| Evidence for balancing selection and local adaptation in the available nucleotide polymorphism data is scarce. We begin by showing that these forms of selection are unlikely to be observed in most regions of the Drosophila genome. One way around this problem is to investigate regions with reduced gene conversion, such as near the breakpoint junction of a chromosomal rearrangement. We describe the pattern of nucleotide variation in a population sample for a region surrounding a breakpoint junction of In( 2L)t, a common polymorphic inversion in Drosophila melanogaster. In(2L) t appears to be a young polymorphism recently elevated to high frequencies by the action of natural selection. A significant deficiency of haplotypes is found at the breakpoint site among standard chromosomes. The observations are consistent with the existence of epistatic interactions among sites that predate the appearance of In(2L )t. However, alternative selection models are also plausible, as are certain demographic models. We sample several geographically diverse populations of D. melanogaster and one population of a sister species, D. simulans, to distinguish between selective and demographic explanations for the data. Significant geographic differentiation is detected between populations of D. melanogaster . As in some (but not all) populations of D. melanogaster , most nucleotide variation in the D. simulans sample is distributed between two deeply diverged haplotype classes. Simple demographic scenarios do not appear to be a sufficient explanation for patterns of nucleotide variation in these two species. Finally, we describe a study of the available polymorphism data for D. melanogaster and D. simulans. A comparison of recombination rates estimated from polymorphism data with those from laboratory crosses suggest a genome-wide excess of linkage disequilibrium in both species. |
