Population ecology of the leopard cat (Prionailurus bengalensis) in three commercial forest reserves in Sabah, Malaysia

The leopard cat (Prionailurus bengalensis) is the most common felid species found in Borneo and Southeast Asia. Although often regarded as adapted well to living in disturbed habitat and frequently sighted in secondary forests and plantations, yet little is known about this species in the wild. By using remote digital camera-traps, the present study was conducted on the leopard cat populations located at three commercial forest reserves in South-central of Sabah, Malaysia. The forest reserves (FR), namely Deramakot FR, Tangkulap FR and Segaliud Lokan FR, have been managed under three different logging approaches in the past. The objectives of the study were to investigate the ecology of the leopard cat with specific emphasis on activity pattern, home range and habitat preferences. This study also provides the first densities estimate of leopard cat based on large camera trapping dataset. At Deramakot FR, 23 different individuals were identified during a 4-month trapping period (47 camera stations, 1916 trap-days). At Tangkulap FR, 41 different individuals were identified (64 camera station, 2302 trap-days) and at Segaliud Lokan FR, 60 individuals were identified (SS camera station, 2933 trap-days) during the surveys. At all study sites, male leopard cats were photographed more frequent than females. Males also tended to have larger home range and moved further distance than female. Minimum observed home range estimate for individuals with > 3 captures at different camera locations in Deramakot FR ranged from 1. 9 to 12.3 km2 for five males and 0.4 to 7 .5 km2 for five females. In Tangkulap FR, minimum observed home range estimate for nine males ranged from 1.2 to 26.4 km2 • Whereas in Segaliud Lokan FR, minimum observed home range for 12 males ranged from 0.8 to 11.4 km2 and 0.3 to 1. 9 km2 for three females. Leopard cats were captured more frequently along logging roads than on the forest trails and at night than during the day, confirming the behaviour that is regularly observed for a nocturnal species. The spatially explicit capture-recapture (SECR) models estimated a higher leopard cat density in the two more disturbed forest reserves; 16.50 ± 2.00 individuals 100 km-2 in Segaliud Lokan FR and 12.40 ± 1.61 individuals 100 km-2 in Tangkulap FR, than in the sustainably managed forest - 9.56 ± 1.63 individuals 100 km-2 in Deramakot FR. The non-spatial model had probably overestimated the leopard cat density because it did not take into account the animal movements and the results were highly influenced by the choice of buffer width (half MMDM or full MMDM). Therefore, the density estimates using SECR models are more reliable and they can be used as baseline data for the leopard cat density in production forests. Using occupancy-based approach, the variables which influenced the occurrence of leopard cat were identified. No single models emerged as the best model but canopy closure, the ratio of climax to pioneer trees and the ratio of large to small trees had a significantly negative impact on the leopard cat occurrences. These results confirm that leopard cat is able to survive in modified landscape and it even appeared to benefit from the opening of forests following logging activities but for other species, the effect might be different