How to catch a ghost? Comparing two camera trap-based monitoring methods for elusive small mustelids in the Italian Alps

Small mustelids are increasingly recognized as species requiring conservation attention. In recent years, several camera-based methodologies have been developed to study them, but studies comparing different methods are still rare. To identify the most effective method to study small mustelid populations, we compared two camera-based monitoring methods in the Italian Alps. We also examined the effects of sampling session and habitat type on the occupancy probability and tested the “umbrella effect” of these methods for rodents. After superimposing a 700 × 700 m grid on an Alpine valley (Maritime Alps Natural Park, northwestern Italy), we surveyed 36 cells over three separate 45-day sessions from June to October 2023. In each cell, we employed (1) an “Alpine Mostela”, a foldable PVC box containing a camera trap and a PVC 9 cm Ø tube, and (2) a stand-alone trail camera. All devices were located at least 150 m from the others, and salmon oil was used as bait in half of the cells. To compare the methods, we used a single-season Bayesian occupancy model. The detection probability of stoats was higher with unbaited Alpine Mostelas and baited external cameras. We found the highest occupancy probability in the second session and non-forested habitats. Bait use positively affected the number of non-target videos. In this study, unbaited Alpine Mostelas and baited external cameras demonstrated reliable performance in detecting stoats. However, with the Alpine Mostela accomplishing slightly better results with much fewer non-target videos, it emerged as the preferred choice for long-term stoat monitoring.