Genetic variation and effective population size in isolated populations of coastal cutthroat trout

Following glacial recession in southeast Alaska, waterfalls created by isostatic rebound have isolated numerous replicate populations of coastal cutthroat trout (Oncorhynchus clarkii clarkii) in short coastal streams. These replicate isolated populations offer an unusual opportunity to examine factors associated with the maintenance of genetic diversity. We used eight microsatellites to examine genetic variation within and differentiation among 12 population pairs sampled from above and below these natural migration barriers. Geological evidence indicated that the above-barrier populations have been isolated for 8,000–12,500 years. Genetic differentiation among below-barrier populations (F _ST = 0.10, 95% C.I. 0.08–0.12) was similar to a previous study of more southern populations of this species. Above-barrier populations were highly differentiated from adjacent below-barrier populations (mean pairwise F _ST = 0.28; SD 0.18) and multiple lines of evidence were consistent with asymmetric downstream gene flow that varied among streams. Each above-barrier population had reduced within-population genetic variation when compared to the adjacent below-barrier population. Within-population genetic diversity was significantly correlated with the amount of available habitat in above-barrier sites. Increased genetic differentiation of above-barrier populations with lower genetic diversity suggests that genetic drift has been the primary cause of genetic divergence. Long-term estimates of N _e based on loss of heterozygosity over the time since isolation were large (3,170; range 1,077–7,606) and established an upper limit for N _e if drift were the only evolutionary process responsible for loss of genetic diversity. However, it is likely that a combination of mutation, selection, and gene flow have also contributed to the genetic diversity of above-barrier populations. Contemporary above-barrier N _e estimates were much smaller than long-term N _e estimates, not correlated with within-population genetic diversity, and not consistent with the amount of genetic variation retained, given the approximate 10,000-year period of isolation. The populations isolated by waterfalls in this study that occur in larger stream networks have retained substantial genetic variation, which suggests that the amount of habitat in headwater streams is an important consideration for maintaining the evolutionary potential of isolated populations.