Spatial Analysis of Genetic Population Structure in Cougars

The cougar is a large predatory feline found widely in the Americas. To assess genetic differentiation in cougars, previous research used a Bayesian clustering model with genetic data and found evidence of genetic population structure in cougars, with suggestive but indistinct spatial boundaries between subpopulations. These patterns are intriguing, particularly because the model ignored spatial information when assessing genetic differentiation, which could arise from limited host interaction due to landscape barriers. Using Bayesian classification models of microsatellite loci data and associated spatial data, we investigate whether the composition of genetic population structure in cougars changes when considering available spatial information. Using a geographic information system, we also explored whether apparent genetic population structure boundaries were associated with landscape features. This paper compares the posterior genetic group membership from two Bayesian models, where one accommodates individual spatial locations and the other does not. There were two clearly differentiated cougar populations using the spatial model, in contrast to the two overlapping populations suggested by the aspatial model. The spatial correspondence of genetic subpopulation groups and certain landscape features, such as average annual precipitation and presence of major rivers and roads, was suggestive of possible landscape barriers to host movement, although the model could not differentiate between the several potential barriers. This study suggests that the use of spatial information and models adds insight when investigating genetic population structure.