Isolation-by-distance (IBD) is a natural pattern not readily incorporated into theoretical models nor traditional metrics for differentiating populations, although clinal genetic differentiation can be characteristic of many wildlife species. Landscape features can also drive population structure additive to baseline IBD resulting in differentiation through isolation-by-resistance (IBR). We assessed the population genetic structure of boreal caribou across western Canada using non-spatial (STRUCTURE) and spatial (MEMGENE) clustering methods and investigated the relative contribution of IBD and IBR on genetic variation of 1221 boreal caribou multilocus genotypes across western Canada. We further introduced a novel approach to compare the partitioning of individuals into management units (MU) and assessed levels of genetic connectivity under different MU scenarios. STRUCTURE delineated five genetic clusters while MEMGENE identified finer-scale differentiation across the study area. IBD was significant and did not differ for males and females both across and among detected genetic clusters. MEMGENE landscape analysis further quantified the proportion of genetic variation contributed by IBD and IBR patterns, allowing for the relative importance of spatial drivers, including roads, water bodies and wildfires, to be assessed and incorporated into the characterization of population structure for the delineation of MUs. Local population units, as currently delineated in the boreal caribou recovery strategy, do not capture the genetic variation and connectivity of the ecotype across the study area. Here, we provide the tools to assess fine-scale spatial patterns of genetic variation, partition drivers of genetic variation and evaluate the best management options for maintaining genetic connectivity. Our approach is highly relevant to vagile wildlife species that are of management and conservation concern and demonstrate varying degrees of IBD and IBR with clinal spatial genetic structure that challenges the delineation of discrete population boundaries.
Usage Notes:
Priadka_etal_2018_genotypes
Genotypes for 1221 boreal caribou at 9 microsatellite loci. Missing data is indicated by "-99". Coordinates for sampling locations have not been included because boreal caribou are listed as Threatened under Canada's Species At Risk Act.
Priadka_etal_2018_Landscape_rasters
Raster .asc files of landscape resistance models. Univariate models include resistance costs of 10, 50 and 100. Optimized models for Cluster 1 and Cluster 2 reflect the optimized resistance costs for roads and water bodies (land1), roads and forestry (land2), water bodies and forestry (land3) and roads, water bodies and forestry (land4). Raster cell size is 250m in planar projection.