Testing the reliability of a camera‐trap based abundance proxy for assessing mountain ungulate populations

Abstract

Monitoring wildlife abundance is important for understanding species population status and trends and, thus, for informing conservation planning and wildlife management. Yet, abundance monitoring can be challenging and costly to implement, particularly for wildlife of conservation concern occurring in remote and rugged terrain. Remotely‐sensed proxies, such as the relative abundance index (RAI) based on camera‐trap data, can be a useful and cost‐efficient approach in such situations. However, RAI is typically assumed to be linearly, continuously, and positively related to actual abundance—the assumptions that are rarely tested. In this study, we explored how well RAI from camera‐traps corresponds to direct field estimates of the abundance of the bezoar goat ( Capra aegagrus ) in the rugged mountains of the South Caucasus. Specifically, we summarized the data of RAI and count estimates obtained during the same one‐month periods (all in winter) and 5 × 5 km ² grid cells in 2018, 2019, and 2022, and used linear regression to determine the relationship between both measures. We found that the RAI‐abundance relationship was reliable, with a positive, statistically significant, accurate and precise effect, and a large effect size. We conclude that camera‐trap based RAI can be a reliable metric for winter monitoring of bezoar goats in remote and rugged mountain regions such as the South Caucasus. This offers a widely applicable, relatively low‐cost pathway to upscale mountain ungulate monitoring to larger areas to inform conservation planning and wildlife management.