Cooperative Fish and Wildlife Research Units Program: Pennsylvania
Education, Research and Technical Assistance for Managing Our Natural Resources

Chronic wasting disease (CWD) can be spread by direct and indirect transmission with direct transmission easier to predict and assess using landscape genetics. Identifying subpopulations of a wild large herbivore allows identification of focal areas to target for effective intervention. Our aim was to assess population structure of white-tailed deer (Odocoileus virginianus) in the northeastern United States at a regional scale to inform managers of gene flow throughout the region. We used ten microsatellites to genotype 5,701 samples collected from wild deer in Maryland, New York, Ohio, Pennsylvania, and Virginia from 2014 to 2022. We conducted a spatial principal component analysis to assess distribution of genetic variability and assessed population genetic structure using two methods: non-spatial Bayesian clustering algorithm (BCA) and spatial BCA. We used results of these clustering methods to create simulated populations of 500 individuals representing each cluster and also created simulated populations representing a captive deer population, among other anthropogenically-derived spatial units to conduct genetic assignment tests. We conducted genetic assignment tests using these simulated reference populations as potential sources, calculating the probability of samples being assigned to their corresponding genetic cluster, state, or physiographic province. Non-spatial BCA identified widespread gene flow with only 2 clusters across the region, while spatial BCA suggested further partitioning into a maximum of nine clusters. Our assignment tests were able to correctly assign deer to captive or wild origin in most cases (94%), as reported in a previous study, but performance varied when trying to assign wild deer to spatial units. Assignments to genetic clusters performed well where the potential sources represented clusters inferred with non-spatial BCA, but efficiency was greatly reduced when assigning samples to clusters inferred via spatial BCA. This indicates that differences between spatial BCA clusters are not strong enough to make assignment tests a reliable method for unequivocally inferring geographic origin of deer using 10 microsatellites. However, genetic distinction between clusters is still important to consider when managing for disease. Future research on what landscape barriers may be leading to these divisions and species-specific single-nucleotide polymorphisms will further our understanding of potential subpopulations in the region.