Quantifying the similarity between genes and geography across Alaska's alpine small mammals

AimQuantitatively evaluate the similarity of genomic variation and geography in five different alpine small mammals in Alaska, and use this quantitative assessment of concordance as a framework for refining hypotheses about the processes structuring population genetic variation in either a species-specific or shared manner. LocationAlaska and adjacent north-western Canada. MethodsFor each taxon we generated 3500-7500 single-nucleotide polymorphisms and applied a Procrustes analysis to find an optimal transformation that maximizes the similarity between principal components analysis maps of genetic variation and geographical maps of sample locations. We generate stability maps using projected distributions from ecological niche models of the Last Glacial Maximum and the present. ResultsSignificant similarity between genes and geography exists across taxa. However, the extent to which geography is predictive of patterns of genetic variation not only differs among taxa, but the correspondence between genes and geography varies over space. Geographical areas where genetic structure aligns poorly with the geographical coordinates are of particular interest because they indicate regions where processes other than isolation by distance (IBD) have influenced genetic variation. The clustering of individuals according to their sample location does not support suppositions of admixture, despite the presumed high vagility of some species (e.g. arctic ground squirrels). Main conclusionsGenomic data indicate a more nuanced biogeographical history for the taxa than suggested by previous studies based on mtDNA alone. These include departures from IBD that are shared among taxa, which suggest some shared processes structuring genetic variation, including new potential ancestral source populations. In addition, some regions fit expectations of IBD where incremental migration and gene flow play a strong role in population structure, despite any ecological difference among taxa. Differences in dispersal capabilities do not result in different species-specific local patterns of population structure, at least at the sampling scale examined here. We highlight how the general fit to, as well as departures from, expectations for patterns of genetic variation based on the Procrustes analyses can be used to generate hypotheses about the underlying processes.