Inferring the History of Speciation from Multilocus DNA Sequence Data: The Case of Drosophila pseudoobscura and Close Relatives

The divergence of Drosophila pseudoobscura from its close relatives, D. persimilis and D. pseudoobscura bogotana, was examined using the pattern of DNA sequence variation in a common set of 50 inbred lines at 11 loci from diverse locations in the genome. Drosophila pseudoobscura and D. persimilis show a marked excess of low-frequency variation across loci, consistent with a model of recent population expansion in both species. The different loci vary considerably, both in polymorphism levels and in the levels of polymorphisms that are shared by different species pairs. A major question we address is whether these patterns of shared variation are best explained by gene flow or by persistence since common ancestry. A new test of gene flow, based on patterns of linkage disequilibrium, is developed. The results from these, and other tests, support a model in which D. pseudoobscura and D. persimilis have exchanged genes at some loci. However, the pattern of variation suggests that most gene flow, although occurring after speciation began, was not recent. There is less evidence of gene flow between D. pseudoobscura and D. p. bogotana. The results are compared with recent work on the genomic locations of genes that contribute to reproductive isolation between D. pseudoobscura and D. persimilis. We show that there is a good correspondence between the genomic regions associated with reproductive isolation and the regions that show little or no evidence of gene flow.