| Pacific rockfishes (genus Sebastes) have attracted wide scientific and public interest from an evolutionary, fishery and conservation standpoints. This dissertation addresses several hypotheses involving spatial and temporal scales of genetic change in two overfished rockfishes, darkblotched (S. crameri) and canary rockfish (S. pinniger), using statistical analyses of genetic variation within microsatellite DNA. First, I isolate and characterize 9 polymorphic microsatellites from canary rockfish useful in at least 13 congeneric species. Second, I investigate spatial differentiation in darkblotched rockfish, and how pooling samples with few individuals affects estimates of genetic parameters. Results indicate (i) significant correlation between pairwise genetic and geographic distances (isolation-by-distance) among samples, implying that scales of larval dispersal are short, and (ii) using an absolute genetic distance to pool samples increases overall genetic differentiation and resolves deviations from Hardy-Weinberg equilibrium without creating data artifacts. Third, I report the variance effective population size ( Ne) for darkblotched rockfish by measuring allele frequency shifts among six abundant cohorts. I obtained an estimate differing by three to four orders with the current census population size, which is most likely explained by high variance in reproductive success among individuals, genetic structure, and demographic perturbations such as historical fishing. This finding stresses the need to clearly defined conservation goals for rockfishes and marine populations in general. Fourth, I analyze contemporary samples and archived otolith samples for the heavily exploited canary rockfish to test for long-term temporal genetic changes. Results suggest that temporal differences over years clearly surpassed spatial differences among contemporary and archived sampling locations, highlighting important effects of genetic drift on this species. Lastly, I employed simulations, spatial data and Ne estimates to rapidly appraise which rockfishes within the Sebastes complex are more or less critical to be managed as spatially separated management units based on the relationship between overall genetic distance (FST) and population assignment (PA) of individuals to their site of origin. Using simulations and some arbitrary criteria, I conclude that it is unrealistic to obtain meaningful PA when FST 0.01 should be more susceptible to lose genetic diversity if local populations are extirpated. |
