Predicting harvest vulnerability for a recovering population of American black bears in western Maryland

Recreational hunting is the tool most commonly used to manage American black bear (Ursus americanus) populations in North America. However, bear populations can be sensitive to overharvest, particularly of mature females that can directly affect population growth. Managers need a thorough understanding of the factors affecting harvest vulnerability when using hunting as a primary management strategy. Here, we coupled Global Positioning System spatial data from female black bears and human hunters in western Maryland, USA, from 2005 to 2007, in order to model bear harvest vulnerability. Specifically, we developed maximum entropy (Maxent) predictive occurrence models for bears and for bear hunters and evaluated the influence of 7 environmental variables on their distributions. We then assessed predicted distribution maps for probability of co-occurrence to identify areas of high and low harvest vulnerability. Slope and land ownership (i.e., private public) were the 2 most important variables determining female bear distributions, whereas land ownership and cover type were the most important variables influencing hunter distributions. We classified roughly 12% and 16% of the study area as being of high relative use for bears and bear hunters, respectively. Only 5.4% of the study area was considered to have high harvest vulnerability (i.e., high probability of co-occurrence). Areas with high bear relative use but low hunter use (i.e., low harvest vulnerability) comprised 0.9% of the study area. We were most interested in areas of high and low harvest vulnerability to enable resource managers to adjust hunting regulations that meet harvest goals.