Predicted Risk of Negative Human-Black Bear Interactions in Maryland under Neutral and Drought Conditions

Reports of negative interactions between humans and bears in Maryland have increased steadily since the 1990s, requiring substantial investments of time and resources for their management. While managers typically possess a strong understanding of the spatial distribution of current and historical interactions, they lack the quantitative tools necessary to predict where future interactions are most likely to occur under changing climate conditions. To address this gap, we created 50-m resolution maps that illustrate the risk of negative human-bear interactions across Maryland under two drought scenarios: no drought (Standardized Precipitation Evapotranspiration Index [SPEI]: 0) and severe summer drought (SPEI: -2). Maps were informed by an advanced machine learning model, specifically eXtreme Gradient Boosting (XGBoost), which was developed by analyzing the relationships between 15 anthropogenic and environmental features and records of negative human-bear interactions in Maryland from 2020 through 2023. Our maps visualize the substantially heightened predicted risk of negative interactions during and following summer drought conditions. Specifically, during neutral conditions, 10% of Maryland displayed a predicted risk ≥0.5, indicating a moderate to high probability of negative human-bear interactions. This percentage increased to 16% during and following summer drought conditions, suggesting that climate-driven events can exacerbate negative human-bear interactions. These predictive maps offer valuable tools for proactive management and mitigation, enabling managers to anticipate and address high-risk areas before they become problematic.