Empirical comparison of single nucleotide polymorphisms and microsatellites for population and demographic analyses of bowhead whales

Interest in bowhead whale stock structure has been high due to the species' extreme historical depletion, differential rates of recovery, the potential effects of climate change, and the need to set appropriate quotas for aboriginal hunts. We present an analysis of 42 linked and unlinked single nucleotide polymorphisms (SNPs) among 3 bowhead whale stocks and within the Bering/Chukchi/Beaufort Seas (BCB) stock, and compare results with previously published results of mtDNA control region sequences and 22 microsatellites. We performed tests of population structure (F_ST, χ², STRUCTURE), population assignment, and estimates of effective population size (N_e), and evaluated different numbers of loci and samples to estimate the relative statistical power of SNPs and microsatellites. Results indicate that this number of SNPs provides similar power to microsatellites to detect low levels of differentiation (F_ST = 0.005-0.03) between bowhead populations with sample sizes of at least 20 per population. Neither marker performed well for Bayesian analysis of population structure (STRUCTURE) for the strata that had high diversity coupled with low differentiation. This example is valuable in cautioning against use of STRUCTURE to exclude demographic independence of relatively abundant populations. Microsatellites provided greater precision for estimates of N_e and for assignment tests. All 3 genetic marker types are consistent with the BCB stock being a single population. For microsatellites, differences were found between individuals born before 1949 and those born after 1979. SNPs are continuing to prove valuable as tools for understanding structure and demography of populations, and are likely to prove beneficial for long-term monitoring of bowhead whales.