Estimating leopard cat Prionailurus bengalensis densities using photographic captures and recaptures

Abstract

Precise estimates of abundance and density are crucial for species conservation. For secretive felids, such as leopard cat Prionailurus bengalensis, acquiring such estimates based on conventional methods is difficult. We demonstrated the possibility of individual identification of leopard cats using coat patterns, and estimated their density using photographic capture‐recaptures in a small watershed (182 km2) of Khangchendzonga Biosphere Reserve, Sikkim, India. On comparing the different body parts, we found that hind‐quarter had the maximum usability (83.9%) for individual identification. The overall photo‐capture rate representing an index of leopard cat relative abundance was calculated as 3.7 ± 1.27 captures/100 trap days. We used both non‐spatial and spatially explicit capture‐ recapture (SECR) approaches to estimate leopard cat abundance and density. Our spatially explicit models estimated leopard cat density as 17 ± 5.33 (maximum‐likelihood based approach) and 17.52 ± 5.52 (Bayesian approach with Markov chain Monte Carlo simulations) individuals/100 km2, while in our non‐spatial model, density estimates varied from 18.01 to 22.25 individuals/100 km2. Camera trap results also indicated that the leopard cat used temperate and subtropical habitats to a large extent. Our study validated the applicability of camera trap based capture‐recapture techniques to estimate the density of leopard cat. Therefore, we recommend the use of this technique with appropriate site‐specific modifications for population estimation and monitoring of this species throughout its distribution range.