A comparison of polymorphism of genetic markers and population sample sizes required for mixed-stock analysis of sockeye salmon (Oncorhynchus nerka) in British Columbia

Abstract

We evaluated two questions: (i) do microsatellites require larger population baseline sample sizes than single nucleotide polymorphisms (SNPs) to allow the accuracy provided by the microsatellites in genetic stock identification (GSI) applications to be expressed, and (ii) do less genetically distinct populations require larger population baseline sample sizes than more distinct populations to improve population-specific accuracy in GSI applications? Forty-six SNP loci were surveyed in 40 populations of sockeye salmon ( Oncorhynchus nerka ) over 16 regions from southern and central British Columbia and were split into two groups: the top 23 SNPs evaluated for stock identification for British Columbia sockeye salmon and the poorest 23 nuclear SNPs. Fourteen microsatellites were surveyed and split into two groups, with loci from the top 7 loci for stock identification accuracy assembled in one group, and the remaining 7 microsatellites assigned to a second group. SNPs and microsatellites with lower stock identification power required larger population sample sizes to allow expression of stock identification potential. To achieve the same level of population-specific accuracy, SNPs required fewer individuals to be sampled in a population than did microsatellites. Less genetically distinct populations required larger population sample sizes to achieve a given level of accuracy in estimated stock compositions.