DESATURASE‐2 , ENVIRONMENTAL ADAPTATION, AND SEXUAL ISOLATION IN DROSOPHILA MELANOGASTER

Abstract

In a previous paper (Coyne and Elwyn 2006), we repeated environmental stress experiments in Drosophila melanogaster that were originally conducted by Greenberg et al. (2003). In their study, Greenberg et al. used targeted gene replacement to construct lines having different alleles of the desaturase-2 (desat2) locus—a gene involved in synthesis of cuticular hydrocarbons. They found that different alleles had different effects on flies’ responses to cold and starvation stress: carriers of the African and Caribbean allele (ds2Z) consistently had less cold resistance and greater starvation resistance than carriers of the Cosmopolitan allele (ds2M). Greenberg et al. (2003) proposed that these differences are adaptive because one might expect African flies to encounter higher temperatures and less available food. Because there was earlier evidence for sexual isolation between carriers of different desat2 alleles (African females homozygous for ds2Z discriminate against non-African males homozygous for ds2M; Fang et al. 2002), possibly because different cuticular hydrocarbons act as mating cues, Greenberg et al. (2003) interpreted the sexual isolation between African and Cosmopolitan races as a byproduct of adaptation to different environments. Using the same lines described by Greenberg et al. (2003), but with a larger sample of transgenic constructs and of tested individuals, Coyne and Elwyn (2006) failed to find consistent ‘‘adaptive’’ differences between ds2Z and ds2M alleles in their tolerance to cold and starvation stress. Rather, the association was more or less random, with differences between the alleles sometimes nonsignificant and sometimes significant in directions opposite to those reported by Greenberg et al. (2003). We attributed our observations of ‘‘nonadaptive’’ differences in stress tolerance between different alleles, as well as the differences among replicate constructs of the same desat2 allele, to differences in the genetic backgrounds of transgenic constructs—differences that arose during their synthesis and that could affect stress tolerance independently of the desat2 genotype. We have no explanation for the discrepancy between the results of Coyne and Elwyn (2006) and those of Greenberg et al. (2003). Coyne and Elwyn also (2006) also reported mating experiments between carriers of various genetically engineered desat2 alleles and found that, although there was some evidence that desat2 may be involved in sexual isolation, this evidence was weak and was contradicted by other observations. After learning of Coyne and Elwyn’s new results, Green-