Abstract
Detecting changes in wildlife populations across broad spatial extents and long time periods requires standardized monitoring approaches that produce consistent, comparable estimates. Indirect population density estimates based on animal field signs may be particularly useful, as they can be implemented over large areas with relatively low field effort and minimal equipment. Here, we aimed to estimate the distribution of sika deer population density across mainland Japan by using fecal pellet-group surveys. We compiled empirical data from fecal pellet decomposition experiments conducted across Japan and developed a hazard model of fecal pellet-group decomposition as a function of climatic conditions by using a meta-regression approach. We then quantified geographical and seasonal variation in decomposition rates and applied the model to nationwide pellet-group survey data. The model revealed substantial variation in fecal pellet decomposition rates according to elapsed time since defecation, climatic conditions, habitat types, and proximity to the coast. Depending on the season, regional differences in the number of pellet groups expected to accumulate under identical deer density conditions ranged from approximately 200-fold to more than 500-fold. Consequently, estimated pellet-group densities were highly sensitive to local environmental conditions, suggesting that the model assumption of constant decomposition rates may not always be appropriate when population density is compared across large spatial extents or seasons. By explicitly incorporating spatial and temporal variation in pellet decomposition, our method improves the reliability of deer population estimates derived from pellet-group surveys. The approach provides a cost-effective and scalable tool for nationwide wildlife monitoring and supports biodiversity assessment.

This publication has 36 references indexed in Scilit: