Canadian polar bear population structure using genome-wide markers

Predicting the consequences of environmental changes, including human-mediated climate change on species, requires that we quantify range-wide patterns of genetic diversity and identify the ecological, environmental, and historical factors that have contributed to it. Here, we generate baseline data on polar bear population structure across most Canadian subpopulations (n=358) using 13,488 genome-wide single nucleotide polymorphisms (SNPs) identified with double digest restriction site-associated DNA sequencing (ddRAD). Our ddRAD dataset showed three genetic clusters in the sampled Canadian range, congruent with previous studies based on microsatellites across the same regions; however due to a lack of sampling in Norwegian Bay, we were unable to confirm the existence of a unique cluster in that subpopulation. These data on the genetic structure of polar bears using SNPs provide a detailed baseline against which future shifts in population structure can be assessed, and opportunities to develop new non-invasive tools for monitoring polar bears across their range.
Methods
ddRAD libraries were constructed following the Peterson et al. (2012) protocol, using MluCI and PstI restriction enzymes and size selecting the DNA insert to 275-365 bp using a BluePippin Prep. Libraries were sequenced using paired-end 125 bp Illumina HiSeq 2500. PCR duplicates were removed and reads were assebled to the polar bear reference genome UrsMar_1.0 using BWA mem. Variant detection and genotype calling were done using samtools mpileup and bcftools call. Loci were filtered using VCFtools.