Estimating Japanese serow abundance using a spatially explicit capture–recapture model

Abstract

Since the beginning of the 2000s, declining densities of the Japanese serow have been reported in many protected areas, and management based on reliable population estimates is needed. To contribute to the management of Japanese serows, we applied an open spatially explicit capture–recapture (open SECR) model, which can consider both process errors and observation errors using a hierarchical model, in Tsumagoi Village, Gunma Prefecture, Japan. We fitted an ordinal SECR model (model 1) and a sex-age (male, female, and juvenile) structured SECR model (model 2) to individual identification data from 25 camera traps installed between September 2017 and December 2019 to estimate the abundance of serows. Individuals were identified by natural marks such as horn length and angle, facial scars, ear loss or slit, facial pattern, subocular gland size, leg loss, and distinctive body color. As a result, the parameters in all models were converged and the SECR models successfully estimated the serow density (2.44 and 2.18 serow/km2 by models 1 and 2, respectively). The results of this study will contribute to future serow management because our method enables more reliable estimations with credible intervals compared to conventional methods.