Spatial predictors of livestock depredation by leopards in a human-use landscape of Kumaon Himalayas, India

Livestock depredation is widely recognised as a primary proximate driver of perceived human-large carnivore conflict, posing challenges for both carnivore conservation and rural livelihoods. In this study, we examine the spatial patterns and ecological correlates of reported livestock depredation by leopards amid persistent conflict incidents in the lesser Himalayan region. The objectives of our study were to determine the ecological predictors influencing livestock depredation and map the spatial distribution of predicted zones of conflict classified as very high - low risk across the study area. Herein, we developed a predictive depredation risk modelling approach by utilizing 94 verified livestock depredation events, recorded through forest department compensation records in Ranikhet block. Employing generalized linear models that integrate anthropogenic, habitat, and environmental covariates, we identified proximity to human settlements as the strongest predictor of reported depredation occurrence (β = −0.87, p < 0.001). Central and southern regions, characterized by clustered towns, villages, and rangelands, emerge as high-risk hotspots to moderate risk zones. Conversely, intact forest patches, including the Chaubatia forest patch have been predicted to have minimal risk. As our analysis did not incorporate independent data on wild prey availability or leopard density, inferences are restricted to spatial patterns of depredation occurrence rather than leopard habitat preference. Our spatially explicit risk map, along with its analytical procedures and visual outputs, offers a replicable framework to identify potential human-carnivore conflict hotspots and equips conservation managers to prioritize targeted interventions within heterogeneous human-use landscapes.