Heredity 68 (1992) 557-563 Received 31 July 1991 Genetical Society of Great Britain

The evolutionary history of buzzatii. XXIV. Second chromosome inversions have differentaverage effects on thorax length

ESTEBAN HASSON, JUAN J. FANARA, CONSTANTINA RODRIGUEZ, JUAN C. VILARDI*, OSVALDO A. REIG & ANTONIO FONTDEVI LA t GIBE Departamento de Ciencias BiolOgicas, Facultad de Ciencias Exactas y Naturales. Universidad de Buenos Aires, Ciudad Universitaria Pab. II, 1428 Buenos Aires, Argentina, *Laboratorio de Genética, Departamento de Ciencias B/old gicas, Facultad Ciencias Exactas y Naturales, Universidad de Buenos Aires, Ciudad Universitaria, 1428 Buenos Aires, Argentina and tDepartmento de Genética y Microbiologia, Un/versidad AutOnoma de Barcelona, 08193 Bellaterra (Barcelona), Spain

Wedemonstrate a genetic correlation between rearrangements of the second chromosome of D. buzzatii and thorax length, as a measure of body size. The results indicate that 2j and 2jz3 arrange- ments are correlated with large size, whereas 2st arrangement is correlated with small size. Some inversions (2st and 2jz3) show dominant effects and others (2j/f3) exhibit overdominance. These results show that at least 25 per cent of body size variation may be accounted for by the studied . The possible integration of the genotypic, phenotypic and fitness levels, and also the possible implications to life-history theories, are discussed. These results suggest that, under moderate to high heritability values, some kinds of chromosomal endocyclic and/or balancing selection may be valuable mechanisms for maintenance of body size variation.

Keywords: bodysize, chromosomal inversions,Drosophila buzzatii, fitness, selection, life-history traits..

ness-related morphometric character that is easy to Introduction observe and measure. Body size is one such trait. Although the adaptive basis of chromosomal poly- Several studies have shown that thorax length, as a morphisms in Drosophila is universally accepted, the measure of body size is correlated with adult fitness selective forces involved are only inferred from components, e.g. fecundity (Robertson 1957a, b), indirect evidence from natural and experimental success (Butlin eta!., 1982; Taylor eta!., 1987; populations (Sperlich & Pfriem, 1986). Direct Partridge et aL, 1987; Santos et a!., 1988; Taylor & demonstrations of the selective significance of these Kekié, 1988), and longevity (Partridge eta!., 1987). inversion polymorphisms in nature are practically non- Nonetheless, the existence of a relationship between existent (Endler, 1986). In recent years, however, it has the phenotype and the chromosomal rearrangements is been possible to show that selection acts directly on rather controversial. Lande (1979) and John (1983) these polymorphisms through several fitness com- claim that chromosomal repatternings have no effect ponents in natural populations (Anderson et a!., 1979; on the morphological exophenotype. However, a posi- Ruiz eta!., 1986; Salceda & Anderson, 1988; Ruiz & tive relationship has been postulated in some Santos, 1989; Santos eta!., 1989; Hasson eta!., 1991). species, although a significant correlation has only been Bearing in mind that selection acts primarily at the demonstrated in a few cases (White & Andrew, 1960; phenotypic level and only secondarily at the genotypic White et a!., 1963; Butlin et aL, 1982; Colombo, one, population biologists have been searching for a fit- 1989). As regards species of Drosophi!a, a relationship between body size and chromosome Correspondence: A. Fontdevila, Department of and Evo- lutionary Biology, University of , Irvine, California was also reported for the D. subobscura A chromo- 92717, U.S.A. some (Krimbas, 1967) and for the D. pseudoobscura 557 558 E. HASSON ETAL. third chromosome (Thomson, 1971). Additional evi- 2st, respectively. The karyotypes of each collected dence pointing to the effect of chromosomal variation female and her mate were identified by analysing the on morphological characters comes from the corre- polytene chromosomes of 11 progeny larvae. In a few lated response of chromosomal frequencies to pheno- cases both parents were homozygous for the 2j typic artificial selection in D. subobscura (Prevosti, arrangement and these isofemale lines were selected as 1960; Prevosti 1967) and D. pseudoobscura (Druger, homokaryotypes. However, most of the lines were 1962). segregating and the homokaryotypic lines had to be Data on the genetic basis of body size in natural obtained by progeny selection of full sib crosses. This populations is of high relevance to the understanding progeny testing was determined with an accuracy of the evolutionary consequences of natural pheno- higher than 95 per cent by the cytological analysis of 13 typic selection. In D. buzzatii, Santos et al., (1988) and progeny larvae (Barbadilla & Naveira, 1988). The lines Ruiz & Santos (1989) have found that body size is did not attain fixation simultaneously and some lines positively correlated with mating success in a Spanish were kept by mass culturing until all the homokaryo- natural population. These studies have also shown that typic lines were obtained. The maximum number of different karyotypes differ in their mean value for generations before fixation was five. thorax length (Ruiz & Santos, 1989). In this paper we The experimental homokaryotypic stocks were present data showing that non-inbred with founded with eggs collected in optimal conditions from different second chromosome karyotypes extracted egg-laying cages, similar to those described by Ruiz et from an Argentinian natural population differ signi- a!. (1986). For each homokaryotypic stock, a total of ficantly in their thorax lengths and we give some infor- 150 mature (4—5 day-old) virgin males and 150 mature mation on the genetic determination of the chromo- virgin females of the same homokaryotypic lines were somal polymorphism on body size. introduced into one of these cages. Each homokaryo- typic line contributed with equal numbers of flies to this founding population. A large sample of collected Materialsand methods eggs was seeded in a culture and this stock was main- tained by mass culturing during four generations The species Drosophila buzzatli before the experiment was initiated. D. buzzatiiis a cactophilic species of the repleta group (mullen complex) that breeds mainly in the rotting cla- Derivationand analysis of different karyotypes dodes of cacti belonging to several Opuntia species. It Atotal of nine (Table 1) homo- and heterogametic probably originated in Argentina (Fontdevila, 1981; crosses were conducted by introducing 100 mature Fontdevila et al., 1982) and has successfully colonized (4—5 day-old) virgin males and 100 mature females of the Mediterranean region (Fontdevila et a!., 1981; the homokaryotypic stocks to egg collecting chambers. Fontdevila, 1981; 1989) and Australia (Barker, 1982) Reciprocal heterogametic crosses were performed in following its natural host plants. Chromosomally, this order to test possible maternal effects. species is highly polymorphic mainly for its second Eggs were allowed to hatch and first instar larvae chromosome (Ruiz eta!., 1985), and fitness differences were seeded in culture vials at optimal density, i.e. 40 have been demonstrated among these arrangements larvae in 6 ml of culture medium, as determined by (Ruiz eta!., 1986; Hasson eta!., 1991). Fanara (1988). Eight to ten replicates were run simul- taneously for each cross. Cultures were maintained at 25°C under continuous light until all adult flies Derivationof homokaryotypic stocks emerged. A random sample of each progeny was Weanalysed flies derived from a collection performed measured 24 h after individual emergences. Thorax in May 1987 at Arroyo Escobar, a locality situated 30 length was measured to the nearest 0.02 5 mm with a km north-west of Buenos Aires city (for a description binocular microscope fitted with an ocular micrometer. of the population, see Fontdevila eta!., 1982). Previous The measure was taken from the anterior margin of the studies showed that this population is polymorphic for thorax to the posterior tip of the scutellum, as laterally four arrangements in the second chromosome (st, j,jz3 viewed. All measurements were done by one of us and jq7) and two in the fourth (Stands) (Fontdevila et (JJF). a!., 1982). A modified formula of David's killed yeast culture Three homokaryotypic stocks involving 17, 14 and medium (David, 1962) was employed. Polytene chro- 4 independently derived chromosomes were obtained mosome slides were obtained following the technique for the second chromosome arrangements: 2j, 2jz3 and described by Fontdevila eta!. (1981). CHROMOSOMAL INVERSION EFFECTS ON DROSOPHILA SIZE 559

Table 1 Mean thorax length and the corresponding Statistical analysis standard error for the different second chromosome karyotypes of D. buzzatii. Sample sizes are indicated in ANOVAwith sex and as fixed factors and replicates as a random factor nested in karyotypes was parenthesis employed for the analysis of data. The mixed model Number ofMale thorax Female thorax employed can be expressed as: Karyotypes (*) replicateslength (mm) length (mm)

X,Jk/=+ af+/3J+FkO)+(a/3)+(aF)k+ EkI st/st 10 1.0273±0.03041.1288±0.0277 (116) (119) whereX,Jklisthe individual value; 1u is the overall mean; j/j 10 1.0396 0.02871.1303 0.0280 a1 is the sex effect, fixed; j3. is the karyotypic effect, (135) (131) 10 1.0530 0.02941.1499 0,0262 fixed; Fk(J)isthe replicate factor nested in karyotypes, jz3/jz3 random; (a/3 )is the sex by karyotypic interaction; (158) (161) 8 1.0365±0.02761.1347±0.0229 (a.F )ikisthe sex by replicate interaction; and e,,klisthe st/i random error. (118) (130) 10 1.0215±0.02881.1285±0.0253 This ANOVA analysis gives only a partial view of i/st the karyotype effects on thorax length. We have (152) (146) st/fr3 10 1,0242±0.02671.1206±0.0257 extended our analysis by fitting our data to a diallel (160) (137) design (Mather & Jinks, 1982) in which all possible iz3/st 10 1.0410±0.02821.1473±0.0288 crosses are performed among a set of fixed homo- (146) (146) karyotypic lines. This diallel design can be tested by a iliz3 10 1.0647 0.02281.1532 0.0225 two-way ANOVA with paternal and maternal karyo- (151) (146) types as fixed factors, with each cell observation as the jz3/j 10 1.0710±0.02411.1701±0.0279 average thorax length for each replicate. (136) (153) A posteriori comparisons were performed utilizing the method of Scheffe. This and all statistical methods (*)The chromosome of male parental origin is indicated in employed can be found in Sokal & Rohlf (1981). Com- the first term. putations were performed with SPSS and BMDP stati- stical packages. Table 2 Analysis of variance testing for differences in thorax length among second chromosome karyotypes of Results Drosophila buzzatii Themean thorax length of males and females bearing Source of different second chromosome karyotypes is shown in variation d.f. SS MS F P Table 1. The ANOVA is shown in Table 2. There are Sex 1 6.246.24 2466,240.000 highly significant differences among sexes and among Karyotypes8 0.660.08 31.620.000 karyotypes. The among-replicate (within karyotype) Sex by variation is also highly significant which suggests that 8 0,022,50x io- 0.990,450 despite the homogeneity of the experimental condi- karyotype Replicates79 0.202,53 x i0 3.890.000 tions, environmental variance could be relatively high. Sex by On the other hand, there is neither sex—karyotype replicates79 0.078,90X i0 0.350.999 interaction nor sex—replicate interaction, indicating Error 23661.536.50x104 that thorax length variation behaves independently of Total 25418.72 sex across genotypes or laboratory environments. The between-sex consistency of the mean thorax lengths (Table 1) was also analysed by using the Kendall rank correlation coefficient (Sokal & Rohlf, 1981). The the total variance. The results of ANOVA performed value of this statistic (n =64)was highly significant independently in each sex were strikingly consistent indicating that the rank order of karyotypes, according between sexes (data not shown), showing that about to their thorax length, is not statistically different 25—29 per cent of size variance may be accounted for between sexes. by the karyotypic factor. The among replicate com- The relative importance of karyotype on thorax ponent showed an acceptable value (9 per cent) for length determination was measured by estimating the most experimental designs. The residual error, which percentage contribution of each source of variation to represents a very important fraction of the total van- 560 E. HASSON ETAL.

Table 3 Two-way analysis of variance according to a diallel variation) can be assigned to 2jz3inthe direction of design testing for differences in thorax length among second increasing body size. No significant contribution of 2j chromosome inversions of D. buzzatii transmitted via male to size variation has been detected. and female parents Interaction between male and female parents is Source of highly significant in both sexes (Table 3), which variation d.f, SS MS F P suggests deviations from additivity. In order to under- stand these non-additive effects, the thorax length of Males different karyotypes carrying 2, 1 or 0 chromosomes Parental with each rearrangement were compared using the a males 2 0,011040.0055239.08 0.000 Parental posteriori Scheffe's method. Rearrangements 2St and females 2 0.007040.0035224.900.000 2jz3 exhibit dominant effects of opposite signs. Thus, Interaction 4 0.007510.0018813.28 0.000 homo- and heterokaryotypes for these rearrangements Error 79 0.011160.00014 exhibit non-significantly different sizes (St/St versus St/ Females +: male F879 =0.116,female F879 =0.359;Jz3/Jz3 Parental versus jz3/ +: male F879 =0.059,female F879 =0.027), males 2 0.010220.0051132.270.000 but both karyotypes are statistically different from non- Parental bearing (+ /+)flies (St/St: male F= 5.84, female females 2 0.001480.00074 4.670.012 F=4.35; St/+: male F=12.78, female F=6.36; Interaction 4 0,006450.00161 10.19 0.000 jz3/jz3: male F3.46, female F=3.37; jz3/+: male Error 79 0.012510.00016 F= 5.98, female F= 6.29. All F values significant at P<0.01 for 8 and 79 d.f.). Similar tests for over- dominant effects show significant values only for the ance (62—66 per cent), may be mainly explained as a 1hz3 heterokaryotype (male F879=7.26, P<0.01; consequence of the genetic background. female F8,79 =4.45,P <0.01). In order to obtain a further understanding of the effects of different arrangements, two types of analyses Discussion were performed: (a) diallel analysis, and (b) Scheffe's paired comparisons. Table 3 shows the two-way Latitudinalphenotypic variation in body size (Prevosti, ANOVA using a diallel design. Differences among 1955; Misra & Reeve, 1964) has been interpreted as a arrangements are highly significant in both parental genetic adaptation (Pfriem, 1983), but a pure pheno- sexes and there are no differential contributions typic response to environmental variation cannot be between sexes to the total variance. The latter can be excluded. In the D. buzzatii natural population of tested by using the F statistics produced by the ratio of Carboneras (Spain), the significant correlation between female-to-male parent variance components. None body size and karyotype frequency (Ruiz & Santos, of these F ratios is statistically significant (males: 1989) may not have a genetic cause. The authors F22 =1.57,P= 0.33; females: F22 =6.91,P= 0.13). mention at least two non-genetic explanations. Firstly, This suggests that maternal effects do not exist. A body size and karyotype might be independently corre- corroboration of this has been provided by paired lated with longevity. Secondly, if karyotype polymor- comparisons among means of reciprocal crosses using phism is subjected to multiple niche selection, some Scheffe's method. Differences were only significant in karyotypes may select niches that favour a specific one out of six possible comparisons and, consequently, adult size. The results presented here may rule out maternal effects were considered negligible and thorax both environmental explanations. Our experiment is lengths of reciprocal heterokaryotypes were averaged performed with flies of the same age reared in similar for future comparisons. conditions. The percentage of total- variance assigned The relative contribution of each rearrangement to to genetic (karyotypic) causes is high-(25--29 per cent) body size variability can be assessed by multiple step- in both sexes. A similar conclusion can be inferred wise regression analysis (BMDP2R subprogram, in from a recent laboratory study in the population of BMDP, 1988) of body size on increasing doses of each Carboneras (Spain) (Ruiz et al., 1991). rearrangement. St rearrangement decreases size signi- It can be argued that the low values of thorax length ficantly (r male value: —0.1070,P< 0.01; r female of 2st/st karyotypes may be a consequence of the value: —0.0130,P<0.0i) and makes a large contribu- inbreeding depression produced by the low number of tion to size variation (32 per cent averaged among independently derived 2st chromosomes that origi- sexes). A smaller, although significant contribution (2.7 nated the homokaryotypic stock. This seems unlikely. per cent) to male size variation (but not to female size When 2st/st homokaryotypes are compared with CHROMOSOMAL INVERSION EFFECTS ON DROSOPHILA SIZE 561

2st/ + heterokaryotypes, non-significant size differ- found by some authors in natural populations of Dros- ences are obtained, a result that contradicts the puta- ophila (Prout & Barker, 1989, but see Coyne & tive depression of 2 st/st. Moreover, the esti- Beechan 1987). Low heritabilities due to high values of mated mean size of the 2s1/st karyotype (Table 1) environmental variance result in nearly neutral genetic shows no significant increase in its standard error, as variance and no balancing selection is necessary to expected if this line was highly inbred. explain this variation. Thus, although the measure of In all these studies with D. buzzatii there is a striking heritability in nature is not easy, it is of utmost import- coincidence in the effects of second chromosome ance to disclose the mechanisms of maintenance of arrangements on body size. The lowest size always morphological variation. In this endeavour, the ecolo- corresponds to the 2 st/st and the effect of arrange- gical structure of populations plays a leading role. ments of the jphylad(j,jz3,jq7) tends on average to Some ecological studies performed with D. buzzatii increase in size. The karyotypic effects are similar in (Santos etal., 1989; Prout & Barker, 1989) have shown field studies (Ruiz & Santos, 1989; Ruiz et a!., 1991), that the ephemeral patchiness of rotting cladodes sub- although the among karyotype differences are statisti- divides the populations in a way that positively cally significant only in males and when the 2j phylad enhances not only the local within-rot heritability esti- arrangements are pooled. In these studies with wild mates, but also the contribution of drift to genetic var- flies, the phenotypic variance increases between 5 and iation. 9 times that in laboratory studies, so the probability of detecting significant differences among karyotypes is diminished accordingly. Acknowledgements If the selective differences among chromosomal Theauthors wish to thank Lic. Rosa Liascovich for cri- inversions of D. buzzatii were mediated by their effects tical reading of the manuscript, Lic. Beatriz Gonzalez on body size, one would be able to predict the expected for her advise on the statisitical processing of data, and correlations among karyotype, phenotype and fitness Drs Mauro Santos, Aifredo Ruiz and Michael Rose for components. Ruiz & Santos (1989) suggest that the helpful discussions and constructive criticisms of effect of inversions on body size is responsible for the earlier versions of this manuscript. Dr Francesc Pens differences in male mating success. Applying this line and Mr Antonio Barbadilla are thanked for critical of thinking one would expect that flies with similar reading of the manuscript and constructive discussions body size (e.g. 2j and 2jz3 carriers) should share a on the statistical treatment of the data. The authors are common pattern of fitness components. This cannot be also indebted to Dr F. Pens for his help in the data pro- adequately tested for mating success because sample cessing with the BMDP package. This work is the sizes are too small for some karyotypic classes in Ruiz result of a co-operative project between Argentina and & Santos (1989) and the method of fitness compo- Spain. It was supported in Argentina by a CONICET nent analysis used by Ruiz et al. ,(1986)and Hasson et grant PIA 004-0422/87 and by Universidad de a!., (1991) is not sensitive enough. Nonetheless, the Buenos Aires grant EX038/87 awarded to 0. A. Reig analysis of fitness components by Hasson et al. (1991), and in Spain by a DGICYT grant PB 850071 awarded on the same population studied here, seems to indicate to A. Fontdevila. that genetically small individuals (2st) would be fitter at early stages and disadvantageous at later stages. In References contrast, the same study by Hasson et a!. (1991) shows that for the 2jz3 arrangement, the antagonistic ANDERSON, W.W. LEVINE,L, 0LvERA,0, POWELL. J. R, DE LA ROSA. M. behaviour is verified between two late fitness compo- E,SALCEDA, v.M, GASO, M.I.ANDGUZMAN, .i.1979. Evidence nents: longevity and fecundity, favouring the former for selection by male mating success in natural popula- and impairing the latter. It is clear from this that in tions of Drosophila pseudoobscura. Proc. Nat!. Acad. Sci., complex chromosomal polymorphisms there is no USA 76, 1519—1523. A. AND NAVEIRA, .i. The estimationof simple relationship between body size karyotype and BARBADILLA, 1988. parentalgenotypes by the analysis of a fixed number of fitness components. their offspring. 119,465—472. The evidence reported here and in previous works BARKER, J. S. F. 1982. Population genetics of Opun.tia breeding (Ruiz et a!., 1986 Hasson et a!., 1991) suggests that a Drosophila in Australia. In: Barker, 1. S. F. and Starmer, W. type of endocycic or balancing selection, although not T.(eds). EcologicalGenetics and Evolution. Academic a simple one, can be invoked to explain the main- Press, Australia, pp. 209—224. tenance of body size variation in nature. Nonetheless, BMDP.1988. Statistical Software. University of California the validity of this assertion has been challenged in Press,Berkeley, CA. view of the low heritability estimates for body size BUTLIN, R.K, READ, 1. L. AND DAY, T.H.1982. The effectsofa 562 E.HASSON ETAL.

chromosomal inversion on adult size and male mating pseudoobscura under field conditions. Anim. Behav. 35, success in the seaweed , Coelopa frigida. Heredity 49, 468—476. 51—62. PFRIEM, i'. 1983. Latitudinal variation in wing size in COLOMBO, i'. 1989. Chromosome polymorphisms affecting Drosophila subobscura and its dependence on polygenes recombination and exophenotypic traits in Leptysma in chromosome 0. Genetica 61,221—232. argentina (Orthoptera): A populational survey. Heredity PREVOSTI, A. 1955. Geographical variability in quantitative 62,289—299. traits in populations of Drosophila subobscura. Cold. Spr, COYNE,J. A. AND BEECHAM, E. 1987. Heritability of two Harb. Symp. Quant. Biol. XX, 294—299. morphological characters within and among natural PREVOSTI, A. 1960. Cambios en la heterocigosis por inversion populations of Drosophila melanogaster. Genetics 117, cromosómica a! variar por selección la longitud del ala en 727—737. Drosphila subobscura. Genet. Iber. 12, 27—4 1. DAVID, J. 1962. A new medium for rearing Drosophila in PREVOSTI, A. 1967. Inversion heterozygosity and selection for axenic conditions. Dros. Inf. Serv. 36, 128. winglength in Drosophila subobscura. Genet. Res. 10, DRUGER, M. 1962, Selection and body size in Drosophila 8 1—93. pseudoobscura at different temperatures. Genetics 47, PROUT, T. AND BARKER, J. S. F. 1989. Ecological aspects of the 209—222. heritability of body size in Drosophila buzzatii. Genetics ENDLER, J. A. 1986. Natural Selection in the Wild. Princeton 123, 803—8 13. University Press, Princeton, N J. ROBERTSON, F. W. 1957a. Studies in quantitative inheritance. X. FANARA, .i. i. 1988. Parámetros de vida en Drosophila buzzatii. Genetic variation of ovary size in Drosophila. J. Genet. 55, Seminario de Licenciatura. Facultad de Ciencias Exactas y 4 10—427. Naturales, Universidad de Buenos Aires. ROBERTSON, F. W. 1 957b. Studies in quantitative inheritance. FONTDEVILA, A. 1981. j,Cuánto polimorfismo cromosómico se XI. Genetic and environmental correlation between body pierde en Ia colonización? In: Reig, 0. A. (ed.), Ecology size and egg production in Drosophila melanogaster. J. and Genetics of Speciation. Equinoccio, Caracas, Genet. 55, 428—443. Venezuela, pp. 69—99. RUIz, A., FONTOEvILA, A., SANTOS, M., SEOANE, M. AND TORROJA, F. FONTDEVILA, A. 1989. Founder effects in colonising popula- 1986. The evolutionary history of Drosophila buzzatii. tions: The case of Drosophila buzzatii. In: Evolutionary VIII. Evidence for endocyclic selection acting on the Biology of Transient Unstable Populations. Springer- inversion polymorphism in a natural population. Evolu- Verlag, Berlin, pp. 74—95. tion 40,740-755. FONTDEVILA, A., RUIZ, A., ALONSO, G. AND OCANA, j.1982.The RUIZ, A., NAVEJRA, H. AND FONTDEV!LA, A. 1985. La historia evolutionary history of Drosophila buzzatii. II How much evolutiva de Drosophila buzzatii.IV. Aspectos has chromosomal polymorphism changed in colonization? citogenéticos de su polimorfismo cromosómico. Genét. Evolution 36,843—851. Ibér. 36, 13—35. FONTDEVILA, A., RUIZ, A., OCANA, J. AND ALONSO, G. 1981. The RuIZ, A. AND sANTO5, M. 1989. Mating probability, body size evolutionary history of Drosophila buzzatii. I. Natural and inversion polymorphism in a colonizing population of chromosomal polymorphism in colonized populations of Drosophila buzzatii. In: Fontdevila, A. (ed.), Evolutionary the Old World. Evolution, 35, 148—157. Biology of Transient Unstable Populations. Springer- FIASSON, E., VILARDI, J. C., NAVEIRA, H., FANARA, J. I., RODRIGUEZ, C., Verlag, Berlin, pp. 96—113. REIG, 0. A. AND FONTDEVILA, A. 1991. The evolutionary RUIZ, A., SANTOS, M., BARBADILLA, A., QUEZADA-DIAZ, J. E., HASSON, history of Drosophila buzzatii. XX. Fitness component E. AND FONTDEVILA, A. 1991. The evolutionary history of analysis in a natural endemic population from Argentina. Drosophila buzzatii. XVIII. Genetic variance for body size J. Evol. Biol. 4, 209—225. in a natural population of Drosophila buzzatii. Genetics, JOHN, B. 1983. The role of chromosome change in the evolu- 128, 739—750. tion of orthopteroid . In: Sharma, A. K. and SALCEDA, V. M. AND ANDERSON, w. w. 1988. Rare male mating Sharma, A. M. (eds), Chromosomes in the Evolution of advantage in a natural population of Drosophila Eukaryotic Groups, Vol. 1, CRC Press, FL, pp. 1-114. pseudoobscura. Proc. NatI. Acad. Sci.,U.S.A. 85, KIUMBAs, C. 1967. The genetics of Drosophila subobscura 9870—9874. populations. III. Inversion polymorphism and climatic SANTOS, M., RUIZ, A., BARBADILLA, A., QUEZADA-DIAZ, J. E., HASSON, factors. Mol. Gen. Genet. 99, 133—150. H. AND FONTDEVILA, A. 1988. The evolutionary history of LANDE, R. 1979. Effective deme sizes during long-term evolu- Drosophila buzzatii. XIV. Larger flies mate more often in tion estimated from chromosomal rearrangements. Evolu- nature. Heredity6l, 255—262. tion33, 324—251. SANTOS, M., RUtZ, A. AND FONTDEVILA, A. 1989. The evolutionary MATHER, K. AND JINKS, J. L 1982. Biometrical Genetics. history of Drosophila buzzatii. XIII. Random differentia- Chapman and Hall, . tion as a partial explanation of chromosomal variation in a MISRA, R. K. AND REEVE, E. C. R. 1964. Clines in body size structured natural population. Am. Natur. 133, 183—197. dimension in populations of Drosophila subobscura. SOKAL, R. R. AND ROHLF, F. J. 1981. Biometry. 2nd ed. W. H.. Genet. Res. 5, 240—256. Freeman, New York. PARTRIDGE, L., HOFFMAN, A. AND JONES, J. s. 1987. Male size and SPERLICH, D. AND PFRIEM, p•1986.Chromosomal poly- mating success in Drosophila melanogaster and Drosophila morphisms in natural and experimental populations. In: CHROMOSOMAL INVERSION EFFECTS ON DROSOPHILA SIZE 563

Carson, H. L., Ashburner, M. and Thompson, J. N. (eds.), and resistance of dissecation in Drosophila pseudoob- Genetics and Biology of Drosophila. Vol. 3e, Academic scura, Can. J. Genet. Cptol. 13,63—69. Press, London, pp. 257—309. WHITE, M. J. 0. AND ANDREW, L. F. 1960. Cytogenetics of the TAYLOR, C. E. AND KEKII, B. 1988. Sexual selection in a natural grasshopper Moraba scurra. V. Biometric effects of population of Drosophila melanogaster. Evolution 42, chromosomal inversions. Evolution 14, 284—292. 197—199. WHITE, M. J. D., LEWONTIN, R. C. AND ANDREW, L. E. 1963. Cyto- TAYLOR, C. E., PEREDA, A. Ii AND FERRARI, J. A. 1987. On the genetics of the grasshopper Moraba scurra. VII. Geo- correlation between mating success and offspring quality graphic variation of adaptive properties of inversions. in Drosophila melanogaster. Am. Natur. 129,721—729. Evolution 17, 147—162, THOMSON, i. A. 1971. Association of karyotype, body weight