3.1 | Genome quality assessments and demographic history
We sequenced 28 caribou genomes from across Canada and two caribou from Greenland to high coverage (35.57 – 43.03X; Table S1) representing all four subspecies and six Canadian Designatable Units (DUs; Figure 1 and Table S1), and used an additional reindeer genome from a domesticated animal in Inner Mongolia, sequenced by Li et al., (2019). Our caribou genomes showed high quality and recovery of BUSCO genes in the assembly, ranging from 92.7% to 93.1% of the more than 4,000 conserved genes surveyed (Table S1), and missing data per individual was low at 0.3%-1.0% with the exception of the previously published reindeer genome at 16.0% (Table S1).
A reconstruction of Rangifer demography over time using PSMC indicated a major population size expansion starting approximately 100-200 kya with peak population sizes around the glacial interstitial stage of a largely ice-free North America 120 kya (Figure 2a-b). This timing corresponds to a divergence of lineages largely concordant with the expansion and intra-specific diversification proposed by Banfield (1961). These differential population trajectories correspond to contemporary subspecies and ecotypes. The NAL and the Greenland caribou have much lower population sizes than BEL caribou, including the boreal caribou from the Northwest Territories, with Peary caribou being intermediate to these groups (Figure 2a-b, Figure S6 for all plotted together) consistent with an earlier divergence (Klütsch, Manseau, Anderson, Sinkins, & Wilson, 2017). Population sizes for all caribou lineages declined during the Wisconsin glaciation which lasted between 75-11 kya (Figure 2a-b), with the exception of the reindeer which has a unique demographic trajectory likely as a result of domestication. Contemporary inbreeding estimates varied greatly between different individuals. For the North American caribou, they ranged from -0.009 to 0.311. They were highest for Greenland at 0.654, and the Inner Mongolia reindeer also had an elevated co-efficient at 0.177, again reflecting the origin of the latter as originating from a domesticated population (Li et al., 2019; Table S1).