Cavicchi et al. 1985. Variación en venación alar

8/12/2019 Cavicchi et al. 1985. Variacin en venacin alar

1/25

Copyright 0 1985 by the Genetics Society of America

T E M P E R A T U R E - R E L A T E D D IV E R GE N CE I NE X P E R I M E N T A L P O P U L A T I O N S OF DROSOPHILAMELANOGASTER. I. G E N E T I C A N D D E V E L O P M E N T A L B A S I S

OF W I N G S IZ E A N D S H A P E V A R I A T I O NSANDRO CAVICCHI, DANIELA GUERRA, GIANFRANCO GIORGI AND

CRISTINA PEZZOLIIstituto di Genetica, Universita degli Studi di Bologna, 40126 Bologna, Italy

Manuscript received February 28, 1984Revised copy accepted November 12, 1984

ABSTRACTThe effects of environmental temperature on wing size and shape of Dro-

sophila melanogaster were analyzed in populations derived from an Oregonlaboratory strain kept at three temperatures ( IS 0, 25, 28 ) for 4 yr. Tem-perature-directed selection was identified for both wing size and shape. Thelength of the four longitudinal veins, used as a test for wing size variations inthe different populations, appears to be affected by both genetic and maternalinfluences. Vein expression appears to be dependent upon developmental pat-tern of the wing: veins belonging to the same compartment are coordinated intheir expression and relative position, whereas veins belonging to differentcompartments are not. Both wing and cell areas show genetic divergence,particularly in the posterior compartment. Cell number seems to compensatefor cell size variations. Such compensation is carried out both at the level ofsingle organisms and at the level of population as a whole. The two compart-ments behave as individual units of selection.

NVIRONMENTAL temperature plays an important role in ecogeograph-E ical differentiation between populations, and much morphological varia-tion between geographic races of several Drosophila species has been inter-preted as an expression of genetic adaptation to temperature (STALKERndCARSON947; PREVOSTI955; MISRAand REEVE1964; DAVID nd BOCQUET1975a,b; DAVID,BOCQUETnd DE SCHEEMAEKER-LOUIS977).

Although temperature does not appear to be the only environmental factorinvolved in population differentiation (SOKOLOFF966; ANDERSON968), AN-DERSON (1 966) showed that temperature-directed selection for body (wing) sizeoccurs in the laboratory, making it plausible that similar selective forces areat work in the laboratory and in nature (ANDERSON973). In addition, POW-ELL (1 974) showed that tropical populations of Drosophila willistoni possessgenetic mechanisms to adjust body size to temperature that are similar to thoseof temperate species.On the other hand, attention has seldom been paid to shape variation(STALKERnd CARSON 947; MISRA and REEVE 1964), although it too mayGenetics 109: 665-689 April, 198.5.


2/25

666 S . CAVICCHI E T ALwell be an adaptive response to climatic differences over the species range. Onthe basis of multivariate analysis of nine metric traits, we observed (CAVICCHI,GIORGI nd MOCHI1978) that a laboratory population of Drosophila melano-gaster reared at two different temperatures (25 and 28 ) exhibits progressivewing shape divergence and no divergence in wing size during nine generations.Differential survivorship was associated with shape differences, supporting thehypothesis that the shape variability has an adaptive meaning in relation totemperature.

Shape variation arises from interactions between genes controlling differenttraits during development and pattern formation. Therefore, in order to de-termine the genetic basis of shape and its evolutionary significance a dualapproach is necessary, employing quantitative developmental genetics and pop-ulation genetics.

Some connections between quantitative and developmental genetics haverecently been provided as regards wing pattern formation. GARCIA-BELLIDO,RIPOLL nd MORATA 1973) showed that the wing blade appears to be subdi-vided in earlier development into two distinct portions originating from pri-mordial cells in the imaginal disc, whose descendants (polyclones) never crossa certain demarcation line. This line represents the boundary of the two wingcompartments (posterior and anteior) which show different and specific celldynamics during development (GARCIA-BELLIDO977).

The compartments appear to be units for the genetic control of developmentand responsible for size and shape of organs and appendages (LAWRENCEndMORATA1976). Very recent studies on quantitative inheritance in Drosophilasupport this idea: CAV ICCH I, EZZOLInd GIORGI 1981) and CAVICCHIt al.(1981) studying variations in body dimensions in lines selected for short wing,or in lines exposed to sudden temperature variations, found that metric traits(vein lengths and their distances) belonging to the same wing compartmentwere well coordinated in their phenotypic expression. THOMPSON,ELLACKand KENNEDY (1982) and THOMPSONnd WOODRUFF1982), dealing withpolygenes affecting the expression of vein mutants, found an interdependencebetween modifiers affecting veins within the same compartment.

The developmental genetic reasons for size and shape variations may beinvestigated by studying cell size and number variation. The contribution ofcell size and number in determining wing surface varies in relation to theorigin of the studied populations. Inadequate nutrition and artificial selectionfor body size (ZARAPKIN 934; ROBERTSON 959b) appear to be associatedwith changes in cell number, whereas variations in body size produced bytemperature seem solely related to variations in cell size (ALPATOV 930; ROB-ERTSON 1959a). Moreover, cell size appears genetically determined as shownby the selection procedure (ROBERTSON959a). It has also been suggested thatcell number may undergo genetic control by genes acting on the mitotic rateduring larval growth of the wing disc whose expression might also be depend-ent on genes controlling the two wing compartments (GARCIA-BELLIDO,IPOLLand MORATA1976).On the basis of these findings we have attempted to connect developmental


3/25

WING SIZE AN D SHAPE VARIATION 667t l m o o f r o a r i n #OREGON - 18'

2 0 y o o r r

A - 1 8 ' b-26' c-28'4 y o r r r

( f r o m c o m m o n p a r o n t r )

2 mo n o r o t i o n s( f r o m c o m m o n p r r o n t r )

FIGURE .-Experimental design. Egg samples from 20 pairs of flies from the Oregon strainkept for 20 yr at 1 8 were placed at 18 , 25 and 28 to found the A, B and C populations,respectively. After 4 yr of mass rearing, from each population 18 pairs were chosen and allowedto lay eggs for 24 hr in their own environment and were then transferred to the other twotemperatures for I-day laying. Measurements were performed after an additional single-culturegeneration at each temperature.genetics to population genetics by studying the genetic basis of size and shapevariations induced by temperature in the Drosophila wing in relation to itsdevelopmental pattern. It is known that wing size is highly correlated withbody size (R OBER TS ON 1962; TANTAWYnd RAKHA 1964; SOKOLOFF966;C A V I C C H I , PEZZOLI nd GIORGI1981) and fitness components (TANTAWYndVETUKHIV 960; TANTAWYnd RAKHA1964; PIERAGOSTINI,ANGIORGIndCAVICCHI979). Therefore, wing size may be confidently used as an index ofbody size in a given population.

18' 2 6 ' 28 18' 26 28' 18' 25 28'

MATERIALS AND METHODSThe starting population consisted of an Oregon laboratory strain reared for almost 20 yr at a

constant temperature of 18 . W e preferred a laboratory rather than a wild strain because weconsider that 20 yr of continuous rearing at this temperature represent a sufficieint time to makeit well adapted to its environment. In addition we were encouraged by the fact that the straindisplayed a heritability of 0.3 for both thorax and wing length when estimated by both Lerner'smethod and parent-offspring regression.

Eggs laid from 20 pairs of flies were used to found new mass populations (Figure 1): the first(A) was maintained at the same 18 temperature; the second (B) and the third (C) were kept for4 yr (about 140 generations) at 25 and 28 , respectively. The last temperature represents theupper extreme thermic environment for the survival of our strain; few progeny were obtainedfrom each generation until the 1st yr of rearing when a gradual improvement was seen. After 4yr. 18 pairs were chosen from each population (A, B and C) and allowed to lay for 1 day at eachtemperature (18 , 25 , 28 ).The flies belonging to the A population transferred to 28 and those from the C one transferredto 18 displayed a ve ry low fitness, whereas those from the B population showed a good survi-vorship at all temperatures.To minimize physiological effects determined by previous rearing at different temperatures,one subsequent generation was reared in each environment: two pairs within each progeny were


4/25

668 S. CAVICCHI E T AL

FIGURE .--Graphic representation of the Drosophila wing. Th e region chosen to represent theposterior (P) compartment is delimited by L4, L5 veins, measured as straight lines (AE = L 4 andAF = L5) and their distance (EF = L4-L5 distance). The region delimited by L2, L3 veins (BC= L2 and BD = L3) and their distance (CD = L2-L3 distance) was considered to represent theanterior (A) compartment.

randomly sampled and reared again in single culture, changing vials every day for 3 days. Theprogeny number obtained in the fittest vials never exceeded 40, so that, considering the quantityof standard medium in a culture (no less than 15 ml), crowding was avoided. Th e progeny obtainedfrom only one of the two pairs and, hence, from a total of nine families (replicate pairs) weretaken for measurements. The right wing of seven females from each family (when available) waspulled out and mounted on a slide. The length of L2, L3, L4 and L5 veins and the distance, atthe margin of the wing, between L2, L3 and L4, L5 veins (Figure 2) were measured under amicroscope at magnification X50, with an ocular micrometer of 100 divisions; all measurementswere then converted into millimeters X 10'. Because the boundary between the two wing com-partments lies just beside the L 4 vein, different effects induced by temperature on the two groupsof veins were taken as wing shape variation resulting from changes in the developmental patternof the wing.

Size and shape differences between populations and transfers were analyzed by univariate andmultivariate analysis. The univariate analysis of size was performed by a joint analysis of variancebetween populations and transfers. Of the multivariate statistics we chose Mahalanobis' distance(0 ) hich has been considered the most appropriate method for size and shape analysis (ATCHLEY1980; KUNKELand CHERRY 980). Total 0 was partitioned into size and shape componentsfollowing the same theoretical background proposed by SPIELMAN1 973) and TEMPLETON1977) .Our method is essentially that used by Spielman except that we transformed our raw data intovalues on uncorrelated coordinates by the pivotal condensation method (see RAO1970) computerprogrammed by MURTYand ARUNACHALAM1967) instead of principal component analysis. Asfar as correlated traits are concerned, some components corresponding to very low eigenvaluescan be discarded by principal component analysis.These represent minor axes of variation withina group but not necessarily minor axes of variation between groups. The test for significance ofD2, ize' and shape' was based on F-test of the Hotelling T .

Variations in cell size and number were evaluated considering two wing surfaces that approxi-mately correspond to the wing compartments (posterior and anterior). For the posterior compart-ment the area of the triangle whose sides are represented by the L4 and L5 veins and the distancebetween them at the margin of the wing was estimated, and for the anterior compartment thearea delimited by L2, L3 veins and the distance between them was considered (Figure 2) . Areaswere computed by Erone's formula

where p is half the perimeter and a, b and c are the lengths of the sides of the triangles.number of bristles (cells) present on a dorsal surface of 96.8 X

Cell area was estimated under a microscope at a total magnification X275 by counting themm' limited by a reticle


5/25

WING SIZE AND SHAPE VARIATIO N 669placed in the eyepiece. The average cell area was estimated by dividing the area of the reticle bythe number of cells counted. For the posterior compartment the reticle was placed close to thecrossvein and at the same distance from the end of the wing in the anterior compartment. Celldensity differs in different wing regions and does not appear to be compartment dependent;nevertheless, the cells are regularly arranged on the wing surface, and counts in different regionsare quite well correlated (ROBERTSON 959a; DELCOURand LINTS 1966). In this experimentdisruption of the correlation will, therefore, be considered evidence of a different response of thetwo compartments to temperature.

Cell number was obtained by dividing wing size by cell area; because areas and number of cellswere not found normally distributed, the measurements were converted to natural logarithms(ROBERTSON959a). In this form, wing area is the sum of cell area and number.

Genetic differences among populations were detected by crossing the original A population withthe B and C ones. The crosses were performed two generations after transferring the populationsto a common environment at 25. Parents and Fls were raised simultaneously with FZS.All ofthe reciprocal crosses were made. Seven female right wings from each of ten families per reciprocalcross were measured.

RESULTSVein lengths and their distances: Mean values of vein lengths and their dis-

tances in the B and C populations kept for 4 yr at 25 and 28 , respectively,and in the original population (A) two generations after their transfers at thethree temperatures are shown in Figure 3. The populations (A, B, C) appearquite differentiated within each environment (18 , 25 O f 28 ), showing a per-sistent temperature-dependent cline in wing size most evident at 18 fliesreared at warm temperatures show smaller wings in all environments consid-ered. Different traits show quite different responses both as effects of transferand of prolonged rearing at different temperatures; the distance between L2and L3 veins (anterior compartment) shows the smallest variation.

The analysis of variance (ANOVA) given in Table 1 provides estimates forthe effect of transfers on both populations and genotypes within populationsas well as the effect of prolonged rearing at different temperatures. Highlysignificant differences between temperatures and between population meansquares are observed, whereas the amount of variance between families isalmost the same, or less than family X temperature interaction. Population Xtemperature interaction is not significant when compared with the betweenfamilies for L3, L4 veins and for the distances between L4-L5 and L2-L3veins, whereas significant interaction family X temperature is observed. Be-cause the families are the replicate pairs of flies sampled from each populationat each temperature, the between families variance represents variability dueto genotypes within populations. Accordingly, the family X temperature inter-action represents a genotype-environment interaction, where genotype meansaverage genotype because flies within any family may differ genetically dueto segregation and recombination. Intraclass correlations quoted in the sametable summarize the results. Intraclass correlation represents a measure of theresemblance between individuals within each class; if the components of thevariances between and within are summed, the intraclass correlation is theratio between the class component and the sum of components.N o difference between veins belonging to different compartments can beobserved; in particular, L3 and L4 veins and likewise L2 and L5 show quite


6/25

670 S. CAVICCHI ET AL.

240

220

200-

W O -

W O -

--

Anterior

60

40

d L 2 - L 3

I18 1 1 125 28 181 I

25 2 8

T E M P E R A T UR E S (transfor.)FIGURE .-Mean values o f the differen t wing traits measured within compa rtments in the A,B and C populations transferred to three temperatures IS0, 5" and 28").

similar behavior. This is not surprising, since both L3 and L 4 veins seem tobe affected by polygenes associated with chromosome ZZZ (THOMPSON975;THOMPSONnd WOODRUFF1982), whereas those for lengthening the L2 veinare associated with chromosome ZZ; he L5 vein seems affected by polygenesand modifiers located on both chromosomes ZZ and ZZZ (THOMPSONnd WOOD-RUFF 1982). On the other hand (Figure 3), we can see that the distances


7/25

WING SIZE AND SHAPE VARIATION 671

+U


8/25

672 S . CAVICCHI ET ALbetween veins in the posterior and anterior compartments show a quite differ-ent behavior, suggesting that the way in which the veins are placed in the twocompartments changes as an effect of temperature.

Both size and shape variations induced by temperature have been studiedby multivariate analysis. Tables 2 and 3 summarize the average size and shapedistance between temperatures within populations (Table 2) and between pop-ulations within temperatures (Table 3). The analysis was performed on boththe wing traits taken as a whole (posterior and anterior) and on those belongingto either the anterior or posterior compartment considered separately. Thesums of the statistics obtained on each compartment (posterior plus anterior)are reported too.

Obviously, the distance (0 ) obtained as a sum is overestimated and largerthan that obtained in the joint analysis (posterior and anterior), in which thecorrelation between traits of different compartments is taken into account. Inaddition, when a different number of characters is dealt with, the significancelevel is not the same in different analyses (see Table 2 and 3 footnotes).Significance of the shape components within compartments reveals only allo-metric variations in the characters belonging to the same compartment. There-fore, in order to compare the behavior of the two compartments, size' andshape' percent on total 0 are also given in the tables both for the sum(posterior plus anterior) and the joint analysis (posterior and anterior).

If the characters belonging to different compartments are subjected to dif-ferent controls for their expression, the amount of shape variation in the jointanalysis must be greater than that obtained in the sum of the within-compart-ment analyses (posterior plus anterior).

If w e first consider the effect of transfers (between temperatures, Table 2)on the original population (A) w e note that the size component largely accountsfor the phenotypic changes observed. The percent contribution of size andshape is almost the same in posterior plus anterior and posterior and anterior,suggesting that the two compartments react quite similarly to transfers. In theB and C populations the amount of the shape contribution increases; moreover,in the population exposed to more severe environmental selection (C), thecontribution of shape is 11% greater in posterior and anterior when comparedwith posterior plus anterior, indicating that the two compartments react dif-ferently to transfers.

In the between-populations within-temperatures analysis (Table 3), total Dappears markedly reduced but highly significant. It is noteworthy that thecontribution of shape is proportionally increased, being almost the same orlarger than that of size in the posterior and anterior analysis. The percentdifferences between shape contributions in posterior plus anterior and poste-rior and anterior are v e r y large (19-3 1%), indicating a great deal of compart-ment x population interaction. The prolonged rearing at different tempera-tures, therefore, seems to have produced different genetic effects on the twocompartments.

In general, when real values are considered, the largest discrimination valuesare observed at 18 which represents the original environment; the selected


9/25

WING SIZE AN D SHA PE VARIATION 673

Y)ti::s.k

UQ

Le

F:T3

3

U

$%L9e

U8r,

g3 s

(U

L5 23 %

W ?

$ 3W F :4 sk -L32&

t2

U

9F:UBnvs-U-UU

El

2E l -Y._t2

* * * 85m 1 4 2: 4 8

.5z% 2c 2g 2 :i: Q f - 3U v) c ? o @?$E 2plh 2 s0 .gsE %* .s :

630 .1 a: 3 2v) c9d: @ ? ? ,:

z 2

* *m - a g - 0Q m -- 0 . 1 m m L a

01- 3 Q c

W OL O

m * *8 * *0 . 1 030 1 22 h s0 .z* * * %$ 5

01 - m a a Z 2 5Q 2: 9. 2 - g . g0.10.1 6 3 m ga hm B- E.2 8

D( 82 .> 0 5g 2s - 0 )5 2 34 sg- W 2 Q + mnz e . ; 3s8 c , X Uzc 3 3 m o0 - 5 2wc ? N . 2 e s0 6 3 mt. 9 8 mc,

oz.&a3 3 3* * : 82 . c 5Q ? c ? 22 2 s 2.2r - m0.1 rT] * ;.s . .c.s

0 2-5 0a y : % 3A Nu 2.z y) czc a h U U -* m M L Z W* * B-Cr I.2 0< c r-0.1 m a mLz e g . zv, 8 % 22 c-;*$CI >:ss:g 2 2 & Us s . Z g g 0.2E = O m - e q* : z m : z s8- :: * U Y n g w o - . w220 jz,s.s:Jm0.1 I n m E C m : : * 222 8 z 2 Y

I % 8 grin 2 2.- z5 5 5 5 8 1 - 2 = c c u Gc c B M.20 0 5 . 2E Y f o S J s E l O g k b W *U a z 2 5 2 4 & . . , * *

* * * V Im W

- E0

3 ct +m W m myd * *

sD,

* c-m8-) :m m & b ? Y Uz E.:.10.1 6 3 6 3

* *m r C O o a

m w -

v) 0 -qfit. - O O 2C O O @

CY Y

2 d d y s C M0 - c s . - ma b 0.a 5 9


10/25

674 S. CAVICCHI ET AL* * : a* *5 2 2 + aq q& & V m

U1

E h hm V4 * * ;E & g V a l m -2 0.I 2 2 Y 5

s01 N 2EY s a * *s .-) * o o *:-e ' ? c9t9U o . ( o

5. _ w * * o.(i , r j&s h

N Z Z

N a* *+ -hp a._

w -

0

h * * *9' Q 5 5 a -3 : g * *% 2 . 2% v ) I n al01

h - 0

? c 9 2:w 9 '

N 5- 1 sa se a 8 * E ;c1 z & & m -i ;E 2: 2;

0 0h 5m $ ?alo

U._U-3$2Q As

N m wf- m l o

* ** * 29 c ? ? 4':a * zalPI 0 1 - 0 ) -B

s ggn & 5 zz 22

w * * *% g * * *Y--U nU *

qc 0 0 :

I n --U csN ::.-) m * a

a m 2-F

4vEeE

FL + .)19 g E g g

dN -Nv).-2

Cavicchi et al. 1985. Variación en venación alar

Documents

Transcript of Cavicchi et al. 1985. Variación en venación alar