Abstract
Understanding the spatial structure of genetic diversity provides insights into a populations' genetic status and enables assessment of its capacity to counteract the effects of genetic drift. Such knowledge is particularly scarce for the snow leopard, a conservation flagship species of Central Asia mountains. Focusing on a snow leopard population in the Qilian mountains of Gansu Province, China, we characterised the spatial genetic patterns by incorporating spatially explicit indices of diversity and multivariate analyses, based on different inertia levels of Principal Component Analysis (PCA). We compared two datasets differing in the number of loci and individuals. We found that genetic patterns were significantly spatially structured and were characterised by a broad geographical division coupled with a fine-scale cline of differentiation. Genetic admixture was detected in two adjoining core areas characterised by higher effective population size and allelic diversity, compared to peripheral localities. The power to detect significant spatial relationships depended primarily on the number of loci, and secondarily on the number of PCA axes. Spatial patterns and indices of diversity highlighted the cryptic structure of snow leopard genetic diversity, likely driven by its ability to disperse over large distances. In combination, the species' low allelic richness and large dispersal ability result in weak genetic differentiation related to major geographical features and isolation by distance. This study illustrates how cryptic genetic patterns can be investigated and analysed at a fine spatial scale, providing insights into the spatially variable isolation effects of both geographic distance and landscape resistance.