Evidence for selection following perturbation of allozyme frequencies in a natural population of Drosophila

Abstract

There is now ample evidence for the ubiquity of genetic variation at allozyme loci in many species^1,2, but the extent to which this variation is actively maintained by natural selection remains unresolved. Although various approaches have suggested the importance of selection at some allozyme loci in laboratory populations^3–6, to be fully convincing, evidence for natural selection must finally come from studies on natural populations. Inevitably, this will entail an understanding of and the potential to manipulate both the ecology and genetics of the population. Because sufficient is known of the ecology, breeding site, nutritional requirements and field behaviour of at least some species of the cactiphilic Drosophila, we^7–9 and others¹⁰ have chosen them for experimental evaluation of the forces operating to maintain genetic variation. If genetic variation at some locus is being actively maintained by natural selection, the action of this selection should be detectable following artificial change in gene frequencies produced either by adding certain genotypes to the population, or by removing them from it. Jones and Parkin¹¹ have reported unsuccessful attempts to detect selection by such perturbation experiments in Cepaea and Drosophila pseudo-obscura, while Halkka et al.¹² inferred strong natural selection at a locus controlling colour polymorphism from exchange of individuals of Philaenus spumarius between two island populations. We report here results for the simultaneous perturbation of gene frequencies at three allozyme loci in Drosophila buzzatii. The changes in gene frequency following this perturbation provide strong evidence for selection, and for differential selection among the three loci.