Where leopards die: identifying mortality hotspots in northern Pakistan

Human–wildlife conflict poses a major threat to large carnivores worldwide, particularly in human-dominated landscapes where their persistence increasingly depends on effective coexistence strategies. This study employs geostatistical and time-series analyses to identify conflict hotspots, temporal trends, and key anthropogenic drivers of human-induced leopard (Panthera pardus) mortality in Azad Jammu and Kashmir, Pakistan. We applied kernel density estimation, hotspot analysis, and multiscale geographically weighted regression to assess spatial patterns of mortality, identify underlying drivers, and examine temporal trends. Based on 178 verified leopard mortality records collected between 2000 and 2023, retaliatory killings following livestock depredation emerged as the primary cause of death, accounting for 63% of cases, with mortality peaking during autumn and winter. Spatial analyses revealed a high concentration of leopard mortalities in areas adjacent to and outside protected area boundaries. Furthermore, distance to protected areas and proximity to human settlements significantly influenced mortality risk, operating across variable spatial scales. Our findings indicate that leopard killings intensify where fear, misinformation, and economic losses intersect. These results emphasize the importance of landscape-level conservation planning, improved conflict mitigation strategies, and the integration of human-dominated areas into carnivore conservation frameworks. This study provides spatially explicit evidence to support targeted management interventions and policy actions for leopard conservation in Azad Jammu and Kashmir.