SEXUAL SELECTION OPERATING IN A WILD POPULATION OF DROSOPHILA ROBUSTA.

Abstract

Almost 60 years have elapsed since the discovery of variation in the polytene chromosomes of Drosophila (Sturtevant and Dobzhansky 1936). Inversion polymorphism studies have revealed some of the intricacies of evolutionary change, but even now, when interest in them has waned, there is little understanding of the particular alleles contained within them that alone or in relation to other genes influence any component of fitness. Perhaps the most important rule about inversion polymorphisms learned from a long history of laboratory experiments (cf. Lewontin et al. 1981) is their sensitivity to environmental features such as temperature and nutrition. Therefore, laboratory studies alone may not predict the consequences due to genotype (karyotype) by environment interactions of selection in wild populations. Only a few studies have documented selective mechanisms associated with preserving inversion polymorphisms in natural populations beyond establishing correlations with temperature, rainfall, humidity, elevation, and other environmental features. Despite the long history of interest in the ecology of Drosophila breeding and feeding sites (cf. Carson 1951, 1971; Heed 1957, 1968, 1971), experimental analyses of inversion polymorphisms involving natural breeding substrates are few because the mating sites and larval ecology of most species with well described inversion polymorphisms are poorly known, e.g., D. pseudoobscura. An early exception to this problem was the comprehensive study of inversion polymorphism in D. flavopilosa, a flower breeder from South America (Brncic 1962). Egg to adult viability differences among karyotypes were observed among preadult stages in flowers returned to the lab and exposed to contrasting temperatures (Brncic 1968). More recently, analysis of the sexratio meiotic drive system in natural populations of D. pseudoobscura suggested that most selection responsible for maintaining this polymorphism must be operating in preadult stages (Beckenbach 1996). A notable non-Drosophila example is the chromosome I polymorphism in seaweed flies, Coelopa frigida, that carry out their life cycle on decomposing seaweed on beaches around the north Atlantic and North Sea. Gene arrangement-based differences in egg to adult development time, adult body size, and mate preference have been demonstrated in seaweed-reared flies (Day et al. 1983; Engelhard et al. 1989; Gilburn et al. 1992). For species that can be collected in nature throughout their life cycle, karyotype frequencies can be estimated from life stages reared in the wild and comparisons can be made between stages of the life cycle in selection component analyses (e.g., Prout 1971; Bundgaard and Christiansen 1972). Inversion polymorphisms in cactophilic D. buzzatii have been studied this way, a species that feeds and breeds in the fermenting tissues of several species of prickly pear cactus. In general, cactophilic Drosophila have taught us much about maintenance of genetic variability, host specificity, chemical ecology, trophic adaptation, and life history evolution (Barker and Starmer 1982; Barker et al. 1990). Despite the fact that 48% (44/91) of all described and undescribed D. repleta group species possess inversion polymorphism (Wasserman 1992), in only a few species has there been concerted effort to unravel the causes for maintenance of inversion polymorphism in nature, that is, D. buzzatii and D. mojavensis. Inversion polymorphism in the latter species, along with D. pachea, a cactophilic member of the D. nannoptera group, has been recently summarized (Etges et al., in press). Ruiz et al. (1986) demonstrated endocyclic selection, directional selection in opposing directions during the life cycle, on gene arrangements of D. buzzatii. However, this pattern is population and host specific (Hasson et al. 1991) and not surprisingly is limited to a few components of fitness including egg to third instar viability and adult body size, but not mating propensity (Santos et al. 1992; Barbadilla et al. 1994). Adult size in Drosophila can be correlated with increased fitness as larger flies typically have greater reproductive potential (Robertson 1957) and dispersal ability (Mangan 1982). These studies of D. buzzatii deserve special note because they illustrate the complexity of factors that must be addressed in understanding the maintenance of inversion polymorphisms in natural populations. Thus, studies of selection on inversion polymorphisms in species with hard-to-find breeding sites have been restricted to adult flies because they can be captured in the wild. If we include the recent D. buzzatii studies mentioned above, several studies of genetic variation in male mating success (Anderson et al. 1979; Levine et al. 1980; Salceda and Anderson 1988), the recent analysis of the sex-ratio polymorphism in D. pseudoobscura (Beckenbach 1996), and mate choice in seaweed flies (Gilburn et al. 1992), most of what we know of selection operating on inversion polymorphisms under natural conditions has only recently been documented. It is interesting to note that not one example documenting an identifiable selective agent operating on inversion polymorphisms was cited by Endler (1986) in his review, Natural Selection in the Wild, despite the long history of population genetics