Drift, selection and adaptive variation in small populations of a threatened rattlesnake.

An important goal of conservation genetics is to determine if the viability of small populations is reduced by a loss of adaptive variation due to genetic drift. Here, we assessed the impact of drift and selection on direct measures of adaptive variation (toxin loci encoding venom proteins) in the eastern massasauga rattlesnake (Sistrurus catenatus), a threatened reptile that exists in small isolated populations. We estimated levels of individual polymorphism in 46 toxin loci and 1,467 control loci across 12 populations of this species, and compared the results with patterns of selection on the same loci following speciation of S. catenatus and its closest relative, the western massasauga (S. tergeminus). Multiple lines of evidence suggest that both drift and selection have had observable impacts on standing adaptive variation. In support of drift effects, we found little evidence for selection on toxin variation within populations and a significant positive relationship between current levels of adaptive variation and long- and short-term estimates of effective population size. However, we also observed levels of directional selection on toxin loci among populations that are broadly similar to patterns predicted from interspecific selection analyses that pre-date the effects of recent drift, and that functional variation in these loci persists despite small short-term effective sizes. This suggests that much of the adaptive variation present in populations may represent an example of "drift debt," a nonequilibrium state where present-day levels of variation overestimate the amount of functional genetic diversity present in future populations.