Abstract: [Objective] Far Eastern Leopard (Panthera pardus orientalis) disperses in the northernmost region and maintains the minimum population of Panthera pardus subspecies. It is evaluated as Critical Endangered in IUCN Red List. In this study, we analyzed long-term dynamic of population size and distribution of Far Eastern Leopard in Wangqing Nature Reserve and expected to provide scientific data for the protection of this endangered species.[Method] From Autumn 2013 to Autumn 2017, cameras were set up with a density of 1 pair per 3 km×3 km grid in the main leopard habitat in Wangqing Nature Reserve. The relative abundance indices (RAI) was used to calculate leopard abundance, the distribution area was defined by 99% kernel density contour, and individual was identified in Extract Compare software. The closure test, model selection and population estimation were performed for the data of effective monitoring period with CAPTURE software. The effective trapping area was estimated by projecting an 8 km radius buffer around each camera trap location. Then we calculated population density using effective trapping area.[Result]The RAI of leopard varied from 0.34 to 2.12 during 9 monitoring seasons. The distribution area of leopard was 201 km² in Autumn 2013 and 992 km² in Autumn 2017. A total of 9 individuals were identified, including 3 females, 3 males, 2 cubs and 1 unrecognizable. Both the value and the power of CAPTURE closure test were low when population size was small. The model selection criterions of M_h (heterogeneity effects model) were highest in 4 monitoring seasons, and second only after M₀ (null model) in the other 5 seasons. The goodness of fit tests of M₀ vs M_b(behaviour effects model) and M₀ vs M_t(time effects model)showed that there were no significant differences between groups. The population density of leopard ranged from 0.12 to 0.88 individual per 100 km² during research period. The value of captured individual/estimated population size was relatively high in general, albeit there was a low value of 0.43 in Spring 2017.[Conclusion] The distribution pattern of leopard obviously showed spatiotemporal heterogeneity due to geographic variance and prey abundance fluctuation. CAPTURE closed population test was ineffective for small sample test. M_h model was proper to evaluate leopard population. The capture rates showed no variance after trap or in different trap occasions. Leopard population density would be overestimated due to trap-shy response.

Key words: Panthera pardus orientalis, camera trap, spatiotemporal patterns, population density evaluation, wild animal