Résumé
ABSTRACT
Maximizing detection probability of elusive species enhances the robustness of population and occupancy estimates, which are essential for supporting impactful conservation strategies. Although the number of camera trap studies on leopards is increasing, few have assessed the drivers influencing leopard detection specifically. We used occupancy modeling embedded in a causal inference framework to provide four biologically robust site use models against which to test the influence of six factors likely influencing leopard detectability at the level of encounter probability, trigger probability, and image quality. In this study, vegetation productivity moderated by rainfall was the top predictor of leopard detectability associated with three of the four site use models. While optimizing detection probability improves estimates of population parameters, the cost‐effectiveness of the study designs is also an essential criterion to consider for long‐term monitoring of elusive species. Camera trap placement involves minimal cost in the early stages of the grid design. Our results suggest that setting cameras in microhabitats of moderate productivity improved leopard detectability in the wet season. This study can inform the design of camera trap studies occurring in semi‐arid bushland ecosystems to improve estimates of leopard population and occupancy.
Camera trapping studies are increasingly used to derive population and occupancy estimates of leopards; however, little is known about the drivers influencing leopard detection when using such devices. We used occupancy modeling embedded in a causal inference framework to identify factors influencing leopard detectability. Our results suggest that setting cameras in microhabitats of moderate productivity improved leopard detectability in the wet season.