values
when the OM comes from N-fixing species Atmospheric fixation of N2 by
cyanobacteria ends in OM with δ15N values close to 0permil (Leng et al 2006)
Phytoplankton preferentially uptake 14N from Dissolved Inorganic Nitrogen (DIN) in
47
the water column and derived OM typically have δ15N values lower than DIN values
When productivity increases the remaining DIN becomes depleted in 14N which in
turn increases the δ15N values of phytoplankton over time especially if the N not
replenished (Torres et al 2012) Thus high POM δ15N values from Laguna
Matanzas reflect elevated phytoplankton productivity with a 14N- depleted DIN In
addition N-watershed inputs also contribute to high δ15N values Heavily impacted
watersheds by human activities are often reflected in isotope values due to land use
changes and associated modified N fluxes For example the input of N runoff
derived from the use of inorganic fertilizers leads to the presence of elevated δ15N
(between ndash4 to +4permil) values in the water bodies (Elliott and Brush 2006 Diebel and
Vander Zanden 2009) Widory et al (2004) reported a direct relationship between
elevated δ15N values and increased nitrate concentration from manure in the
groundwater of Brittany (France) Terranes and Bernasconi (2000) found a good
correlation between augmented nutrient loading and a progressive increase in δ15N
values of sedimentary OM related to agricultural land use
Post-depositional diagenetic processes can further affect C and N isotope
signatures Several studies have shown a decrease in δ13C values of OM in anoxic
environments particularly during the first years of burial related to the selective
preservation of OM depleted in 13C (Hollander and Smith 2001 Lehmann et al
2002 Gӓlman et al 2009 Torres et al 2012) Diagenetic processes can also lead
to post-burial N isotope enrichment of the sediments Indeed 14N is consumed more
rapidly than 15N by denitrifying bacteria which intensifies under anoxic conditions
(Diebel and Vander Zanden 2009) Thus the remaining OM pool will be enriched
in 15N leaving to elevated δ15N values (Granger et al 2008 Menzel et al 2013)
48
In summary the relatively high δ15N values in sediments of Laguna Matanzas
reflect N input from an agriculturegrassland watershed with positive synergetic
effects from increased lake productivity enrichment of DIN in the water column and
most likely denitrification The increase of algal productivity associated with
increased N terrestrial input andor recycling of lake nutrients (and lesser extent
fixing atmospheric N) and denitrification under anoxic conditions can all increase
δ15N values (Fig 3) In addition elevated lake productivity without C replenishing
(eg by terrestrial C input) produces shifts towards positive δ13C values whereas C
input from the watershed generates more negative δ13C values
42 Recent evolution of the Laguna Matanzas watershed
Sedimentological compositional and geochemical indicators show three
depositional phases in the lake evolution under the human influence in the Laguna
Matanzas over the last two hundred years Although the record is longer (around
1000 years) we analyzed the last two centuries (unit 2 and 1) to provide a recent
historical context for the large changes detected during the 20th century
The first phase lasted from the beginning of the 19th century until ca 1940
(unit 2a Fig 3 4) and was characterized by moderate productivity with elevated
sediment input from the watershed as indicated by our geochemical proxies (BrTi
= 004 AlTi gt 01 Fig 6) Higher lake levels and dominant anoxic bottom conditions
(MnFe = 001 Fig3) can be related to increased precipitation (mean= 557 mmyr)
and lower temperatures (summer annual temperature lt19ordmC) During the Spanish
colonial period the Laguna Matanzas watershed was used as a livestock farm
(Jesuits 1627-1780 unit 3 followed by the Pedro Balmaceda era 1780-1810- unit
2a) (Contreras-Loacutepez et al 2014) The main buildings and farm facilities at the El
49
Convento village During this period livestock grazing and lumber extraction for
mining would have involved extensive deforestation and loss of native vegetation
(eg Armesto et al 1994 2010) However the Matanzas pollen record does not
show any significant regional deforestation during this period (Villa Martiacutenez 2002)
suggesting that the impact may have been highly localized
Lake productivity sediment input and elevated precipitation (Fig 6) all
suggest that N availability was related to this increased input from the watershed
The N from cow manure and soil particles would have led to higher δ15N values
(Elliot and Brush 2016) In addition anoxic lake bottom conditions would lead to
even further enrichment of buried sediment N The δ13C values lend further support
to our interpretation of increased sediment input -and N- from the watershed
Decreased δ13C values reach their lowest values in the entire record (ndash270permil) at
ca 1910 CE (Fig 4 6)
During most of the 19th century human activities in Laguna Matanzas were
similar to those during the Spanish Colonial period However the appearance of
Pinus radiata and Eucalyptus globulus pollen ca 1890 CE (Gibson 1972) the dune
stabilization-afforestation program which began ca 1900 CE (Albert 1900) and the
application of the Chilean forestry law of 1931 (DFL nordm265) contributed to an
increased capacity of the surrounding vegetation to retain nutrients and sediments
The law subsidized forest plantations in areas devoid of vegetation and prohibited
the cutting of forest on slopes greater than 45ordm These land use changes were coeval
with decreased sediment inputs (AlTi trend) from the watershed slightly increased
lake productivity (BrTi from 001 to 003) and a decrease in annual precipitation
(Fig 6) N isotope values become more negative during this period although they
remained high (from +49permil to +37permil) whereas the δ13C trend towards more
50
positive values reflects changes in the N source from watershed to in-lake dynamics
(e g increased endogenic productivity)
The second phase started after 1940 and is clearly marked by an abrupt
change in the general trend of δ15N that oscillates around +41permil (SD = 04permil) during
the first phase to +24permil (SD = 14permil) during the second These δ15N trends reflect
the lowest watershed nutrient and sediment inputs (based on the AlTi record)
decreased precipitation (mean = 318 mm year) and a slight increase in lake
productivity (increased BrTI) Depositional dynamics in the lake likely crossed a
threshold as human activity intensified throughout the watershed and lake levels
decreased
During the Great Acceleration δ15N values shifted towards higher values to
ca 3permil with an increase in δ13C values that are not reflected either in lake
productivity or lake level As the sediment input from the watershed increased and
precipitation remained as low as the previous decade δ15N values during this period
are likely related to watershed clearance which would have increased both nutrient
and sediment input into the lake
The δ13C trend to more positive values reaching the peaks in the 1960s (ndash
212permil) at the same time as the peaks in temperature (196 ordmC mean annual) with a
downward trend in precipitation A shift in OM origin from macrophytes and
watershed input influences to increased lake productivity could explain this trend
(Fig 4 1b)
In the 1970s the Laguna Matanzasacute watershed was mostly covered by native
forest (36) and grassland areas were intended for livestock grazing (44 Fig 5)
Both soils have δ15NSoil lt 5permil (Fig 4) Farming fields occupied a very small area and
tree plantations were almost nonexistent The decreasing trend in δ15N values seen
51
in our record is interrupted by several large peaks that occurred between ca 1975
and ca 1989 when the native forest and grassland areas fell by 23 and 27
respectively largely due to fires affecting 17 of the forests Agriculture fields
increased by 4 and forest plantations increased by 9 (unit 1a) Concomitantly
sediment ndash and likely N - inputs from the watershed decreased (as indicated by the
trend in AlTi) the precipitation was still relatively low (Fig 6) These changes are
likely related to the increase of vegetation cover especially of tree plantations (which
have more negative δ15N values) The small increase in productivity in the lake could
have been favored by increased temperature (von Gunten et al 2009) After 1989
the increase in agriculture land (17 in 2016) correlates with increasing δ15N δ13C
and TOC trends in spite of declining rainfall The increase of forest plantations was
mostly in response to the implementation of the Law Decree of Forestry
Development (DL 701 of 1974) that subsidized forest plantation After 1989 the
increase in agricultural land (17 in 2016) is synchronous with increasing δ15N
δ13C TOC and MnFe trends despite the decline in rainfall and overall lower lake
levels as more water is used for irrigation
The third phase started c 1990 CE (unit 1a) when OM accumulation rates
increase and δ13C δ15N decreased reaching their lowest values in the sequence
around 2000 CE Afterward during the 21st century δ13C and δ15N values again
began to increase The onset of unit 1 is marked by increased lake productivity and
decreased sediment input (AlTi lt 02) synchronous with intensive farming replacing
forestry and extensive agriculture (Fig 5 6)
A change in the general trend of δ15N values which decreased until 1990
(unit 1a) as the native forest and grassland area fell to 23 and 27 respectively
is most likely due to deforestation and fires Agriculture surface increased to 4 and
52
forest plantations increased by 9 (Fig 5) Concomitantly sediment ndash and likely N
ndash inputs from the watershed decreased probably related to the low precipitation (Fig
1b) and the increase of vegetation cover in the watershed in particularly by tree
plantations (with more negative δ15N Fig 4)
At present agriculture and tree plantations occupy around 34 of the
watershed surface whereas native forests and grassland cover 21 and 25
respectively Lake productivity as indicated by BrTi (Fig6) is higher and generates
OM with lower δ15N and higher δ13C values (about +2permil and ndash20permil ca 2000 CE
respectively) due to in-lake processes (ie biological N fixation and nutrient
recycling) and driven by changes in the arboreal cover which diminishes nutrient
flux into the lake (Fig6)
53
Figure 6 Anthropogenic and climatic forcing and lake dynamics response
(productivity sediment input N and C cycles) at Matanzas Lake over the last two
54
centuries Mean annual precipitation reconstructed and temperatures (von Gunten
et al 2009) Vertical gray bars indicate mega-droughts
5 CONCLUSIONS
Human activities have been the main factor controlling the N and C cycle in
the Laguna Matanzas during the last two centuries The N isotope signature in the
lake sediments reflects changes in the watershed fluxes to the lake but also in-lake
processes such as productivity and post-depositional changes Denitrification could
have been a dominant process during periods of increased anoxic conditions which
were more frequent prior to Great Acceleration (1950 CE) Higher δ15N and lower
δ13C values are associated with increased nutrient input from the watershed due to
increased livestock grazing and agriculture pressure (phase 1 Fig 7a) whereas
lower isotope values occurred during periods of increased forest plantations (phase
3 Fig 7c) During periods of increased lake productivity - such as in the last few
decades - δ15N values increased significantly
The most important change in C and N dynamics in the lake occurred after the
1950s (Phase 2 and 3) as tree plantations dominated the watershed Recent
changes in N dynamics can be explained by the higher nutrient contribution
associated with intensive agriculture (i e fertilizers) since the 1990s Although the
replacement of livestock activities with forestry and farming seems to have reduced
nutrient and soil export from the watershed to the lake the inefficient use of fertilizer
(by agriculture) can be the ultimate responsible for lake productivity increase during
the last decades
55
Figure 7 Schematic diagrams illustrating the main factors controlling the
isotope N signal in sediment OM of Laguna Matanzas N input from watershed
depends on human activities and land cover type Agriculture practices and cattle
(grassland development) contribute more N to the lake than native forest and
plantations Periods of higher productivity tend to deplete the dissolved inorganic N
in 14N resulting in higher δ15N (OM) The denitrification processes are more effective
in anoxic conditions associated with higher lake levels
6 METHODS
Short sediment cores were recovered from Laguna Matanzas using an Uwitec
gravity piston corer in 2011 and 2013 (Fig 1a) Sediment cores (MAT11-6A 129 cm
MAT13-2A 149 cm MAT13-3A 33 cm MAT13-4A 99 cm) were split
photographed sub-sampled and stored at the Pyrenean Institute of Ecology (IPE-
CSIC Spain) Core MAT11-6A was obtained from the central sector of the lake and
56
was selected for detailed multiproxy analyses (including elemental geochemistry C
and N isotope analyses XRF and 14C dating)
The isotope analyses (δ13C and δ15N) were performed at the Laboratory of
Biogeochemistry and Applied Stable Isotopes (LABASI-PUC Chile) using a Delta
V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via a
Conflo IV interface Isotope results are expressed in standard delta notation (δ) in
per mil (permil) relative to the standards Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Sediment samples
for δ13Corg were pre-treated with 50 ml of HCl and 50 ml of deionized water and
dried at 60ordmC for 4 hr to remove carbonates (Harris et al 2011)
Total Carbon (TC) Total Organic Carbon (TOC) Total Inorganic Carbon (TIC)
and Total Sulphur (TS) were measured every cm with a LECO SC 144 DR at IPE-
CSIC XRF measurements were carried out every 4 mm in MAT11-1A-1G core using
an AVAATECH X-ray Fluorescence II core scanner at the University of Barcelona
(Spain) Results are expressed as element intensities in counts per second (cps)
Tube voltage was operated at 30 kV and 10 kV to obtain the abundances of 15
elements (Al Si S Cl K Ca Ti V Mn Fe Br Rb Sr Y Zr) with an average at
least of 1600 cps (less for Br=1000)
Biogenic silica content mineralogy and grain size were measured every 4
cm Biogenic silica was measured following Mortlock and Froelich (1989) and
Bernaacuterdez et al (2005) using an Auto Analyzer Technicon AAII for dissolved silicate
analysis Mineralogy was analyzed with a Siemens D-500 X-ray diffractometer (Cu
kα 40 kV 30 mA graphite monochromator) at the ICTJA-CSIC (Spain) Grain size
analyses were performed in a Beckmann Coulter LS 13 320 Particle Size Analyzer
57
at the IPE-CSIC The samples were classified according to textural classes as
follows clay (lt2 μm) silt (20-2 microm) and sand (gt2 microm) fractions
The age-depth model for the Laguna Matanzas sedimentary sequence was
constructed using 210Pb and 137Cs dating techniques (MAT13-4A) as well as two 14C
AMS dates on bulk sediment samples (MAT11-6A fig 2) We dated the dissolved
inorganic carbon (DIC) in the water column and no significant reservoir effect is
present in the modern-day water column (10454 + 035 pcmc Table 2) An age-
depth model was obtained with the Bacon R package to estimate the deposition
rates and associated age uncertainties along the core (Blaauw and Christen 2011)
To estimate LUCC of Laguna Matanzasacutes watershed we use satellite images
Landsat MSS for 1975 Landsat TM for 1989 and Landsat OLI for 2016 all taken in
summer or autumn (Table 1) We performed supervised classification of land uses
(maximum likelihood algorithm) for each year (1975 1989 and 2016) and the results
were mapped using software ArcGIS 102 in 2017
Satellite Images Acquisition Date
Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat OLI 20160404 30 m
Table 1 Landsat imagery
Surface water samples were filtered for obtained particulate organic matter In
addition soil samples from the main land usecover present in the Laguna Matanzas
watershed were collected Elemental C N and their corresponding isotopes from
POM and soil were obtained at the LABASI and used here as end members
Daily precipitation at Santo Domingo (33ordm39`S 71ordm3 6`W)- the nearest weather
station to Laguna Matanzasndash was compiled using the redPrec R package (Fig1 d
Serrano-Notivoli et al 2017 Sarricolea et al 2019) To extend our precipitation
58
reconstruction back to 1824 we correlated this dataset with that available for
Santiago The Santiago data was compiled from data published in the Anales of
Universidad of Chile (1850) for the years 1824 to 1849 Almeyda (1940) for the years
1849 to 1864 and the Quinta Normal series from 1866 to the present (Direccioacuten
Meteoroloacutegica de Chile) We generated a linear regression model between the
presentday Santo Domingo station and the compiled Santiago data with a Pearson
coefficient of 087 and p-valuelt 001
Acknowledgments This research was funded by grants CONICYT AFB170008
to the Institute of Ecology and Biodiversity (IEB) and FONDECYT 1150763 (to CL)
Doctoral grant Becas Chile 21150224 MEDLANT (Spanish Ministry of Economy
and Competitiveness grant CGL2016-76215-R) Additional funding was provided
by the Laboratorio Internacional de Cambio Global (LINCGlobal PUC-CSIC) We
thank R Lopez E Royo and M Gallegos for help with sample analyses We thank
the Laboratory of Biogeochemistry and Applied Stable Isotopes (LABASI) of the
Department of Ecology (PUC) for sample analyses
References
Albert F 1900 Las dunas o sean arenas volantes voladeros arenas muertas invasion de las arenas playas I meacutedanos del centro de Chile comprendiendo el litoral desde el liacutemite norte de la provincia de Aconcagua hasta el liacutemite sur de la de Arauco Memorias cientiacuteficas I Lit
Anderson NJ W D Fritz SC 2009 Holocene carbon burial by lakes in SW
Greenland Glob Chang Biol httpsdoiorg101111j1365-2486200901942x
Armesto J Villagraacuten C Donoso C 1994 La historia del bosque templado
chileno Ambient y Desarro 66 66ndash72 httpsdoiorg101007BF00385244
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A
59
historical framework for land cover change in southwestern South America in the past 15000 years Land use policy 27 148ndash160 httpsdoiorg101016jlandusepol200907006
Arnell NW Gosling SN 2013 The impacts of climate change on river flow
regimes at the global scale J Hydrol 486 351ndash364 httpsdoiorg101016JJHYDROL201302010
Bernaacuterdez P Prego R Franceacutes G Gonzaacutelez-Aacutelvarez R 2005 Opal content in
the Riacutea de Vigo and Galician continental shelf Biogenic silica in the muddy fraction as an accurate paleoproductivity proxy Cont Shelf Res httpsdoiorg101016jcsr200412009
Blaauw M Christen JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474 httpsdoiorg10121411-BA618
Brush GS 2009 Historical land use nitrogen and coastal eutrophication A
paleoecological perspective Estuaries and Coasts 32 18ndash28 httpsdoiorg101007s12237-008-9106-z
Camarero L Catalan J 2012 Atmospheric phosphorus deposition may cause
lakes to revert from phosphorus limitation back to nitrogen limitation Nat Commun httpsdoiorg101038ncomms2125
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego
R Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and environmental change from a high Andean lake Laguna del Maule central Chile (36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Catalan J Pla-Rabeacutes S Wolfe AP Smol JP Ruumlhland KM Anderson NJ
Kopaacuteček J Stuchliacutek E Schmidt R Koinig KA Camarero L Flower RJ Heiri O Kamenik C Korhola A Leavitt PR Psenner R Renberg I 2013 Global change revealed by palaeolimnological records from remote lakes A review J Paleolimnol httpsdoiorg101007s10933-013-9681-2
Chen Y Y Huang W Wang W H Juang J Y Hong J S Kato T amp
Luyssaert S (2019) Reconstructing Taiwanrsquos land cover changes between 1904 and 2015 from historical maps and satellite images Scientific Reports 9(1) 3643 httpsdoiorg101038s41598-019-40063-1
Contreras S Werne J P Araneda A Urrutia R amp Conejero C A (2018)
Organic matter geochemical signatures (TOC TN CN ratio δ 13 C and δ 15 N) of surface sediment from lakes distributed along a climatological gradient on the western side of the southern Andes Science of The Total Environment 630 878-888 httpsdoiorg101016jscitotenv201802225
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia UC
60
de V 2014 ELEMENTOS DE LA HISTORIA NATURAL DEL An Mus Hist Natulas Vaplaraiso 27 51ndash67
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-010-9453-1
Diebel M Vander Zanden MJ 201209 Nitrogen stable isotopes in streams
stable isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19 1127ndash1134 httpsdoiorg10189008-03271
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Elser J 2011 A world awash with nitrogen Science 334 1504ndash1506 Espizua LE Pitte P 2009 The Little Ice Age glacier advance in the Central
Andes (35degS) Argentina Palaeogeogr Palaeoclimatol Palaeoecol 281 345ndash350 httpsdoiorg101016JPALAEO200810032
Figueroa JA Castro S A Marquet P A Jaksic FM 2004 Exotic plant
invasions to the mediterranean region of Chile causes history and impacts Invasioacuten de plantas exoacuteticas en la regioacuten mediterraacutenea de Chile causas historia e impactos Rev Chil Hist Nat 465ndash483 httpsdoiorg104067S0716-078X2004000300006
Fowler D Coyle M Skiba U Sutton MA Cape JN Reis S Sheppard
LJ Jenkins A Grizzetti B Galloway N Vitousek P Leach A Bouwman AF Butterbach-bahl K Dentener F Stevenson D Amann M Voss M 2013 The global nitrogen cycle in the twenty- first century Philisophical Trans R Soc B httpsdoiorghttpdxdoiorg101098rstb20130164
Frank D Reichstein M Bahn M Thonicke K Frank D Mahecha MD
Smith P van der Velde M Vicca S Babst F Beer C Buchmann N Canadell JG Ciais P Cramer W Ibrom A Miglietta F Poulter B Rammig A Seneviratne SI Walz A Wattenbach M Zavala MA Zscheischler J 2015 Effects of climate extremes on the terrestrial carbon cycle Concepts processes and potential future impacts Glob Chang Biol httpsdoiorg101111gcb12916
Fritz SC Anderson NJ 2013 The relative influences of climate and catchment
processes on Holocene lake development in glaciated regions J Paleolimnol httpsdoiorg101007s10933-013-9684-z
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
61
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS) implications for past sea level and environmental variability J Quat Sci 32 830ndash844 httpsdoiorg101002jqs2936
Galloway JN Townsend AR Erisman JW Bekunda M Cai Z Freney JR
Martinelli LA Seitzinger SP Sutton MA 2008 Transformation of the Nitrogen Cycle Science (80- ) 320 889ndash892 httpsdoiorg101126science1136674
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924 httpsdoiorg104319lo20095430917
Garciacutea-Ruiz JM 2010 The effects of land uses on soil erosion in Spain A
review Catena httpsdoiorg101016jcatena201001001
Garciacutea-Ruiz JM Loacutepez-Moreno JI Lasanta T Vicente-Serrano SM
Gonzaacutelez-Sampeacuteriz P Valero-Garceacutes BL Sanjuaacuten Y Begueriacutea S Nadal-Romero E Lana-Renault N Goacutemez-Villar A 2015 Los efectos geoecoloacutegicos del cambio global en el Pirineo Central espantildeol una revisioacuten a distintas escalas espaciales y temporales Pirineos httpsdoiorg103989Pirineos2015170005
Garciacutea-Ruiz JM Nadal-Romero E Lana-Renault N Begueriacutea S 2013
Erosion in Mediterranean landscapes Changes and future challenges Geomorphology 198 20ndash36 httpsdoiorg101016jgeomorph201305023
Garreaud RD Vuille M Compagnucci R Marengo J 2009 Present-day
South American climate Palaeogeogr Palaeoclimatol Palaeoecol httpsdoiorg101016jpalaeo200710032
Gayo EM Latorre C Jordan TE Nester PL Estay SA Ojeda KF
Santoro CM 2012 Late Quaternary hydrological and ecological changes in the hyperarid core of the northern Atacama Desert (~21degS) Earth-Science Rev httpsdoiorg101016jearscirev201204003
Ge Y Zhang K amp Yang X (2019) A 110-year pollen record of land use and land
cover changes in an anthropogenic watershed landscape eastern China Understanding past human-environment interactions Science of The Total Environment 650 2906-2918 httpsdoiorg101016jscitotenv201810058
Goyette J Bennett EM Howarth RW Maranger R 2016 Global
Biogeochemical Cycles 1000ndash1014 httpsdoiorg1010022016GB005384Received
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and
oxygen isotope fractionation during dissimilatory nitrate reduction by
62
denitrifying bacteria 53 2533ndash2545 httpsdoiorg10230740058342
Gruber N Galloway JN 2008 An Earth-system perspective of the global
nitrogen cycle Nature 451 293ndash296 httpsdoiorg101038nature06592
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater eutrophic lake Limnol Oceanogr 51 2837ndash2848 httpsdoiorg104319lo20065162837
Gurnell AM Petts GE Hannah DM Smith BPG Edwards PJ Kollmann
J Ward J V Tockner K 2001 Effects of historical land use on sediment yield from a lacustrine watershed in central Chile Earth Surf Process Landforms 26 63ndash76 httpsdoiorg1010021096-9837(200101)261lt63AID-ESP157gt30CO2-J
Heathcote A J et al Large increases in carbon burial in northern lakes during the
Anthropocene Nat Commun 610016 doi 101038ncomms10016 (2015) Hollander DJ Smith MA 2001 Microbially mediated carbon cycling as a
control on the δ13C of sedimentary carbon in eutrophic Lake Mendota (USA) New models for interpreting isotopic excursions in the sedimentary record Geochim Cosmochim Acta httpsdoiorg101016S0016-7037(00)00506-8
Holtgrieve GW Schindler DE Hobbs WO Leavitt PR Ward EJ Bunting
L Chen G Finney BP Gregory-Eaves I Holmgren S Lisac MJ Lisi PJ Nydick K Rogers LA Saros JE Selbie DT Shapley MD Walsh PB Wolfe AP 2011 A coherent signature of anthropogenic nitrogen deposition to remote watersheds of the Northern Hemisphere Science (80) 334 1545ndash1548 httpsdoiorg101126science1212267
Hooper DU Adair EC Cardinale BJ Byrnes JEK Hungate BA Matulich
KL Gonzalez A Duffy JE Gamfeldt L Connor MI 2012 A global synthesis reveals biodiversity loss as a major driver of ecosystem change Nature httpsdoiorg101038nature11118
Horvatinčić N Sironić A Barešić J Sondi I Krajcar Bronić I Borković D
2016 Mineralogical organic and isotopic composition as palaeoenvironmental records in the lake sediments of two lakes the Plitvice Lakes Croatia Quat Int httpsdoiorg101016jquaint201701022
Howarth RW 2004 Human acceleration of the nitrogen cycle Drivers
consequences and steps toward solutions Water Sci Technol httpsdoiorg1010382Fscientificamerican0490-56
Jenny B Valero-Garceacutes BL Urrutia R Kelts K Veit H Appleby PG Geyh
M 2002 Moisture changes and fluctuations of the Westerlies in Mediterranean Central Chile during the last 2000 years The Laguna Aculeo record (33deg50primeS) Quat Int 87 3ndash18 httpsdoiorg101016S1040-
63
6182(01)00058-1
Jenny B Wilhelm D Valero-Garceacutes BL 2003 The Southern Westerlies in
Central Chile Holocene precipitation estimates based on a water balance model for Laguna Aculeo (33deg50primeS) Clim Dyn 20 269ndash280 httpsdoiorg101007s00382-002-0267-3
Leng M J Lamb A L Heaton T H Marshall J D Wolfe B B Jones M D
amp Arrowsmith C 2006 Isotopes in lake sediments In Isotopes in palaeoenvironmental research (pp 147-184) Springer Dordrecht
Le Roux JP Nielsen SN Kemnitz H Henriquez Aacute 2008 A Pliocene mega-
tsunami deposit and associated features in the Ranquil Formation southern Chile Sediment Geol 203 164ndash180 httpsdoiorg101016jsedgeo200712002
Lehmann M Bernasconi S Barbieri a McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66 3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Lenihan JM Drapek R Bachelet D Neilson RP 2003 Climate change
effects on vegetation distribution carbon and fire in California Ecol Appl httpsdoiorg101890025295
McLauchlan K K Gerhart L M Battles J J Craine J M Elmore A J
Higuera P E amp Perakis S S (2017) Centennial-scale reductions in nitrogen availability in temperate forests of the United States Scientific reports 7(1) 7856 httpsdoi101038s41598-017-08170-z
Martel-Cea A Maldonado A Grosjean M Alvial I de Jong R Fritz SC von
Gunten L 2016 Late Holocene environmental changes as recorded in the sediments of high Andean Laguna Chepical Central Chile (32ordmS 3050 masl) Palaeogeogr Palaeoclimatol Palaeoecol 461 44ndash54 httpsdoiorg101016jpalaeo201608003
Martiacuten-Foreacutes I Casado MA Castro I Ovalle C Pozo A Del Acosta-Gallo
B Saacutenchez-Jardoacuten L Miguel JM De 2012 Flora of the mediterranean basin in the chilean ESPINALES Evidence of colonisation Pastos 42 137ndash160
Marzecovaacute A Mikomaumlgi a Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54 httpsdoiorg103176eco2011105
Matesanz S Valladares F 2014 Ecological and evolutionary responses of
Mediterranean plants to global change Environ Exp Bot httpsdoiorg101016jenvexpbot201309004
64
McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE 2007
Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash1643 httpsdoiorg1011770959683613496289
Menzel P Gaye B Wiesner MG Prasad S Stebich M Das BK Anoop A
Riedel N Basavaiah N 2013 Influence of bottom water anoxia on nitrogen isotopic ratios and amino acid contributions of recent sediments from small eutrophic Lonar Lake central India Limnol Oceanogr 58 1061ndash1074 httpsdoiorg104319lo20135831061
Meyers PA Teranes JL 2001 Sediment organic matter in Tracking
Environmental Change Using Lake Sediments Volume 2 Physical and Geochemical Methods httpsdoiorg1018971551-5028(1999)018lt0231SOMCAAgt23CO2
Michelutti N Wolfe AP Cooke CA Hobbs WO Vuille M Smol JP 2015
Climate change forces new ecological states in tropical Andean lakes PLoS One httpsdoiorg101371journalpone0115338
Muntildeoz-Rojas M De la Rosa D Zavala LM Jordaacuten A Anaya-Romero M
2011 Changes in land cover and vegetation carbon stocks in Andalusia Southern Spain (1956-2007) Sci Total Environ httpsdoiorg101016jscitotenv201104009
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Poraj-Goacuterska A I Żarczyński M J Ahrens A Enters D Weisbrodt D amp
Tylmann W (2017) Impact of historical land use changes on lacustrine sedimentation recorded in varved sediments of Lake Jaczno northeastern Poland Catena 153 182-193 httpdxdoiorg101016jcatena201702007
Pongratz J Reick CH Raddatz T Claussen M 2010 Biogeophysical versus
biogeochemical climate response to historical anthropogenic land cover change Geophys Res Lett 37 1ndash5 httpsdoiorg1010292010GL043010
Prudhomme C Giuntoli I Robinson EL Clark DB Arnell NW Dankers R
Fekete BM Franssen W Gerten D Gosling SN Hagemann S Hannah DM Kim H Masaki Y Satoh Y Stacke T Wada Y Wisser D 2014 Hydrological droughts in the 21st century hotspots and uncertainties from a global multimodel ensemble experiment Proc Natl Acad Sci httpsdoiorg101073pnas1222473110
65
Ruumlhland KM Paterson AM Smol JP 2015 Lake diatom responses to
warming reviewing the evidence J Paleolimnol httpsdoiorg101007s10933-
015-9837-3
Sarricolea P Meseguer-Ruiz Oacute Serrano-Notivoli R Soto M V amp Martin-Vide
J (2019) Trends of daily precipitation concentration in Central-Southern Chile Atmospheric research 215 85-98 httpsdoiorg101016jatmosres201809005
Sernageomin 2003 Mapa geologico de chile version digital Publ Geol Digit Serrano-Notivoli R de Luis M Begueriumlia S 2017 An R package for daily
precipitation climate series reconstruction Environ Model Softw httpsdoiorg101016jenvsoft201611005
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Stine S 1994 Extreme and persistent drought in California and Patagonia during
mediaeval time Nature 369 546ndash549 httpsdoiorg101038369546a0
Stocker TF Dahe Q Plattner G-K Alexander L V Allen SK Bindoff NL
Breacuteon F-M Church JA Cubash U Emori S Forster P Friedlingstein P Talley LD Vaughan DG Xie S-P 2013 Technical Summary in Climate Change 2013 The Physical Science Basis Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change httpsdoiorg101017 CBO9781107415324005
Syvitski JPM Voumlroumlsmarty CJ Kettner AJ Green P 2005 Impact of humans
on the flux of terrestrial sediment to the global coastal ocean Science 308(5720) 376-380 httpsdoiorg101126science1109454
Thomas RQ Zaehle S Templer PH Goodale CL 2013 Global patterns of
nitrogen limitation Confronting two global biogeochemical models with observations Glob Chang Biol httpsdoiorg101111gcb12281
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of different trophic status J Paleolimnol 47 693ndash706 httpsdoiorg101007s10933-012-9593-6
Urbina Carrasco MX Gorigoitiacutea Abbott N Cisternas Vega M 2016 Aportes a
la historia siacutesmica de Chile el caso del gran terremoto de 1730 Anuario de Estudios Americanos httpsdoiorg103989aeamer2016211
Vargas G Ortlieb L Chapron E Valdes J Marquardt C 2005 Paleoseismic
inferences from a high-resolution marine sedimentary record in northern Chile
66
(23degS) Tectonophysics httpsdoiorg101016jtecto200412031 399(1-4) 381-398httpsdoiorg101016jtecto200412031
Valero-Garceacutes BL Latorre C Morelliacuten M Corella P Maldonado A Frugone M Moreno A 2010 Recent Climate variability and human impact from lacustrine cores in central Chile 2nd International LOTRED-South America Symposium Reconstructing Climate Variations in South America and the Antarctic Peninsula over the last 2000 years Valdivia p 76 (httpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-yearshttpwwwpagesunibechproductsmeeting-products1024-2nd-international-lotred-south-america-symposium-reconstructing-climate-variations-in-south-america-and-the-antarctic-peninsula-over-the-last-2000-years
Vicente-Serrano SM Gouveia C Camarero JJ Begueria S Trigo R
Lopez-Moreno JI Azorin-Molina C Pasho E Lorenzo-Lacruz J Revuelto J Moran-Tejeda E Sanchez-Lorenzo A 2013 Response of vegetation to drought time-scales across global land biomes Proc Natl Acad Sci httpsdoiorg101073pnas1207068110
Villa Martiacutenez R P 2002 Historia del clima y la vegetacioacuten de Chile Central
durante el Holoceno Una reconstruccioacuten basada en el anaacutelisis de polen sedimentos microalgas y carboacuten PhD
Villa-Martiacutenez R Villagraacuten C Jenny B 2003 Pollen evidence for late -
Holocene climatic variability at laguna de Aculeo Central Chile (lat 34o S) The Holocene 3 361ndash367
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Human alteration of the global nitrogen cycle sources and consequences Ecological applications 7(3) 737-750 httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M Rein B Urrutia R amp Appleby P (2009) A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 The Holocene 19(6) 873-881 httpsdoiorg1011770959683609336573
von Gunten L Grosjean M Eggenberger U Grob P Urrutia R Morales A
2009 Pollution and eutrophication history AD 1800-2005 as recorded in sediments from five lakes in Central Chile Glob Planet Change 68 198ndash208 httpsdoiorg101016jgloplacha200904004
Widory D Kloppmann W Chery L Bonnin J Rochdi H Guinamant JL
2004 Nitrate in groundwater An isotopic multi-tracer approach J Contam Hydrol 72 165ndash188 httpsdoiorg101016jjconhyd200310010
67
Woodward CA Potito AP Beilman DW 2012 Carbon and nitrogen stable isotope ratios in surface sediments from lakes of western Ireland Implications for inferring past lake productivity and nitrogen loading J Paleolimnol 47 167ndash184 httpsdoiorg101007s10933-011-9568-z
68
Supplementary material
Facie Name Description Depositional Environment
F1 Organic-rich
mud
Massive to banded black
organic - rich (TOC up to 14 )
mud with aragonite in dm - thick
layers Slightly banded intervals
contain less OM (TOClt4) and
aragonite than massive
intervals High MnFe (oxic
bottom conditions) High CaTi
BrTi and BioSi (up to 5)
Distal low energy environment
high productivity well oxygenated
and brackish waters and relative
low lake level
F2 Massive to
banded silty clay
to fine silt
cm-thick layers mostly
composed by silicates
(plagioclase quartz cristobalite
up to 65 TOC mean=23)
Some layers have relatively high
pyrite content (up to 25) No
carbonates CaTi BrTi and
BioSi (mean=48) are lower
than F1 higher ZrTi (coarser
grain size)
Deposition during periods of
higher sediment input from the
watershed
69
F3 Banded to
laminated light
brown silty clay
cm-thick layers mostly
composed of clay minerals
quartz and plagioclase (up to
42) low organic matter
(TOC mean=13) low pyrite
and BioSi content
(mean=46) and some
aragonite
Flooding events reworking
coastal deposits
F4 Laminated
coarse silts
Thin massive layers (lt2mm)
dominated by silicates Low
TOC (mean=214 ) BrTi
(mean=002) MnFe (lt02)
TIC (lt034) BioSi
(mean=46) and TS values
(lt064) and high ZrTi
Rapid flooding events
transporting material mostly
from within the watershed
F5 Breccia with
coarse silt
matrix
A 17 cm thick (80-97 cm
depth) layer composed by
irregular mm to cm-long ldquosoft-
clastsrdquo of silty sediment
fragments in a coarse silt
matrix Low CaTi BrTi and
MnFe ratios and BioSi
Rapid high energy flood
events
70
(mean=43) and high ZrTi
(gt018)
Table Sedimentological and compositional characteristics of Laguna Matanzas
facies
71
CAPIacuteTULO 2 STABLE ISOTOPES TRACK LAND USE AND COVER
CHANGES IN A MEDITERRANEAN LAKE IN CENTRAL CHILE OVER THE
LAST 600 YEARS
72
Stable isotopes track land use and cover changes in a mediterranean lake in
central Chile over the last 600 years
Magdalena Fuentealbaabc Claudio Latorreabc Matiacuteas Frugone-Aacutelvarezabc Pablo
Sarricolead Leonardo Villaciacutese M Laura Carrevedob Blas Valero-Garceacutescg
a Departamento de Ecologiacutea Pontificia Universidad Catoacutelica de Chile Alameda 340 Santiago Chile
b Institute of Ecology and Biodiversity (IEB) Las Palmeras 3425 Nuntildeoa Santiago Chile
c Laboratorio Internacional de Cambio Global LINCGlobal PUC-CSIC Spain
d Departamento de Geografiacutea Universidad de Chile Marcoleta 250 Santiago Chile
e Departamento de Ecologiacutea Universidad de ChileLas Palmeras 3425 Nuntildeoa Santiago Chile
f Instituto Pirenaico de Ecologiacutea (IPE-CSIC) Avenida Montantildeana 1005 Zaragoza 50059 Spain
Corresponding authors E-mail addresses clatorrebiopuccl magdalenafuentealbagmailcom
Key words Anthropocene lake sediment δsup1⁵N δsup1sup3C Great Acceleration organic
geochemistry watershedndashlake system Stable Isotope Analyses land usecover
change Nitrogen cycle mediterranean ecosystems central Chile
73
Abstract
Nutrient fluxes in many aquatic ecosystems are currently being overridden by
anthropic controls especially since the industrial revolution (mid-1800s) and the
Great Acceleration (mid-1900s) Land use and cover changes (LUCC) alter the
availability and fluxes of nutrients such as nitrogen that are transferred via runoff
and groundwater into lakes By altering lake productivity and trophic status these
changes are often preserved in the sedimentary record Here we use
geolimnological proxies and stable isotope analyses (δsup1⁵N δsup13C) on lake sediments
to reconstruct the lake-watershed nutrient transfer in a central Chilean lake (Lago
Vichuqueacuten) over the past 600 years We compare our multiproxy record from recent
lake sediments to the soilvegetation relationship across the watershed as well as
land usecover changes from 1975 to 2014 derived from satellite images Our results
show that lake sediment δsup1⁵N values increased with meadow cover but decreased
with tree plantations suggesting increased nitrogen retention when trees dominate
the watershed Our δsup1⁵N record from central Chile can thus be interpreted as a proxy
for nutrient availability over the last 600 years mainly derived from land use changes
coupled with climate drivers Although variable sources of organic matter and in situ
fractionation often hinder straightforward environmental interpretations of stable N
isotope signatures our study shows that lake sediment δsup1⁵N can be a useful tool for
assessing the contribution of past human activities in nutrient and nitrogen cycling
1 Introduction
Nitrogen (N) is a limiting nutrient in terrestrial and aquatic ecosystems (Vitousek
et al 1997) Changes in its availability can drive eutrophication and increase
pollution in these ecosystems (McLauchlan et al 2013) Although recent human
74
impacts on the global N cycle have been significant the consequences of increased
anthropogenic N on these ecosystems are unclear (Gerhart and Mclauchlan 2014
Battye et al 2017) Isotopic signatures are key to understand N dynamics in lakes
nevertheless in situ andor diagenetic fractionation along with multiple sources of
organic matter (OM) often hinder straightforward environmental interpretations from
isotopes Monitoring δ15N and δ13C values as components of the N cycle
specifically those related to the link between terrestrial and aquatic ecosystems can
help differentiate between effects from processes versus sources in stable isotope
values (eg from Particulate Organic Matter -POM- soil and vegetation) and
improve how we interpret variations in δ15N (and δ13C) values at longer temporal
scales
The main processes controlling stable N isotopes in bulk lake OM are source
lake productivity diagenesis and denitrification (Talbot 2001 Leng et al 2006
Fariacuteas et al 2009 Torres et al 2012 Brahney et al 2014) N sources depend on
contributions from the watershed (ie soil and biomass) the transfer of atmospheric
N to the lake (ie atmospheric N2 fixation) and nutrient recycling (Leng et al 2006)
Atmospheric N2 (isotopically defined as δ15N =0) is fixed by cyanobacteria with
minimal fractionation resulting in δ15NOM values that fluctuate from -2 to +2permil (Fogel
and Cifuentes 1993 Talbot 2001 Elliot and Brush 2006) Besides N2 fixation by
cyanobacteria phytoplankton species assimilate dissolved inorganic nitrogen (DIN)
and values typically range from +7 to +10permil (Xu et al 2016 Meyers et al 2003) In
addition seasonal changes in POM occur in the lake water column Gu et al (2006)
sampled the water column of Lake Wauberg (Florida USA) during a hydrologic year
and found a higher development of N fixing species during the summer A major
factor behind this increase are human activities in the watershed which control the
75
inputs of the sediment and nutrients to the lake (Woodward et al 2011) Some
studies have shown higher δ15N values in lake sediments from watersheds that are
highly affected by human activities (Bruland and Mackenzie 2010 Woodward et al
2011) Syntenic fertilizers displayed δ15N values between -4 to +4permil cattle manure
around +8 to +22permil and surface and groundwater OM between 0 to +10permil (Elliott
and Brush 2006 Leng et al 2006) Although relatively low δ15N values from
fertilizers constitute major N input to human-altered watersheds the elevated loss
of 14N via volatilization of ammonia and denitrification leaves the remaining total N
input enriched in 15N (Bruland and Mackenzie 2010)
In addition to the different sources and variations in lake productivity early
diagenesis at the sedimentndashwater interface in the sediment can further alter
sediment OM isotope values (Brahney et al 2014 Galman et al 2009) During
diagenetic loss of N in lacustrine systems kinetic fractionation occurs and the
remaining N is enriched in 15N leading to higher δ15N values (Galman et al 2006
Talbot 2001) Denitrification tends to enrich lake sediments in 15N due to the
assimilation of nitrates by denitrifying bacteria (Granger et al 2008) and is more
prevalent in anoxic environments (Brahney et al 2016 Lehman et al 2002)
Carbon isotopes in lake sediments can also provide useful information about
paleoenvironmental changes OM origin and depositional processes (Meyers et al
2003) Allochthonous organic sources (high CN ratios) produce isotope values
similar to values from catchment vegetation Autochthonous organic matter (low CN
ratio) is influenced by fractionation both in the lake and the watershed leading up to
carbon assimilation by lacustrine biota (Woodward et al 2011) Changes in
productivity greatly affect δ13COM values (Torres et al 2012) as during C uptake
plankton preferentially assimilate 12C leaving the dissolved inorganic carbon (DIC)
76
pool enriched in 13C (Gu et al 2006) with phytoplankton averaging ca 20 permil lower
than the respective δ13CDIC values (Leng et al 2006) Under conditions of low to
moderate primary productivity plankton preferentially uptake the lighter 12C
resulting in low δ13C values displayed by the OM (Torres et al 2012) Conversely
during high primary productivity phytoplankton will uptake 12C until its depletion and
is then forced to assimilate the heavier isotope resulting in an increase in δ13C
values Higher productivity in C-limited lakes due to slow water-atmosphere
exchange of CO2 also results in high δ13C values (Galman et al 2009) In these
cases algae are forced to uptake dissolved bicarbonate with δ13C values between
7 to 10permil higher than those of the atmospheric CO2 dissolved in water (Sun et al
2016 Torres et al 2012 Galman et al 2009)
Stable isotope analyses from lake sediments are thus useful tools to
reconstruct shifts in lake-watershed dynamics caused by changes in limnological
parameters and LUCC Our knowledge of the current processes that can affect
stable isotope signals in a watershed-lake system is limited however as monitoring
studies are scarce Besides in order to use stable isotope signatures to reconstruct
past environmental changes we require a multiproxy approach to understand the
role of the different variables in controlling these values Hence in this study we
carried out a detailed survey of current N dynamics in a coastal central Chilean lake
(Lago Vichuqueacuten) and a multiproxy study of a sediment record spanning the last
600 years The characterization of the recent changes in the watershed since 1970s
is based on satellite images to compare recent changes in the lake and assess how
these are related with climate variability and an ever increasing human footprint
(Lara et al 2012 Gallardo et al 2015 Steffen et al 2015) Our main goal is to
investigate how stable isotope values from lake sediment reflect changes in the lake
77
ndash watershed system during periods of high watershed disruption (eg Spanish
Conquest late XIX century Great Acceleration) and recent climate change (eg
Little Ice Age and current global warming)
2 Study Site
Lago Vichuqueacuten (34ordm49`S 72ordm04W 4 m asl 30 m of depth Fig 1) is a
mesotrophic to eutrophic lake with dominant anoxic bottom conditions that is
stratified year around (Frugone-Aacutelvarez et al 2017) The main inlets are the
Vichuqueacuten and Huintildee rivers and the only outlet is the Llico estuary that drains into
the Pacific Ocean High tides can sporadically shift the flow direction of the Llico
estuary which increases the marine influence in the lake Dune accretion gradually
limited ocean-lake connectivity until the estuary was almost completely closed off
by ca 10 ky BP and the current lake regime formed (Frugone-Aacutelvarez et al 2017)
The area is characterized by a mediterranean climate with cold-wet winters and
hot-dry summers and an annual precipitation of ~650 mm and a mean annual
temperature of 15ordmC During the austral winter months (June - August) precipitation
is modulated by north-west shifts of the South Pacific Anticyclone (SPA) followed by
an increased frequency of storm fronts stemming off the South Westerly Winds
(SWW) A strengthened SPA during austral summers (December - March) which
are typically dry and warm blocks the northward migration of storm tracks stemming
off the SWW (Fig1 Frugone-Aacutelvarez et al 2017)
78
Figure 1 a) The location of the Lago Vichuqueacuten in central Chile b) Map of land
uses and cover in 2016 c) Ombrothermic diagram mediterranean climates are
characterized by cold-wet winters with surplus moisture from June to August and
hot-dry summers d) Lake bathymetry showing location of cores and water sampling
sites used in this study
Although major land cover changes in the area have occurred since 1975 to the
present as the native forests were replaced by tree (Monterey pine and eucalyptus)
plantations the region was settled before the Spanish conquest (Frugone-Alvarez
et al 2018) Historical documents indicate that Vichuqueacuten was occupied by a
Mitimaes colony as part of the Inka empire (Odone 1998) Unlike other Andean
areas during the Inka period the introduction of corn and quinoa in the Vichuqueacuten
watershed do not seem to have intensified land use The Spanish colonial period in
Chile lasted from 1542 CE to the independence in 1810 CE The first historical
document (1550 CE) shows that the areas around Vichuqueacuten were settled by the
Spanish early on as the Vichuqueacuten watershed was given as an ldquoencomiendardquo
system to Juan Cuevas The ldquoencomiendardquo system provided the owners with land
79
and indigenous people to work but also the introduction of wheat wine cattle
grazing and logging of native forests for lumber extraction and increasing land for
agricultural activities (Odone et al 1998 Armesto et al 2010) During the 19th
century (the Republic) the export of wheat to Australia and Canada generated
intensive changes in land cover use The town of Vichuqueacuten became the regional
capital and plans to build a maritime port in the bay of Vichuqueacuten were drawn
However the fall of international markets in 1880 paralyzed these plans During the
20th century the plantation of trees (Pinus radiata and Eucalyptus globulus) in areas
cleared of native forests was subsidized by two national laws DFL nordm265 (1931) and
DFL nordm 701 (1974) both of which provided funds for such plantations During the last
decades the urbanization with summer vacation homes along the shorelines of
Lago Vichuqueacuten has greatly increased Extensive lake eutrophication is currently a
large environmental problem (EULA 2008)
3 Methods
Coring campaigns were organized from 2011 to 2015 (Fig 1d) We recovered
12 short cores and 4 long cores (up to 14 m long) at several sites using a hammer-
modified UWITEC gravity corer Sediment cores (VIC11-2A 170 cm VIC13-2B 170
cm VIC13-2D 1200 cm and VIC15-1A 119 cm) were split photographed sub-
sampled and stored in the Pyrenean Institute of Ecology (IPE-CSIC Spain) Core
VIC13-2B was selected for detailed multiproxy analyses (including elemental
geochemistry C and N isotopes XRF and 14C dating) Stable isotope analyses
(δ13C δ15N) were performed at the Laboratory of Biogeochemistry and Applied
Stable Isotopes (LABASI-PUC Chile) Sediment samples for δ13Corg were pre-
treated with 50 ml of HCl and 50 ml of deionized water and dried at 60ordmC for 4 hr to
remove carbonates (Harris et al 2011) Isotope analyses were conducted using a
80
Delta V Advantage IRMS coupled to a Thermo Flash 2000 Elemental Analyzer via
a Conflo IV interface Isotope results are expressed in standard delta notation (δ)
and expressed in per mil (permil) relative to the Pee Dee Belemnite (Vienna Pee Dee
Belemnite) for C and to atmospheric N2 for N (Leng et al 2006) Total carbon (TC)
Total Organic Carbon (TOC) Total Inorganic Carbon (TIC) and Total Sulfur (TS)
were measured every 2 cm with a LECO SC 144 DR at the IPE-CSIC
An AVAATECH X-Ray Fluorescence II core scanner with a Rh X-ray tube from
the University of Barcelona was used to obtain XRF logs every 4 mm of resolution
Results are expressed as element intensities in counts per second (cps) Tube
voltage was operated at 30 kV and 10 kV to obtain the abundances of 18 elements
(Pb Al Si S Cl K Ca Ti V Mn Fe Co Zn Br Rb Sr Y Zr) with an average of
at least of 1800 cps (less for Pb=437 and Zn=934 cps) Pb was included due to
similar behavior with Co and Fe Element ratios were calculated to describe changes
in redox conditions (MnFe) endogenic productivity (BrTi) carbonate formation
(SrCa and CaTi) and sediment input (ZrTi) (Barreiro-Lostres et al 2014
Carrevedo et al 2015 Frugone-Aacutelvarez et al 2017 Kylander et al 2011 Moreno
et al 2007a)
Several campaigns were carried out to sample the POM from the water column
two per hydrologic year from November 2015 to August 2018 A liter of water was
recovered in three sites through to the lake two are from the shallower areas (with
samples taken at 2 and 5 m depth at each site) and one in the deeper central portion
(at 2 5 and 20 m depth) of the lake (Fig 1d) The samples were filtered using glass
fiber filters (25 m) and then frozen to obtain the stable carbon and nitrogen isotope
signal of lacustrine POM Additionally soil and vegetation samples from the
following communities native species meadow hydrophytic vegetation and
81
Monterrey pine (Pinus radiata) were taken for stable isotope analyses (see in
supplementary material)
The age model for the complete Lago Vichuqueacuten sedimentary sequence is
based on Frugone-Aacutelvarez et al (2017) with the last 1000 years based on
210Pb137Cs dating techniques in VIC15-1A and 14C AMS dates on bulk sediment
samples (Supplementary Table S1) The 14C measurements of lake water DIC show
a moderate reservoir effect of 180 plusmn 25 14C yr) The revised age-depth model used
here includes three more 14C AMS dates performed with the program Bacon to
establish the deposition rates along the core (Blaauw and Christen 2011 Fig 2)
The age-depth model indicates that average resolution between 0 to 87 cm is lt2
cm per year and from 88 to 170 cm it is lt47 cm per year
82
Figure 2 Chronological age-depth model for the Lago Vichuqueacuten sedimentary
sequence spanning the last 1000 yr (see also Frugone-Alvarez et al 2017)
To estimate land use changes in the watershed we use Landsat MSS images
for 1975 and Landsat TM for 1989 and 2014 all taken in either summer or autumn
(Table 1) We performed supervised classification of land uses (maximum likelihood
83
algorithm) for each year (1975 1989 and 2014) and results were mapped using
ArcGIS 102
Table 1 Images using for LUCC reconstruction
Source of LUCC
Acquisition
Date Resolution
Landsat MSS 19750322 60 m
Landsat TM 19890217 30 m
Landsat TM 19991226 30 m
CONAF 2009 30 m
Land cover Chile 2014 30 m
CONAF 2016 30 m
Previous Work on Lago Vichuqueacuten sedimentary sequence
The sediments are organic-poor dark brown to brown laminated silt with some
intercalated thin coarser clastic layers Lacustrine facies have been classified
according to elemental composition (TOC TS TIC and TN) grain size and
sedimentary textures (Frugone-Aacutelvarez et al 2017) Four main types of lacustrine
facies were identified in this short core Facies L1 is a laminated (1cm) black to dark
brown diatom-poor silt with relatively higher TOC percentages (TOC about 185)
TS (about of 18) and TOCTN (about 92) Facies L2 is characterized by a
homogeneous black organic-poor silty clay with low TOC (mean= 13) TS (mean=
13) TOCTN (mean= 88) values Facies L3 is a laminated (1cm) black organic-
poor (TOC mean 14) silts with higher TS (mean= 21) and TOCTN ratios
(mean= 88) Facies L3 is interpreted as ldquobaselinerdquo deposition on the distal areas
84
of Vichuqueacuten lake Mineralogy is dominated by mica (muscovite) clay minerals
(chlorite) and plagioclase (albite) Quartz and calcite occur at the top and pyrite
occurs in the lower part of the sequence Facies T is composed by massive banded
sediments 4 cm thick and coarse grain size It is interpreted as an instantaneous
depositional event (for more detail see Frugone-Aacutelvarez et al 2017) In this work
we identified four subunits based on geochemical and stable isotope signals
4 Results
41 Geochemistry and PCA analysis
High positive correlations exist between Al Si K and Ti (r = 078 ndash 096
supplementary Fig S1) and between Zn Zr Rb and Y (r = 072 - 096) which reflect
the silicate content of the watershed-derived sediments (Marzecoacuteva et al 2011) Zr
is commonly associated with minerals more abundant in coarser deposits Thus the
ZrTi profile could reflect grain size (Cuven et al 2010) showing higher variability
in the upper part of the Lake Vichuqueacuten sequence and in the alternation between
facies L1 and L2 Another group of elements made up of Co Fe and Pb displayed
positive correlations (r = 067ndash 097) and represents the input of heavy metals Br
Cl and Ca show a positive correlation (r = 049 ndash 06 p-value = 00001) BrTi ratio
is interpreted as a productivity indicator due to Br having a strong affinity with humic
and organic compounds (Marzecoacuteva et al 2011 Frugone-Aacutelvarez et al 2017) In
our Lago Vichuqueacuten sequence BrTi values are high from 170 to 132 cm and from
36 cm to the top of core where it shows several sharps peaks (Fig 3) The MnFe
ratio is indicative of lake bottom oxygenation (Naecher et al 2013) as under
reducing conditions Mn tends to become more mobile than Fe leading to a
decreased MnFe (Marzecoacuteva et al 2011) Several sharp peaks in MnFe occurred
85
from 127 to 120 cm and from 57 to 30 cm (MgFegt03) The silicate group and the
Br Cl Ca Mn group are negatively correlated (r= -012 and -066)
Principal Component Analysis (PCA) was undertaken on the XRF
geochemical data to investigate the main factors controlling sediment deposition in
Lago Vichuqueacuten (Fig 3) The four main eigenvectors explain 801 of the variance
(supplementary material Table S2) The principal component (PC1) explain 437
of the total of variance and grouped elements are associated with terrigenous input
to the lake Positive values of the biplot have been attributed to higher heavy metals
deposition (Pb Co and Fe) and negative values to higher silicate input (Al K Ti and
Si) in a high energy catchment environment (Zr) The PC2 explains 198 of the
total of variance and highlights the endogenic productivity in the lake The positive
loading is compound by Br Cl Ca and Sr and to a lesser extent by Y Zn Rb and
Mn The PC2 may explain both the precipitation of carbonates (Ca Sr) and biological
production (Br)
86
Figure 3 Principal Component Analysis of XRF geochemical measurements in
VIC13-2B Lago Vichuqueacuten lake sediments
42 Sedimentary units
Based on geochemical and stable isotope analysis we identified four
lithological subunits in the short core sedimentary sequence Our PCA analyses and
Pearson correlations pointed out which variables were better for characterizing the
subunits (Fig 4) in terms of endogenic productivity heavy metals and terrestrial
input The unit 2c (170-131 cm) is characterized by the occurrence of some clastic
layers (L2 from 160 and 150 cm) high δsup1⁵N values (mean= +59 plusmn 04permil) with
Magnetic Susceptibility (MS) BrTi ZrTi and SrCa values increase towards the top
Also it has relatively low δsup1sup3C values (mean= -265 plusmn 11permil) and CNmolar ratios
(mean=78 plusmn 04) The ZrTi show upwards increasing values reaching the highest
values of the sequence at the top of this unit suggesting a coarsening upward trend
and relatively higher depositional energy The MS trend also indicates higher
erosion in the watershed and enhanced delivery of ferromagnetic minerals likely
from soil particles particularly from 150 to 160 cm (Frugone-Aacutelvarez at al 2017)
The subunit 2b (130-118 cm) is also composed of black silts but it has the
lowest MS values of the whole sequence and its onset is marked by a sharp
decrease in ZrTi This subunit is marked by sharp peaks of MnFe (at 126 and 120
cmgt027) SrCa (at 125 and 120 cmgt033) CaTi (at 124 and 120 cmgt199) TOC
(about 26 at 130 124 122 and 118 cm) and δsup1⁵N (gt+67permil at 126 and 120 cm)
BrTi oscillates around low values (about 01) whereas the δsup1sup3C values range
between -262 and -282permil
87
The unit 2a (58-117 cm) shows increasing and then decreasing MS values
and displayed sharps peaks of δsup1⁵N (gt64 at 102 to 99 87 and 75 cm) and CN
(about 10 at 86 74 66 and 60 cm) SrCa (mean = 04 plusmn 013) BrTi (mean = 008
plusmn 002) MnFe (mean = 002 plusmn 001) and δsup1sup3C (mean = -260 plusmn 07permil) oscillate in
low values The upper part of this unit (72 ndash 57 cm) shows an increase of SrCa
(from 03 to 05)
The upper unit 1a (0 - 57 cm) starts with a fine clastic layer (T at 57 to 54
cm) interpreted as deposition during a high-energy event It is characterized by
lower BrTi (mean = 003 plusmn 002) TOC (mean = 12 plusmn 03) and δsup1sup3C (mean= -
266 plusmn 06permil) higher δsup1⁵N (mean = +55 plusmn 08 permil) This unit is made of alternating
fine (lt 1cm) black and dark brown laminae with carbonate (L1 facies) Recently
deposited sediments show decreasing MS values increasing TOC (mean= 15 plusmn
04 ) sharp peaks of BrTi (gt015 at 33 21 5 and 1 cm) and the lowest δsup1⁵N values
of the entire record (mean= 45 plusmn 06 permil) and δsup1sup3C values (mean= -277 plusmn 09 permil)
Heavy metal content increases abruptly in the upper section of the core (2 - 20 cm)
(peaks of FeTi CoTi and PbTi)
88
Figure 4 Sedimentary units of Lago Vichuqueacuten sedimentary sequence Selected
variables illustrate watershed input (Magnetic Susceptibility ZrTi CoTi and MnFe)
endogenic productivity (SrCa CaTi TS) organic matter source (BrTi TOC
CNmolar and stable isotope records (δ13Corg and δ15Nbulk)
43 Recent seasonal changes of particulate organic matter on water column
The average of δ15NPOM from the three sites across to the lake was +86 plusmn 58
permil oscillating widely from -14 to +193permil (Fig 5) Important seasonal differences
occur at 2 and 5 m depth with more negative values during summer (~53 plusmn 52 permil)
than winter (~122 plusmn 40permil) At 20 m depth δ15NPOM values exhibit small seasonal
ranges (mean = +10permil for winter and +87permil for summer) The δ13CPOM average was
-296 plusmn 33permil with slightly seasonal and water column depth differences However
more positive δ13CPOM values occurred during winter (mean=-290 plusmn 41permil) than in
summer (mean= -301 plusmn 19 permil) Like the δ15NPOM values CNPOM (mean= 83 plusmn 17)
displayed important seasonal and water depth differences Lower CNPOM ratios
89
occurred during summer at 2 and 5 m depth (mean= 95 plusmn 17) and higher and more
constant values during winter (mean = 73 plusmn 09) at the same depth At 20 m CNPOM
shows similar values in both winter (70) and summer (74)
Figure 5 Seasonal POM averages from samples taken of the Lago Vichuqueacuten
water column Samples were taken from 2015 to 2018 at 2 (n=16) 5 (n=14) and 20
(n=8) meters depth
44 Stable isotope values across the Lake Vichuqueacuten watershed
Figure 6 shows modern vegetation soil and sediment isotope values found for
the watershed Huintildee tributary and Vichuqueacuten lake The δsup1⁵NFoliar values from
meadow plantations and macrophytes have similar range values with a mean of
+89 plusmn50 permil (n=18) +74 plusmn 75 permil (n=5) +82 plusmn 36 permil (n=6) respectively Native
vegetation has overall lower δsup1⁵NFoliar values with a mean of 14 plusmn 21 permil (n=28 see
Table 2 in supplementary material) The δsup1sup3CFoliar from terrestrial samples exhibit
similar values across the different plant communities (tree plantation mean=-274 plusmn
13permil meadow mean = -275 plusmn 23 permil native species mean=-272 plusmn 14permil) whereas
macrophytes display slightly more negative values with a mean of -287 plusmn 23permil
Macrophytes substrate displayed the highest δsup1⁵N values with a mean of +60 plusmn
14permil (n=3) Similar and relatively high δsup1⁵N values for soil of plantation (mean= +54
plusmn 13permil n=4) meadow (mean= +49 plusmn 04permil n=8) and surface river sediment
90
(mean= +52 plusmn 17permil n=10) Native forest soils oscillate around slightly more
negative values with a mean of +45 plusmn 12permil (n=4) Slightly more positive soil δsup1sup3C
values occur both underneath native forests and in tree plantations with means of -
284 plusmn 23permil and -298 plusmn 13permil respectively compared to either meadow soils
(mean= -302 plusmn 11permil) aquatic sediments underneath macrophytes (-311 plusmn 10permil)
or from surface river sediments (mean= -312 plusmn 10permil)
Figure 6 Stable isotope content of modern soil river sediment and foliar vegetation
used as end members in the sedimentary sequence of Lago Vichuqueacuten a)
Scatterplot of foliar δsup1⁵N and δsup13C values for vegetation from Lago Vichuqueacuten
watershed (plantation meadow and native species) and macrophytes on Lake
Vichuqueacuten See supplementary material for more detail of vegetation types b)
Scatterplot of soil δsup1⁵N and δsup13C values across the watershed the sediments of the
Huintildee and Vichuqueacuten rivers and the fluvial sediment corresponding to the
macrophyte vegetation
45 Land use and cover change from 1975 to 2014
Major land use changes between 1975 CE and 2016 CE in the Lago
Vichuqueacuten watershed are summarized in Figure 7 The watershed has a surface
area of 535329 km2 of which native vegetation (26) and shrublands (53)
represent 79 of the total surface in 1975 Meadows are confined to the valley and
91
represent 17 of watershed surface Tree plantations initially occupied 1 of the
watershed and were first located along the lake periphery By 1989 the areas of
native forests shrublands and meadows had decreased to 22 31 and 14
respectively whereas tree plantations had expanded to 30 These trends
continued almost invariably until 2016 when shrublands and meadows reached 17
and 5 of the total areas while tree plantations increased to 66 Native forests
had practically disappeared by 1989 and then increased up to 7 of the total area
in 2016 (Fig 6) preferentially occupying the southeastern sector of the watershed
Figure 7 Changes in land use and cover between 1975 and 2016 in the Lago
Vichuqueacuten watershed as measured from satellite images The major change is
represented by the replacement of native forest shrubland and meadows by
plantations of Monterrey pine (Pinus radiata)
Figure 8 shows correlations between lake sediment stable isotope values and
changes in the soil cover from 1975 to 2013 Positive relationships occurred
between sediment δsup1⁵N and δsup13C values from core VIC13-2B (unit 1) and the
92
percentage of native forest (r=089 for δsup1⁵N 082 for δsup13C) shrublands (r = 071 for
δsup1⁵N 057 for δsup13C) and meadow cover (r = 088 for δsup1⁵N 081 for δsup13C) All of these
correlations are significant (p value lt 0001) In contrast significant negative
correlations (p lt0001) occurred between tree plantation cover and lake sediment
stable isotope values (δsup1⁵N r = -082 and δsup13C r = -067) native forest (r = -097)
meadows (r = -086) and shrubland (r =-093)
Figure 8 Correlation plots of land use and cover change versus lake sediment
stable isotope values The δsup1⁵N values are positively correlated with native forests
agricultural fields and meadow cover across the watershed Total Plantation area
increases are negatively correlated with native forest meadow and shrubland total
area Significance levels are indicated by the symbols p-values (0 0001 001
005 01 1) lt=gt symbols ( )
93
5 Discussion
51 Seasonal variability of POM in the water column
The stable isotope values of POM can vary during the annual cycle due to
climate and biologic controls namely temperature and length of the photoperiod
which affect phytoplankton growth rates and isotope fractionation in the water
column (Gu et al 2006 Leng et al 2006) POM from Lago Vichuqueacuten surface
samples (2 and 5 m depth) displayed lower δ13CPOM values during the summer than
in winter During C uptake phytoplankton preferentially utilize 12C leaving the
DICpool enriched in 13C Therefore as temperature increases during the summer
phytoplankton growth generates OM enriched in 12C until this becomes depleted
and then the biomas come to enriched u At the onset of winter the DICpool is now
enriched in 13C and despite an overall decrease in phytoplankton production the
OM produced will also be enriched in 13C In contrast δ13CPOM values at 20 m depth
did not reflect these seasonal differences probably due to water-column
stratification that maintains similar temperatures and biological activity throughout
the year
Lake N availability depends on N sources including inputs from the
watershed and the atmosphere (ie deposition of N compounds and fixation of
atmospheric N2) which varies during the hydrologic year The fixation of atmospheric
N2 is an important natural source of N to the lake occurring mainly during the
summer season associated with higher temperature and light (Gu et al 2006)
Because the δ15NPOM of N2 fixers is close to 0permil with almost no overall isotope
fractionation (Leng et al 2006) when N2 is the major N source δ15NPOM values are
typically low However when DIN concentrations are high or alternatively when little
94
N2 fixation occurs δ15NPOM values will be higher (Gu et al 2006) δ15NPOM values
from summer Lago Vichuqueacuten samples were lower than those from winter with large
differences between the seasons (~5permil Fig 5 and 9) Furthermore δ15NPOM values
were high when monthly average temperature was low and monthly precipitation
was high (Fig 9) This pattern is consistent with the dominance of high N2 fixation
by cyanobacteria associated with increased summer temperatures This correlation
of δ15NPOM values with temperature further suggests a functional group shift i e
from N fixers to phytoplankton that uptake DIN The correlation between wetter
months and higher δ15NPOM values could be caused by increased N input from the
watershed due to increased runoff during the winter season The lack of data of the
δ15NPOM between the 30 mm and the 160 mm of precipitation is due to the
mediterranean-type climate that concentrates precipitations in the winter months
Finally Lago Vichuqueacuten is characterized by pronounced seasonality leading to
higher phytoplankton biomass in summer characterized by low δ15NPOM In winter
low biomass production and increased input from watershed is associated to high
δ15NPOM
95
Figure 9 Seasonal changes can drive δ15NPOM values in Lago Vichuqueacuten Data
correspond to average monthly temperature and total monthly precipitation for the
months when the water samples were taken (years 2015 - 2018) P-valuelt005
52 Stable isotope signatures in the Lake Vichuqueacuten watershed
The natural abundance of 15N14N isotopes of soil and vegetation samples
from the Lago Vichuqueacuten watershed appear to result from a combination of factors
isotope fractionation different N sources for plants and soil microorganisms (eg N2
fixation Nitrates) depth in the soil where N uptake by vegetation occurs and N loss
mechanisms (ie denitrification leaching and ammonia volatilization Hogberg
1997) The lowest δsup1⁵Nfoliar values are associated with native species and are
probably due to the contribution of N-fixing species (eg Sophora) (Fig6 and for
more detail see Table S3 in supplementary material) The number native N-fixers
species present in the Chilean mediterranean vegetation are not well known
however so there could be other N-fixing species in this group Alternatively δsup1⁵Nfoliar
values reflect soil N uptake (Kahmen et al 2008) In environments limited by N
plants have δsup1⁵N values similar to the soil δsup1⁵N values however the denitrification
and volatilization of ammonia can lead to the remain N of soil to come enriched in
15N (Cifuentes et al 1989 Craine et al 2009 Bruland and Mckenzie 2010) Soil N
isotope samples from native species communities tends to display relatively high
δsup1⁵N values respect to foliar samples due to loss of N-soil
The higher foliar and soil δsup1⁵N values obtained from samples of meadows
aquatic macrophytes and tree plantations can be attributed to the presence of
greater amounts of nitrate available for plant uptake (Fig 6) Houmlgberg (1997)
suggests that the availability of different N sources in soils (ie nitrates versus
96
ammonia) with different residence times can also explain these δsup1⁵NFoliar values
Indeed Feigin et al (1974) described differences of up to 20permil between ammonia
and nitrates sources Denitrification and nitrification discriminate much more against
15N than N2 fixation does (Craine et al 2009) leading to soils (and plants after
uptake) enriched in 14N
In general multiple processes that affect the isotopic signal result in similar
δsup1⁵N values between the soil of the watershed and the sediments of the river
However POM isotope fluctuations allow to say that more negative δsup1⁵N values are
associated to lake productivity while more positive δsup1⁵N values are associated with
N input from the watershed
δ13C values in Lago Vichuqueacuten watershed reflect CO2 gas exchange between
C3 plants and algae with the atmosphere During photosynthesis plants
discriminate against the heavier stable isotope of carbon (13C) in favor of the lighter
isotope 12C which results in δ13CFoliar values between -220permil and -320permil (Browman
and Cook 2002 Liu et al 2007) Likewise the δ13CFoliar values from Lago Vichuqueacuten
oscillate around -27permil (Fig 6) Plants transform atmospheric CO2 into soil organic
carbon (C) which in turn reflects this initial discrimination against 13C during C
uptake and post photosynthetic fractionation (Farquhar et al 1982 1989 Badeck
et al 2005 Pausch and Kuzyakov 2017) Slightly more positive δ13CFoliar values
(about 15permil) were measured in comparison with their δ13CSoil values This may be
reflecting the C transference from plants to the soil but also a soil-atmosphere
interchange The preferential assimilation of the light isotopes (12C) during soil
respiration carried by the roots and the microbial biomass that is associated with the
decomposition of litter roots and soil organic matter explain this differential
(Berhardt et al 2006 Blagodatskaya et al 2010 Midwood and Millard 2011)
97
In general the δ13CSoil values from the Lago Vichuqueacuten watershed oscillated
around -290permil and did not vary with our plant classification types Here we use
these values as terrestrial-end members to track changes in source OM (Fig 6)
Thus more negative δ13C values in lake sediment samples are attributed to higher
OM inputs from the terrestrial watershed By the other hand more positive δ13C
values most likely reflect an increased aquatic OM component as indicated by POM
isotope fluctuations (Fig 9)
53 Recently land use and cover change and its influences on N inputs to the lake
Tree plantations in the Lago Vichuqueacuten watershed have increased greatly in
the last few decades (from 1 in 1975 to 66 in 2016 Fig 7) replacing previous
native forests (from 26 to 7) meadows (17 to 5) and shrublands (53 to
17) In 1975 tree plantations were confined to the lake perimeter with discrete
patches distributed throughout the watershed The ldquoForest Lawrdquo (DFL 701- passed
in 1974) allocated state funding to afforestation efforts and management of tree
plantations which greatly favored the replacement native forests by introduced trees
This increase is marked by a sharp and steady decrease in lake sediment δ15N and
δ13C values because tree plantations function as a nutrient sink whereas other land
uses such as meadows favor nutrient loss from the watershed (Fig8) Bruland and
Mackenzie (2014) noted a decrease in wetland δ15N values when watershed
forested cover increased and concluded that N inputs to the wetlands are lower from
the forested areas as they generally do not export as much N as agricultural lands
A positive correlation between native vegetation and δ15Ncore values can be
explained by the relatively scarce arboreal cover in the watershed in 1975 when
native forest occupied just 26 of the watershed surface whereas shrublands and
98
meadows occupied more than the 70 of the surface of the watershed with the
concomitant elevated loss of N (Fig 7 and Fig 8)
54 Nitrogen dynamics in Lago Vichuqueacuten over the last six hundred years
Sedimentological compositional and geochemical indicators all show changes in
the N cycle associated to LUCC lake dynamics and climate changes (Fig 10) From
the pre-Columbian indigenous settlement including the Spanish colonial period up
to the start of the Republic (1300 - 1800 CE) the introduction of crops such as
quinoa and wheat but also the clearing of land for extensive agriculture would have
favored the entry of N into the lake Conversely major changes observed during the
last century were characterized by a sharp decrease of N input that were coeval
with the increase of tree plantations
From 1365 until 1550 CE (unit 2c 2b and half of 2a Fig 3 and Fig 10) pre-
Columbian agricultural societies planted mostly crops of corn and quinoa (Ramirez
and Vidal 1985) This period is characterized by large peaks seen in the δsup1⁵N record
(eg 1403 1452 1476 and 1505) with overall high δsup1⁵N values (up to 5permil) indicating
that N input from watershed was elevated and oscillating to the beat of the NT These
positive δsup1⁵N peaks could be due to several causes including a) the clearing of land
for farming b) N loss via denitrification which would be generally augmented in
anoxic environments (Diebel et al 2009) such as Lago Vichuqueacuten (low MnFe
values about 0001 Fig 4) or c) climate especially cold-wet winters or hot-dry
summers can also exert control on the δsup1⁵N record Indeed tree-ring records and
summer temperature reconstructions show overall wetcold conditions during this
period (Christie et al 2009 Von Gunten et al 2009 Fig 10) Increased
precipitation would bring more sediment (and nutrients) from the watershed into the
99
lake and increase lake productivity which is also detected by the geochemical
proxies for productivity (peaks of BrTi SrCa and TOC Fig 4 and Fig 10) (see also
Frugone-Alvarez et al 2017)
Figure 10 Changes in the N availability during the last six centuries in Lago
Vichuqueacuten a) lake sediment δsup1⁵N values displayed more positive values during the
prehistoric period Spanish Colony and the starting 19th century which is associated
with enhanced N input from the watershed by extensive clearing and crop
plantations The inset shows this relationship between sediment δsup1⁵N and
100
percentage of meadow cover over the last 30 years b) Summer temperature
reconstruction from central Chile (von Gunten et al 2009) showing a
correspondence with cold phases and high δsup1⁵N values at Vichuqueacuten except for the
last 100 years c) Palmer Drought Severity Index (PDSI) is an interannual moisture
variability reconstruction for late springndashearly summer during the last six centuries
(Christie et al 2009) Grey shadow indicating higher precipitation periods
From 1550 and 1810 CE (unit 2a) and during the Colonial period several peaks
of δ15N could be further related not only to high precipitation (Fig 10 grey shadow)
but also pulses of enhanced N input from the watershed linked to human land use
In 1550 CE Juan Cuevas was granted lands and indigenous workers under the
encomienda system for agricultural and mining development of the Vichuqueacuten
village (Ramiacuterez and Vidal 1985) Historical documents show that around 1579 CE
the Vichuqueacuten watershed was occupied by indigenous communities dedicated to
wheat plantations and vineyards wood extraction and gold mining (Odone 1998)
The introduction of the Spanish agricultural system implied not just a change in the
types of crops used (from quinoa to vineyards and wheat) but also a clearing of
native species for the continuous increase of agricultural surface and wood
extraction During the Colonial period wheat from Vichuqueacuten was exported to Peru
(Ramiacuterez and Vidal 1985) Armesto et al (2009) indicate that during the XVIII and
XIX centuries the extraction of wood for mining operations was important enough to
cause extensive loss of native forests The independence and instauration of the
Chilean Republic did not change this prevailing system Increases in the
contributions of N to the lake during the second half of the XIX century (peaks in
δsup1⁵N ca 1870 CE) could be due to expanding farm land dedicated for wheat
101
production and increased commercial trade with California and Canada (Ramiacuterez
and Vidal 1985)
In contrast LUCC in the last century are clearly related to the development of
large-scale tree plantations (Fig 7) The δsup1⁵N record reaches the lowest values of
the entire sequence in the last few decades (Fig 10) A marked increase in lake
productivity NT concentration and decreasing sediment input is synchronous (unit
1 Fig 4) with trees replacing meadows shrublands and areas with native forests
(Fig 7 and 8) The implementation of the lsquoForest Lawrsquo in 1974 had a stronger impact
on the landscape and lake ecosystem dynamics than the impacts of ongoing climate
change in the region which is much more recent (Garreaud et al 2018) although
the prevalence of hot dry summers seen over the last decade would also be
associated with low δsup1⁵N values and increase of NT (Fig 10) Heavy metals ratios
(FeTi CoTi and PbTi) show notable increases in lake sediments from 1992 to 2011
CE (Fig 4) Although this could be related to mining in the El Maule region the
closest mines are 60 Km away (Pencahue and Romeral) so local factors related to
shoreline urbanization for the summer homes and an increase in tourist activity
could also be a major factor
6 Conclusions
The N isotope signal in the watershed depends on the rates of exchange
between vegetation and soil cover (Fig 11) Plants preferentially uptake 14N and the
underlying soils become enriched in 15N especially when the terrestrial ecosystem
is N-limited andor significant N loss occurs (ie denitrification andor ammonia
volatilization) LUCC in the Lago Vichuqueacuten watershed have heavily modified the
links between terrestrial and aquatic ecosystems with agriculture practices
102
contributing more N to the lake than tree plantations or native forests In situ lake
processes can also fractionate N isotopes An increase of N-fixing species results
in OM depleted in 15N which results in POM with lower δsup1⁵N values during these
periods During winter phytoplankton is typically enriched in 15N due to the
decreased abundance of N-fixing species and increased N input from the
watershed This in turn generates the highest δsup1⁵NPOM values in Lago Vichuqueacuten
Periods of elevated productivity tend to deplete the dissolved inorganic N of 14N
resulting in even higher δ15N values
Our study illustrates the usefulness of δsup1⁵N as a tool for reconstructing the past
influence of LUCC on N availability in lake ecosystems To constrain the relative
roles of the diverse forcing mechanisms that can alter N cycling in mediterranean
ecosystems all main components of the N cycle should be monitored seasonally
(or monthly) including the measurements of δ15N values in land samples
(vegetation-soil) as well as POM
103
Figure 11 Summary of human and environmental factors controlling the δ15N
values of lake sediments Particulate organic matter(POM) δ15N values in
mediterranean lakes are driven by N input from the watershed that in turn depend
on land use and cover changes (ie forest plantation agriculture) andor seasonal
changes in N sources andor lake ecosystem processes (ie bioproductivity redox
condition denitrification) Arrows indicate the direction of N fluxes (inputoutput from
the N cycle) N cycle processes that deplete lake sediments of 15N are shown in
blue whereas those that enrich sediments in 15N are shown in red
104
Supplementary material
Figure S1 Pearson correlate coefficient between geochemical variables in core
VIC13-2B Positive and large correlations are in blue whereas negative and small
correlations are in red (p valuelt0001)
Figure S2 Principal Component Analysis of geochemical elements from core
VIC13-2B
105
Table S1 Lago Vichuqueacuten radiocarbon samples
RADIOCARBON
LAB CODE
SAMPLE
CODE
DEPTH
(m)
MATERIAL
DATED
14C AGE ERROR
D-AMS 029287
VIC13-2B-
1 043 Bulk 1520 24
D-AMS 029285
VIC13-2B-
2 085 Bulk 1700 22
D-AMS 029286
VIC13-2B-
2 124 Bulk 1100 29
Poz-63883 Chill-2D-1 191 Bulk 945 30
D-AMS 001133
VIC11-2A-
2 201 Bulk 1150 44
Poz-63884
Chill-2D-
1U 299 Bulk 1935 30
Poz-64089
VIC13-2D-
2U 463 Bulk 1845 30
Poz-64090
VIC13-2A-
3U 469 Bulk 1830 35
D-AMS 010068
VIC13-2D-
4U 667 Bulk 2831 25
Poz-63886
VIC13-2D-
4U 719 Bulk 3375 35
106
D-AMS 010069
VIC13-2D-
5U 775 Bulk 3143 27
Poz-64088
VIC13-2D-
5U 807 Bulk 3835 35
D-AMS-010066
VIC13-2D-
7U 1075 Bulk 6174 31
Poz-63885
VIC13-2D-
7U 1197 Bulk 6440 40
Poz-5782 VIC13-15 DIC 180 25
Table S2 Loadings of the trace chemical elements used in the PCA
Elementos PC1 PC2 PC3 PC4
Zr 0922 0025 -0108 -0007
Zn 0913 -0124 -0212 0001
Rb 0898 -0057 -0228 0016
K 0843 0459 0108 0113
Ti 0827 0497 0060 -0029
Al 0806 0467 0080 0107
Si 0803 0474 0133 0136
Y 0784 -0293 -0174 0262
V 0766 0455 0090 -0057
Br 0422 -0716 -0045 0226
Ca 0316 -0429 0577 0489
Sr 0164 -0420 0342 -0182
Cl 0151 -0781 -0397 0162
107
Mn -0121 -0091 0859 0095
S -0174 -0179 -0051 0714
Pb -0349 0414 -0282 0500
Fe -0700 0584 -0023 0280
Co -0704 0564 -0107 0250
Table S3 Stable Isotope values from vegetation of Lago Vichuqueacutenacutes watershed
Taxa Classification δsup1⁵N δsup1sup3C CN
molar
Poaceae Meadow 1216 -2589 3602
Juncacea Meadow 1404 -2450 3855
Cyperaceae Meadow 1031 -2596 1711
Taraxacum
officinale Meadow 836 -2400 2035
Poaceae Meadow 660 -2779 1583
Poaceae Meadow 453 -2813 1401
Poaceae Meadow 966 -2908 4010
Juncus Meadow 1247 -2418 3892
Poaceae Meadow 747 -3177 6992
Poaceae Meadow 942 -2764 3147
Poaceae Meadow 1479 -2634 2895
Poaceae Meadow 1113 -2776 1795
Poaceae Meadow 2215 -2737 7971
Poaceae Meadow 1121 -2944 2934
Poaceae Meadow 638 -3206 1529
108
Macrophytes Macrophytes 886 -3044 2286
Macrophytes Macrophytes 1056 -2720 2673
Macrophytes Macrophytes 769 -3297 1249
Macrophytes Macrophytes 967 -2763 1442
Macrophytes Macrophytes 959 -2670 2105
Macrophytes Macrophytes 334 -2728 1038
Acacia dealbata
Introduced
species 656 -2696 1296
Acacia dealbata
Introduced
species 487 -2941 1782
Acacia dealbata
Introduced
species 220 -2611 3888
Luma apiculata Native species 433 -2542 4135
Luma apiculata Native species 171 -2664 7634
Luma apiculata Native species -001 -2736 6283
Luma apiculata Native species 029 -2764 6425
Azara sp Native species 159 -2868 8408
Azara sp Native species 101 -2606 2885
Baccharis concava Native species 104 -2699 5779
Baccharis concava Native species 265 -2488 4325
Baccharis concava Native species 287 -2562 7802
Baccharis concava Native species 427 -2781 5204
Baccharis linearis Native species 190 -2610 4414
Baccharis linearis Native species 023 -2825 5647
109
Peumus boldus Native species 042 -2969 6327
Peumus boldus Native species 205 -2746 4110
Peumus boldus Native species 183 -2743 6293
Chusquea quila Native species 482 -2801 4275
Poaceae meadow 217 -2629 7214
Lobelia sp Native species 224 -2645 3963
Lobelia sp Native species -091 -2565 4538
Aristotelia chilensis Native species -035 -2785 5247
Aristotelia chilensis Native species -305 -2889 2305
Aristotelia chilensis Native species 093 -2836 5457
Chusquea quila Native species 173 -2754 3534
Chusquea quila Native species 045 -2950 6739
Quillaja saponaria Native species 223 -2838 9385
Scirpus meadow 018 -2820 7115
Sophora sp Native species -184 -2481 2094
Sophora sp Native species -181 -2717 1721
Pinus radiata
Introduced
trees 1581 -2602 3679
Pinus radiata
Introduced
trees 1431 -2784 4852
Pinus radiata
Introduced
trees -091 -2708 9760
Pinus radiata
Introduced
trees 153 -2568 3470
110
Salix sp
Introduced
trees 632 -2878 1921
LITERATURE CITED
Armesto JJ Manuschevich D Mora A Smith-Ramirez C Rozzi R Abarzuacutea
AM Marquet PA 2010 From the Holocene to the Anthropocene A historical
framework for land cover change in southwestern South America in the past 15000
years Land use policy 27 148ndash160
httpsdoiorg101016jlandusepol200907006
Battye W Aneja VP Schlesinger WH 2017 Earth rsquo s Future Is nitrogen the next
carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014
httpsdoiorg101002eft2235
Blaauw M Christeny JA 2011 Flexible paleoclimate age-depth models using an
autoregressive gamma process Bayesian Anal 6 457ndash474
httpsdoiorg10121411-BA618
Bowman DMJS Cook GD 2002 Can stable carbon isotopes (δ13C) in soil
carbon be used to describe the dynamics of Eucalyptus savanna-rainforest
boundaries in the Australian monsoon tropics Austral Ecol
httpsdoiorg101046j1442-9993200201158x
Brahney J Ballantyne AP Turner BL Spaulding SA Otu M Neff JC 2014
Separating the influences of diagenesis productivity and anthropogenic nitrogen
deposition on sedimentary δ15N variations Org Geochem 75 140ndash150
httpsdoiorg101016jorggeochem201407003
111
Bruland GL MacKenzie RA 2010 Nitrogen Source Tracking with δN Content
of Coastal Wetland Plants in Hawaii J Environ Qual 39 409
httpsdoiorg102134jeq20090005
Carrevedo ML Frugone M Latorre C Maldonado A Bernaacuterdez P Prego R
Caacuterdenas D Valero-Garceacutes B 2015 A 700-year record of climate and
environmental change from a high Andean lake Laguna del Maule central Chile
(36degS) Holocene 25 956ndash972 httpsdoiorg1011770959683615574584
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the
Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from
tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Craine JM Elmore AJ Aidar MPM Dawson TE Hobbie EA Kahmen A
Mack MC Kendra K Michelsen A Nardoto GB Pardo LH Pentildeuelas J
Reich B Schuur EAG Stock WD Templer PH Virginia RA Welker JM
Wright IJ 2009 Global patterns of foliar nitrogen isotopes and their relationships
with cliamte mycorrhizal fungi foliar nutrient concentrations and nitrogen
availability New Phytol 183 980ndash992 httpsdoiorg101111j1469-
8137200902917x
Cuven S Francus P Lamoureux SF 2010 Estimation of grain size variability
with micro X-ray fluorescence in laminated lacustrine sediments Cape Bounty
Canadian High Arctic J Paleolimnol 44 803ndash817 httpsdoiorg101007s10933-
010-9453-1
Diebel M Vander Zanden MJ 2012 Nitrogen stable isotopes in streams stable
isotopes Nitrogen effects of agricultural sources and transformations Ecol Appl 19
1127ndash1134 httpsdoiorg10189008-03271
112
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in freshwater
wetlands record long-term changes in watershed nitrogen source and land use SO
- Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash
2916
Fariacuteas L Castro-Gonzales M Cornejo M Charpentier J Faundez J
Boontanon N Yoshida N 2009 Denitrification and nitrous oxide cycling within the
upper oxycline of the oxygen minimum zone off the eastern tropical South Pacific
Limnol Oceanogr 54 132ndash144
Farquhar GD Ehleringer JR Hubick KT 1989 Carbon Isotope Discrimination
and Photosynthesis Annu Rev Plant Physiol Plant Mol Biol
httpsdoiorg101146annurevpp40060189002443
Farquhar GD OrsquoLeary MH Berry JA 1982 On the relationship between
carbon isotope discrimination and the intercellular carbon dioxide concentration in
leaves Aust J Plant Physiol httpsdoiorg101071PP9820121
Fogel ML Cifuentes LA 1993 Isotope fractionation during primary production
Org Geochem httpsdoiorg101007978-1-4615-2890-6_3
Frugone-Aacutelvarez M Latorre C Giralt S Polanco-Martiacutenez J Bernaacuterdez P
Oliva-Urcia B Maldonado A Carrevedo ML Moreno A Delgado Huertas A
Prego R Barreiro-Lostres F Valero-Garceacutes B 2017 A 7000-year high-
resolution lake sediment record from coastal central Chile (Lago Vichuqueacuten 34degS)
implications for past sea level and environmental variability J Quat Sci 32 830ndash
844 httpsdoiorg101002jqs2936
Gaumllman V Rydberg J Bigler C 2009 Decadal diagenetic effects on δ 13 C and
δ 15 N studied in varved lake sediment Limnol Oceanogr 54 917ndash924
httpsdoiorg104319lo20095430917
113
Gerhart LM McLauchlan KK 2014 Reconstructing terrestrial nutrient cycling
using stable nitrogen isotopes in wood Biogeochemistry 120 1ndash21
httpsdoiorg101007s10533-014-9988-8
Granger J Sigman DM Lehmann MF Tortell PD 2008 Nitrogen and oxygen
isotope fractionation during dissimilatory nitrate reduction by denitrifying bacteria 53
2533ndash2545 httpsdoiorg10230740058342
Gu B Chapman AD Schelske CL 2006 Factors controlling seasonal
variations in stable isotope composition of particulate organic matter in a softwater
eutrophic lake Limnol Oceanogr 51 2837ndash2848
httpsdoiorg104319lo20065162837
Harris D Horwaacuteth WR van Kessel C 2001 Acid fumigation of soils to remove
carbonates prior to total organic carbon or CARBON-13 isotopic analysis Soil Sci
Soc Am J 65 1853 httpsdoiorg102136sssaj20011853
Houmlgberg P 1997 Tansley review no 95 natural abundance in soil-plant systems
New Phytol httpsdoiorg101046j1469-8137199700808x
Kylander ME Ampel L Wohlfarth B Veres D 2011 High-resolution X-ray
fluorescence core scanning analysis of Les Echets (France) sedimentary sequence
New insights from chemical proxies J Quat Sci 26 109ndash117
httpsdoiorg101002jqs1438
Lara A Solari ME Prieto MDR Pentildea MP 2012 Reconstruccioacuten de la
cobertura de la vegetacioacuten y uso del suelo hacia 1550 y sus cambios a 2007 en la
ecorregioacuten de los bosques valdivianos lluviosos de Chile (35o - 43o 30acute S) Bosque
(Valdivia) 33 03ndash04 httpsdoiorg104067s0717-92002012000100002
Lehmann M Bernasconi S Barbieri A McKenzie J 2002 Preservation of
organic matter and alteration of its carbon and nitrogen isotope composition during
114
simulated and in situ early sedimentary diagenesis Geochim Cosmochim Acta 66
3573ndash3584 httpsdoiorg101016S0016-7037(02)00968-7
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos Transactions
American Geophysical Union httpsdoiorg1010292007EO070007
Liu G Li M An L 2007 The Environmental Significance Of Stable Carbon
Isotope Composition Of Modern Plant Leaves In The Northern Tibetan Plateau
China Arctic Antarct Alp Res httpsdoiorg1016571523-0430(07-505)[liu-
g]20co2
Marzecovaacute A Mikomaumlgi A Koff T Martma T 2011 Sedimentary geochemical
response to human impact on Lake Notildemmejaumlrv Estonia Est J Ecol 60 54
httpsdoiorg103176eco2011105
McLauchlan KK Williams JJ Engstrom DR 2013 Nutrient cycling in the
palaeorecord Fluxes from terrestrial to aquatic ecosystems Holocene 23 1635ndash
1643 httpsdoiorg1011770959683613496289
Meyers PA 2003 Application of organic geochemistry to paleolimnological
reconstruction a summary of examples from the Laurention Great Lakes Org
Geochem 34 261ndash289 httpsdoiorg101016S0146-6380(02)00168-7
Naeher S Gilli A North RP Hamann Y Schubert CJ 2013 Tracing bottom
water oxygenation with sedimentary MnFe ratios in Lake Zurich Switzerland
Chem Geol 352 125ndash133 httpsdoiorg101016jchemgeo201306006
Odone C 1998 El Pueblo de Indios de Vichuqueacuten siglos XVI y XVII Rev Hist
Indiacutegena 3 19ndash67
Pausch J Kuzyakov Y 2018 Carbon input by roots into the soil Quantification of
rhizodeposition from root to ecosystem scale Glob Chang Biol
httpsdoiorg101111gcb13850
115
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98
httpsdoiorg1011772053019614564785
Sun W Zhang E Jones RT Liu E Shen J 2016 Biogeochemical processes
and response to climate change recorded in the isotopes of lacustrine organic
matter southeastern Qinghai-Tibetan Plateau China Palaeogeogr Palaeoclimatol
Palaeoecol httpsdoiorg101016jpalaeo201604013
Torres IC Inglett PW Brenner M Kenney WF Reddy KR 2012 Stable
isotope (δ13C and δ15N) values of sediment organic matter in subtropical lakes of
different trophic status J Paleolimnol 47 693ndash706
httpsdoiorg101007s10933-012-9593-6
Vitousek PM Aber JD Howarth RW Likens GE Matson PA Schindler
DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl
httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
von Gunten L Grosjean M 2009 High-resolution quantitative climate
reconstruction over the past 1000 years and pollution history derived from lake
sediments in Central Chile Philos Fak PhD 246
Woodward C Chang J Zawadzki A Shulmeister J Haworth R Collecutt S
Jacobsen G 2011 Evidence against early nineteenth century major European
induced environmental impacts by illegal settlers in the New England Tablelands
south eastern Australia Quat Sci Rev 30 3743ndash3747
httpsdoiorg101016jquascirev201110014
Xu H Yu K Lan J Sheng E Liu B Ye Y Hong B Wu H Zhou K Yeager
KM 2016 Different responses of sedimentary δ15N to climatic changes and
116
anthropogenic impacts in lakes across the Eastern margin of the Tibetan Plateau
J Asian Earth Sci 123 111ndash118 httpsdoiorg101016jjseaes201603024
117
DISCUSION GENERAL
El Nitroacutegeno (N) es un elemento clave que controla la dinaacutemica y
funcionamiento de ecosistemas tanto terrestres como acuaacuteticos (Vitousek et al
1997) Pese a que el N (N2) es muy abundante en la atmoacutesfera (78) es escaso
en los ecosistemas pues la mayoriacutea de la biota no puede acceder a su forma
molecular (Galloway et al 1995 Zaehle et al 2013) En este sentido la entrada
natural de N en los ecosistemas es viacutea fijacioacuten bioloacutegica del N2 por bacterias que lo
convierten en compuestos bioloacutegicamente disponibles (Battye et al 2017) Debido
a que la fijacioacuten es un proceso energeacuteticamente costoso los ecosistemas
comuacutenmente estaacuten limitados por N (Vitousek and Howarth 1991) Los LUCC
contribuyen al incremento del N disponible y son una de las principales causas de
eutroficacioacuten y contaminacioacuten de los cuerpos de agua (Woodward et al 2012)
En Chile central los LUCC principalmente relacionados con las actividades
agriacutecolas y forestales han causado grandes cambios en el ciclo del N debido al
118
reemplazo de la cobertura natural por un monocultivo y al uso de fertilizantes que
modifican los aportes de MO y N a los cuerpos de agua El programa de
estabilizacioacuten de laderas impulsada por el gobierno (Albert 1900) y la ley forestal
de 1974 han fomentado la forestacioacuten con plantaciones de Pinus radiata y
Eucalyptus globulus y en el caso de Lago Vichuqueacuten estas actualmente cubren maacutes
del 60 de su cuenca (Cap2 Fig 5) Ademaacutes en Laguna Matanzas la
sobreexplotacioacuten del agua en conjunto con el cambio climaacutetico actual el que ha
conducido a una de las megasequiacuteas maacutes severas en las uacuteltimas deacutecadas
(Garreaud et al 2017) generoacute una combinacioacuten letal que secoacute el lago en tan solo
10 antildeos (Cap1 Fig1) Las evidencias obtenidas a partir de estos dos lagos
permiten identificar las huellas del Antropoceno en Chile central basadas en el
registro sedimentario lacustre
La transferencia de N desde los ecosistemas terrestres a acuaacuteticos es un
proceso natural con controles bioacuteticos climaacuteticos y geoloacutegicos El enlace
hidroloacutegico entre los ecosistemas terrestres y acuaacuteticos hace que el estado troacutefico
de los lagos esteacute vinculado al de su cuenca (McLauchlan et al 2013) En Chile
central se sabe poco del impacto de los LUCC sobre el ciclo de nutrientes en los
ecosistemas lacustres aunque al menos desde la colonizacioacuten espantildeola se tienen
registros de influencia humana en las cuencas Durante la colonia espantildeola
Laguna Matanzas fue una hacienda ganadera que exportaba productos caacuternicos al
Peruacute (Contreras-Lopez et al 2014) A su vez en el Lago Vichuqueacuten se realizaban
extensas actividades agriacutecolas hasta comienzos del s XX como por ejemplo
cultivos de vid y trigo ademaacutes de extraccioacuten de madera y mineriacutea de oro (Odone
1997) En las uacuteltimas deacutecadas los LUCC han estado vinculados principalmente con
el desarrollo forestal al compaacutes de una estrategia gubernamental de fomentar con
119
incentivos financieros (Ley de Bosques DFL nordm265 de 1931 y DFL 701 de 1974)
esta actividad El incremento de la superficie forestal es especialmente fuerte en
ambos lagos a partir de la deacutecada de 1980 y hasta 2016 (Laguna de Matanzas 0-
17 Lago Vichuqueacuten 1-66) Este aumento habriacutea sido en detrimento del bosque
nativo y los pastizales (Cap 1 Fig 5 y Cap 2 Fig 8) El incremento de la superficie
forestal generariacutea una disminucioacuten de los aportes de sedimentos y nutrientes al lago
y en este sentido un cambio de estado en los flujos de N (e g tipping points) que
a su vez ha quedado registrado en la sentildeal isotoacutepica (δ15N) y la acumulacioacuten de
MO en los sedimentos lacustres
Isoacutetopos estables y la dinaacutemica de N en los lagos de Chile central
Los sedimentos lacustres son verdaderos ldquosumiderosrdquo que pueden llegar a
registrar lo que ocurre en cuanto a la historia ambiental de su cuenca En esta tesis
se utilizan los isoacutetopos estables de N (15N y 14N) en sedimentos lacustres para
reconstruir los cambios en la disponibilidad de N en el tiempo y establecer la
magnitud de impacto generado por actividades humanas El fraccionamiento
cineacutetico en los lagos es mediado por procesos bioloacutegicos que mediante la
asimilacioacuten de N genera un producto (eg fitoplancton) maacutes ldquoligerordquo (valores maacutes
bajos de 15N) que su remanente (eg DIN) (Fogel y Cifuentes 1993) Sin embargo
en periacuteodos en que los lagos estaacuten limitados por N el fitoplancton discrimina menos
y la MO resultante queda enriquecida en 15N (Fig 1 a y b) Ademaacutes la
desnitrificacioacuten acentuada en lagos de fondo anoacutexicos (eg Laguna Matanzas
entre 1800 y 1940 Cap1 Fig 6) conduce a un enriquecimiento en 15N de los
sedimentos y de la columna de agua Esta informacioacuten queda reflejada en la MO
120
de los sedimentos lacustres lo que permite conocer la disponibilidad del N en el
tiempo a partir de las variaciones de 15N
En esta tesis analizamos la MO de los sedimentos lacustres para reconstruir
la influencia de los LUCC en los aportes de sedimentos y N al lago En teacuterminos de
asimilacioacuten de N se puede distinguir entre dos grupos principales de productores
primarios que componen el POM (Fig1)
1 Los que asimilan el DIN compuesto por amonio nitratos y nitritos Aquiacute el
δ15N resultante fluctuaraacute en torno a valores maacutes positivos en la medida que
la productividad se incrementa o no hay reposicioacuten del DIN (Fig 1)
2 Los fijadores de N Dado que es un proceso energeacuteticamente costoso en
ambientes que no estaacuten limitados por N muchas veces son excluiacutedas
competitivamente por el resto del fitoplancton Si el DIN queda agotado por
el incremento de la productividad (o bien si el ambiente estaacute limitado ya sea
por N o por otros nutrientes) los fijadores de N pueden florecer y la MO que
se genera a partir de ello tendraacute un δ15N con valores en torno a 0permil
De este modo la MO en los sedimentos lacustres dependeraacute de la
composicioacuten de productores primarios (ie fijadores vs asimiladores del DIN)
ademaacutes de los aportes desde la cuenca tanto de MO como del N de la cuenca que
pasa a formar parte del DIN (eg fertilizantes estieacutercol) (Fig 1)
La MO de los lagos estudiados en esta tesis ha sido analizada a partir de
variables geoquiacutemicas isotoacutepicas frotis y estaacute compuesta principalmente por
diatomeas y MO amorfa A partir de los datos obtenidos en esta tesis los valores
de δ15N aumentan cuando hay una mayor cobertura agriacutecola o pastizales A su vez
tiende a disminuir cuando aumenta la cobertura arboacuterea independiente si esta es
por plantaciones forestales o por bosque nativo
121
Por otra parte la dinaacutemica del lago tambieacuten imprime caracteriacutesticas
especiales en el POM observaacutendose variaciones estacionales en los valores
δ15NPOM Durante la estacioacuten caacutelida y seca se observan valores maacutes bajos que
durante la estacioacuten friacutea y huacutemeda posiblemente relacionado con el incremento de
la contribucioacuten de especies fijadoras de N (Lago Vichuqueacuten Fig 9 Cap 2) durante
el verano En la estacioacuten friacutea y huacutemeda el DIN seriacutea la principal fuente de N Las
mayores entradas de MO y N terrestre debidos a un incremento del lavado de la
cuenca como consecuencia de las precipitaciones y la mineralizacioacuten de la MO
podriacutean estimular la produccioacuten de MO lacustre por crecimiento del fitoplancton
Como consecuencia se observan tendencias decrecientes de los valores de
δ15N en la MO sedimentaria cuando la productividad en el lago estaacute relacionada
con especies fijadoras de N mientras que el δ15N muestra aumentos cuando la
productividad del lago estaacute asociada principalmente al consumo del DIN pero
tambieacuten cuando predominan procesos de perdida de N (eg desnitrificacioacuten Fig
1)
Hemos identificado tres fases en la dinaacutemica del δ15N para ambos lagos
Entre 1800 y 1940 CE la cuenca de laguna de Matanzas estaba dominada por
actividades agriacutecolas y ganaderas (δ15Nsuelo gt 4permil Cap 1 Fig 4 y 6) Las entradas
de sedimentos y MO desde la cuenca son altas (δ13C lt -209permil ZrTi ~029 AlTi
~013) y coincide con un clima maacutes huacutemedo y friacuteo (Von Gunten et al 2009
Christiansen et al 2011) El lago teniacutea un nivel de agua maacutes profundo dominado
por condiciones anoacutexicas (MnFe ~0013) que contribuiriacutean a mayores valores de
δ15N en la columna de agua y por ende de los sedimentos acumulados en el fondo
debido a la peacuterdida diferencial de 14N viacutea desnitrificacioacuten (Fig 7 Cap1) La baja
122
produccioacuten de MO lacustre (BrTi ~002 TOC~17) coincide con un bajo contenido
de NT (~478 μg) y valores maacutes positivos de δ15N (gt 4permil)
La actividad agriacutecola tambieacuten dominaba en la cuenca del Lago Vichuqueacuten
durante este periacuteodo (δ15Nsuelo gt4permil Cap2 Fig 6) El aporte sedimentario desde la
cuenca es alto (AlTi ~029 ZrTi ~032) y fue favorecido por mayores
precipitaciones a las actuales (Christiansen et al 2011) El Lago Vichuqueacuten es un
lago estratificado anoacutexico (MnFe ~002) y profundo (~30 m) por lo que la
desnitrificacioacuten juega un rol importante en el enriquecimiento de la MO
sedimentaria La baja produccioacuten de MO (BrTi ~007 TOC ~12) de origen
lacustre (δ13C ~ -268permil) coincide con un bajo contenido de NT (~309μg +6) y
valores maacutes positivos de δ15N (56permil +03)
Durante esta fase en ambos lagos los aportes de N de la cuenca parecen
ser los principales responsables en las fluctuaciones del δ15N Esta sentildeal podriacutea
estar amplificada por bajas temperaturas (que afectan la productividad lacustre) y
altas precipitaciones que favorecen el lavado de la cuenca y aumentan el aporte de
sedimentos y MO desde la cuenca predominantemente agriacutecola
Entre 1940 y 1980 CE en Laguna Matanzas se observa un incremento en
la acumulacioacuten de MO algal (TOC ~2 +1 δ13C ~-230+09permil) El ambiente
deposicional es ligeramente menos turbulento que el anterior (AlTi ~011+001
ZrTi ~03+001) y el lago estaacute en una fase de transicioacuten hacia condiciones maacutes
oacutexicas (MnFe ~002+0004) y maacutes productivas (BrTi ~004+0007) A su vez δ15N
tiende hacia valores maacutes negativos (δ15N ~30permil+03) mientras que el NT aumenta
oscilando en antifase con el δ15N
En Lago Vichuqueacuten en cambio se observa un ligero incremento en la
acumulacioacuten de MO (TOC ~13 +02) de origen terrestre (δ13C ~-27permil +04) La
123
productividad es similar al periacuteodo anterior (BrTi ~007+003) y el ambiente
deposicional es menos turbulento (AlTi ~02 +009 Zrti ~033 +007) El δ15N y el
NT oscilan en torno a valores ligeramente maacutes bajos (δ15N ~51permil +04 NT ~292μg
+6) Este periacuteodo se caracteriza por ser maacutes caacutelido que el anterior lo que
posibilitariacutea un incremento en la productividad lacustre en Laguna Matanzas pero
que no es observada en el Lago Vichuqueacuten
Entre 1980 y la actualidad en Laguna Matanzas habriacutea un incremento de la
acumulacioacuten de MO (TOC ~59 + 3) asociada a un incremento en la productividad
del lago (BrTi ~009+005) y a los aportes de MO terrestres (δ13C ~-24 + 2permil) El
lago se vuelve maacutes oacutexico (Mn ~0003 +0006) y las entradas de sedimento
disminuyen (AlTi~ 009 +002) El δ15N tiende a valores maacutes negativos (δ15N ~13permil
+1) y el NT aumenta de 20 a 55 μg (NT ~346 + 9 μg) tomando valores sin
precedentes en todo el registro En los sedimentos lacustres del Lago Vichuqueacuten
tambieacuten se observa un incremento de la acumulacioacuten de MO (TOC ~17 + 05)
asociada a un incremento en la productividad del lago (BrTi ~01 + 003) y las
entradas de sedimento desde la cuenca disminuyen (AlTi ~005 + 005) El δ15N
(~43 + 05permil) tiende a valores maacutes negativos y el NT incrementa de 20 a 55 μg (NT
~346 + 9 μg) oscilando en antifase
Durante esta fase en ambos lagos se observa un aumento en la
acumulacioacuten de MO sedimentaria un incremento en el NT y valores maacutes negativos
de δ15N que coincide con el incremento de la superficie forestal de las cuencas
(Laguna Matanzas gt17 Lago Vichuqueacuten gt60)
124
Figura 2 Comparacioacuten de los cambios del ciclo del N entre Laguna Matanzas y
Lago Vichuqueacuten con los procesos biogeoquiacutemicos contrastantes en rojo En L
Matanzas las condiciones oacutexicas podriacutean estar favoreciendo la nitrificacioacuten del
amonio En Vichuqueacuten las condiciones anoacutexicas favorecen un aumento en δ15N de
la MO en los sedimentos por desnitrificacioacuten y mineralizacioacuten
Los ambientes mediterraacuteneos en el que los lagos del presente estudio se
encuentran insertos estaacuten caracterizados por una estacioacuten seca prolongada y las
precipitaciones ocurren en eventos puntuales alcanzando altos montos
pluviomeacutetricos en poco tiempo Las lluvias favorecen un lavado de la cuenca la
perdida de N y otros nutrientes Por lo que es esperable que los sedimentos del
lago reflejen mejor los valores isotoacutepicos de los suelos de la cuenca durante los
periacuteodos con altos montos pluviomeacutetricos En el Lago Vichuqueacuten por ejemplo el
125
POM superficial (2 y 5 m de profundidad) despliega valores isotoacutepicos maacutes
positivos en invierno presumiblemente como resultado de mayores aportes de MO
y N de su cuenca (Fig 1b) En ambos lagos los valores maacutes positivos en los
sedimentos lacustres ocurren durante periacuteodos con mayores montos pluviomeacutetricos
(Cap1 Fig 6 y Cap 2 Fig 12)
Ademaacutes del efecto de la precipitacioacuten en el aporte de nutrientes al lago en
esta tesis hemos encontrado que los mayores aportes de N desde la cuenca se dan
cuando la cobertura vegetal del suelo es menor En Laguna Matanzas el desarrollo
de la actividad ganadera desde la llegada de los espantildeoles a la cuenca habriacutea
incrementado los aportes de N al lago Los valores de δ15N en los sedimentos
lacustres depositados durante este periacuteodo son los maacutes positivos de todo el registro
(~4permil) A su vez en Lago Vichuqueacuten los valores isotoacutepicos maacutes positivos se
registran entre 1300 y 1900 CE que coinciden con el mayor desarrollo de
actividades agriacutecola y de extraccioacuten de maderas de la cuenca (Fig 10 Cap 2)
Los recientes LUCC (a partir de 1975) en las cuencas de Laguna Matanzas
y Lago Vichuqueacuten estaacuten caracterizados por un incremento de la superficie forestal
y agriacutecola en reemplazo de la cobertura de pastizal (Laguna Matanzas) y bosque
nativo (Lago Vichuqueacuten) Los valores de δ15N en los testigos sedimentarios fueron
maacutes positivos cuanto mayor la superficie agriacutecola de la cuenca y maacutes negativos
cuanto mayor la superficie forestal Aunque por los alcances de esta tesis no
podemos ser concluyentes respecto del origen del NT (aloacutectonoautoacutectono) parece
ser que esta maacutes ligado a los aportes de la cuenca pese a la disminucioacuten del aporte
sedimentario observado en ambos lagos Las plantaciones forestales a diferencia
del bosque nativo son fertilizadas con una mezcla de N P y K (Toral et al 2005)
Por lo que pese a la disminucioacuten del aporte sedimentario la concentracioacuten de
126
nutrientes en el aporte al lago puede verse incrementada por el tipo de manejo
forestal con respecto al bosque nativo
Los resultados del primer capiacutetulo demuestran que 1) las plantaciones
forestales yo bosque nativo retiene maacutes sedimentos y MO mientras que el uso de
suelo agriacutecola y los pastizales destinados al desarrollo de la actividad ganadera lo
libera 2) Al parecer la desnitrificacioacuten es el proceso natural maacutes importante de
perdida de N en lagos costeros y favorece el enriquecimiento de 15N tanto en la
columna de agua (ie 15N-DIN) como en los sedimentos una vez enterrados y 3) la
desnitrificacioacuten depende de los cambios en las condiciones REDOX en el lago La
oxigenacioacuten en lagos poco profundos -como Laguna Matanzas- es altamente
fluctuante a escalas decadal-centenal depende de la profundidad de la laacutemina de
agua y de la variabilidad de la precipitacioacuten En este sentido en Laguna Matanzas
habriacutea condiciones anoacutexicas en ambientes cuando el lago alcanza niveles maacutes
altos Estas condiciones se observan entre 1820 y 1940 CE y son coetaacuteneas con
episodios maacutes huacutemedos y friacuteos (von Gunten et al 2009 Christie et al 2009) pero
tambieacuten con una fuerte actividad ganadera en la cuenca
Las fuentes variables de N y su fluctuacioacuten en el tiempo hacen necesario
contar con un anaacutelogo moderno que permite usar al δ15N de los sedimentos
lacustres como un indicador indirecto de los cambios en la disponibilidad de N en
el tiempo Por ello en el segundo capiacutetulo integramos muestras de suelo-
vegetacioacuten y un monitoreo de POM de la columna de agua en verano e invierno La
composicioacuten isotoacutepica de POM estariacutea reflejando cambios de la fuente de N (fijacioacuten
vs consumo del DIN) que variacutea durante el antildeo hidroloacutegico Durante el verano la
mayor temperatura y horas-luz favoreceriacutean las entradas de N al lago viacutea fijacioacuten
bioloacutegica de N2 atmosfeacuterico resultando en un incremento de la MO con un valor
127
isotoacutepico bajo (POM verano~5permil Fig 1b) El N2 (δ15N = 0permil) es fijado praacutecticamente
sin fraccionamiento isotoacutepico generando un δ15N de la OM cercano a 0permil (Leng et
al 2006) En invierno las precipitaciones pueden estar favoreciendo un incremento
en la entrada de nutrientes al lago El DIN de la columna de agua llega a contener
valores de δ15N maacutes altos reflejando estos mayores aportes de la cuenca y el POM
del Lago Vichuqueacuten queda enriquecido en 15N (δ15N ~11permil Cap 2 Fig 9) Estas
variaciones estacionales pueden estar evidenciando un cambio en la composicioacuten
de especies de POM desde especies fijadoras a especies que consumen el N de la
columna de agua Sin embargo para corroborar esta informacioacuten seriacutea deseable
contar con el anaacutelisis de la composicioacuten de especies de las muestras de agua
extraidas
Entre los resultados maacutes importantes estaacuten los valores isotoacutepicos del suelo
y la biomasa representativa de la cuenca que incluye un listado de las especies
nativas de la zona que hasta ahora no se contaba (Cap 2 Tabla en material
suplementario) Las especies colectadas despliegan valores de δ15Nfoliar maacutes
positivos (plantaciones forestales y pastizal ~89permil) que el suelo (35permil) salvo por
las especies nativas las que oscilan en torno a valores cercanos al atmosfeacuterico
(~14permil) La razoacuten isotoacutepica 15N14N refleja la dinaacutemica de intercambio de N entre la
vegetacioacuten y el suelo (Koba et al 2002) La discriminacioacuten es positiva en la mayoriacutea
de los sistemas bioloacutegicos por lo que las plantas tienen valores de δ15N maacutes bajos
que el suelo (Evans et al 2001) como sucede con las especies nativas en Lago
Vichuqueacuten En consecuencia las plantas quedan empobrecidas en 15N y conducen
a un enriquecimiento en 15N del suelo sobretodo si no hay reposicioacuten de N por otras
viacuteas (eg fijacioacuten de N) Por lo tanto los valores maacutes negativos de δsup1⁵Nfoliar de las
especies nativas pueden estar relacionados con el consumo preferencial de 14N del
128
suelo donde la discriminacioacuten isotoacutepica de la vegetacioacuten contra el 15N conduce a
valores del δsup1⁵Nsuelo maacutes altos (Houmlgberg 1997) Por el contrario los valores maacutes
positivos de δsup1⁵Nfoliar de las plantaciones forestales y pastos con respecto al suelo
puede deberse por una parte que el suelo no cuenta con mecanismos naturales de
reposicioacuten de N salvo por la descomposicioacuten de la hojarasca que puede ser maacutes
lenta en Pinus radiata que en bosque nativo (Lusk et al 2001) y por otra por el alto
impacto de los aportes de N (y otros nutrientes) derivado de las actividades
humanas (eg uso de fertilizantes) en el suelo
El alcance maacutes significativo de esta tesis se relaciona con un cambio en la
tendencia maacutes negativa de δsup1⁵N y el incremento del NT a partir de 1940 y a partir
de 1970 CE en ambos lagos La causa maacutes probable de esta tendencia es el
reemplazo de la cobertura natural o preexistente a 1940 CE por plantaciones
forestales
En la figura 2 se observa una siacutentesis de los principales procesos que
afectan la sentildeal isotoacutepica de la MO en los sedimentos lacustres de L Matanzas y
L Vichuqueacuten La entrada de N y MO desde la cuenca depende del tipo de cobertura
Las plantaciones forestales y el bosque nativo retienen maacutes nutrientes y sedimentos
en la cuenca mientras que la agricultura los favorece Sin embargo las praacutecticas
de manejo que incluyen la aplicacioacuten de N (y otros nutrientes) pueden aportar maacutes
nutrientes al lago que la cobertra de bosque nativo Cuando las actividades
forestales cubren una mayor superficie de la cuenca (1940 en adelante) δsup1⁵N oscila
en valores maacutes negativos y el NT tiende a incrementarse en los sedimentos
lacustres Cabe destacar que los valores de δsup1⁵N observados en los testigos
sedimentarios a fines del s XX no tienen precedente durante la conquista y colonia
espantildeola o durante el resto del periodo de la Repuacuteblica
129
Figura 2 Modelo de transferencia de N entre los ecosistemas terrestres y
acuaacuteticos El δ15N de los sedimentos lacustres depende de la interaccioacuten entre los
aportes de la cuenca y porcesos ldquoin-lakerdquo Flechas indican la direccioacuten del flujo de
N En azul las salidasentradas de 14N en rojo las salidasentradas de 15N δ15N de
la interaccioacuten plantas-suelo depende del tipo de cobertura de suelo
130
CONCLUSIONES GENERALES
La transferencia de N entre cuencas y lagos es un factor de control del ciclo
del N en los lagos y queda reflejado en la sentildeal isotoacutepica de los sedimentos
lacustres (Leng et al 2006 McLauchlan et al 2007) En Laguna Matanzas el
suelo de las especies nativas y las plantaciones forestales despliegan valores de
δ15N maacutes bajos (~1permil) que los de Lago Vichuqueacuten (~35permil) Coincidentemente los
sedimentos lacustres de Laguna Matanzas oscilan con valores de δ15N maacutes bajos
(-15 a +45permil Fig 1) que los de Lago Vichuqueacuten (+3 a +8permil)
Por otra parte el estudio de la MO de los sedimentos lacustres ha permitido
reconstruir los cambios en los aportes de MO y N desde la cuenca Aunque no es
posible afirmar queacute tipo de N estaacute entrando al lago (eg Nitroacutegeno orgaacutenico e
inorganico) si podemos decir que los cambios en las fluctuaciones de δ15N son
coincentes con el crecimiento de la actividad forestal (1980 - hasta 2016 L
Matanzas 0-17 y L Vichuqueacuten 1-66) El δ15N fluctuacutea hacia valores maacutes
negativos Ademaacutes se observa un incremento del NT en los sedimentos lacustres
cuanto mayor es la superficie forestal
Entre 1800-1940 las cuencas eran fundamentalmente agriacutecolas y
ganaderas (Cap1 Fig 7 y Cap 2 Fig 12) el δ15N de los sedimentos lacustres
oscila con valores maacutes positivos (L Matanzas +40 plusmn 05permil y L Vichuqueacuten 56 plusmn
033permil Fig 1) hay una baja bioproductividad en el lago (BrTi ~002 TOC~17)
lo que coincide con un bajo contenido de NT (~478 μg) Este es un periacuteodo de altas
precipitaciones (Christie et al 2009) que tienden a favorecer el lavado de la cuenca
131
y con ello el aporte de MO sedimentos y nutrientes desde la cuenca tiende a verse
favorecido Aunque las principales actividades humanas en estas cuencas son
diferentes (ganaderiacutea en Laguna Matanzas Contreras-Lopez et al 2014
agricultura en Lago Vichuqueacuten Odone 1997) en ambos casos hubo un reemplazo
de la cobertura natural de bosques nativo que favorecioacute mayores aportes de N y
sedimentos desde la cuenca en un efecto sumado con el aumento de las
precipitaciones
A partir de 1980 CE en ambos lagos se observa una disminucioacuten en los
valores de δ15N y un incremento del NT que no tiene precedentes en todo el registro
y pese a que ambos lagos son limnologicamente muy diferentes En Lago
Vichuqueacuten el δ15N tiende a oscilar en antifase con el NT (Fig 1) En el caso de
Laguna Matanzas se observa una tendencia hacia valores maacutes negativos y a partir
de 1990s a valores maacutes positivos mientras que el NT tiende a incrementarse de
manera sostenida Ello podriacutea estar relacionado con el crecimiento de la actividad
forestal habriacutea una mayor retencioacuten de N y sedimentos en la cuenca ligado al
incremento de la superficie forestal Ademaacutes en el caso de L Matanzas el
incremento de la actividad agriacutecola (sumado al de las plantaciones forestales)
podriacutea expicar el cambio en la tendencia en la deacutecada de los 1990s
En el contexto de Antropoceno esta tesis nos permite identificar un gran
impacto de las actividades humanas en Chile central a partir de la deacutecada de 1940
y una Gran Aceleracoacuten a partir de la deacutecada de 1970s En el registro sedimentario
de los lagos en estudio se observa una gran acumulacioacuten de MO el δ15N oscila
hacia valores maacutes negativos y un gran incremento del NT y el crecimiento de la
actividad forestal parece ser la principal responsable Ademaacutes la sobreexplotacioacuten
132
del agua junto al cambio climaacutetico global ha generado una combinacioacuten letal para
los lagos costeros de Chile central
Figura 1 Comparacioacuten de las fluctuaciones de δ15N durante los uacuteltimos 300
antildeos en Laguna Matanzas y Lago Vichuqueacuten
133
Referencias
Battye W Aneja VP Schlesinger WH 2017 Earthrsquos Future Is nitrogen the next carbon Earthrsquos Future Global Biogeochem Cycles 30 1000ndash1014 httpsdoiorg101002eft2235
Christie DA Boninsegna JA Cleaveland MK Lara A Le Quesne C
Morales MS Mudelsee M Stahle DW Villalba R 2009 Aridity changes in the Temperate-Mediterranean transition of the Andes since AD 1346 reconstructed from tree-rings Clim Dyn 36 1505ndash1521 httpsdoiorg101007s00382-009-0723-4
Contreras-Loacutepez M (Playa AUHVC (Universidad de VRFS (Pontificia
UC de V 2014 Elementos de la historia natural del An Mus Hist Natulas Vaplaraiso 27 51ndash67
Elliott EM Brush GS 2006 Sedimented organic nitrogen isotopes in
freshwater wetlands record long-term changes in watershed nitrogen source and land use SO - Environmental Science amp Technology 40(9)2910-2916 2006 May 1 40 2910ndash2916
Evans RD Evans RD 2001 Physiological mechanisms influencing
plant nitrogen isotope composition 6 121ndash126 Fogel ML Cifuentes LA 1993 Isotope fractionation during primary
production Org Geochem 73ndash98 httpsdoiorg101007978-1-4615-2890-6_3 Galloway JN Schlesinger WH Ii HL Schnoor L Tg N 1995
Nitrogen fixation Anthropogenic enhancement-environmental response reactive N Global Biogeochem Cycles 9 235ndash252
Garreaud RD Alvarez-Garreton C Barichivich J Pablo Boisier J
Christie D Galleguillos M LeQuesne C McPhee J Zambrano-Bigiarini M 2017 The 2010-2015 megadrought in central Chile Impacts on regional hydroclimate and vegetation Hydrol Earth Syst Sci 21 6307ndash6327 httpsdoiorg105194hess-21-6307-2017
Koba K Koyama LA Kohzu A Nadelhoffer KJ 2003 Natural 15 N
Abundance of Plants Coniferous Forest httpsdoiorg101007s10021-002-0132-6
Leng MJ 2006 Isotopes in Palaeoenvironmental Research Eos
Transactions American Geophysical Union httpsdoiorg1010292007EO070007 McLauchlan KK Craine JM Oswald WW Leavitt PR Likens GE
2007 Changes in nitrogen cycling during the past century in a northern hardwood forest Proc Natl Acad Sci 104 7466ndash7470 httpsdoiorg101073pnas0701779104
134
Phytologist N Oct N 2006 Tansley Review No 95 15N Natural Abundance in Soil-Plant Systems Peter Hogberg 137 179ndash203
Steffen W Broadgate W Deutsch L Gaffney O Ludwig C 2015 The
trajectory of the anthropocene The great acceleration Anthr Rev 2 81ndash98 httpsdoiorg1011772053019614564785
Turner BL Kasperson RE Matson PA McCarthy JJ Corell RW
Christensen L Eckley N Kasperson JX Luers A Martello ML Polsky C Pulsipher A Schiller A 2003 A framework for vulnerability analysis in sustainability science Proc Natl Acad Sci U S A httpsdoiorg101073pnas1231335100
Vitousek PM Aber JD Howarth RW Likens GE Matson PA
Schindler DW Schlesinger WH Tilman DG 1997 Vitousek et al 1997 Ecol Appl httpsdoiorg1018901051-0761(1997)007[0737HAOTGN]20CO2
Vitousek PM Howarth RW 1991 Nitrogen limitation on land and in the
sea How can it occur Biogeochemistry httpsdoiorg101007BF00002772 von Gunten L Grosjean M Rein B Urrutia R Appleby P 2009 A
quantitative high-resolution summer temperature reconstruction based on sedimentary pigments from Laguna Aculeo central Chile back to AD 850 Holocene 19 873ndash881 httpsdoiorg1011770959683609336573
Xu R Tian H Pan S Dangal SRS Chen J Chang J Lu Y Maria
Skiba U Tubiello FN Zhang B 2019 Increased nitrogen enrichment and shifted patterns in the worldrsquos grassland 1860-2016 Earth Syst Sci Data httpsdoiorg105194essd-11-175-2019
Zaehle S 2013 Terrestrial nitrogen-carbon cycle interactions at the global
scale Philos Trans R Soc B Biol Sci 368 httpsdoiorg101098rstb20130125
