Original investigationThe influence of human disturbance on occupancy and activity patterns of mammals in the Italian Alps from systematic camera trapping

As human activities increase in natural areas, so do threats to wildlife, potentially leading to immediate and long-term impacts on species distribution, activity, reproduction and survival. This is particularly relevant for large-bodied vertebrates that are especially sensitive to human presence and human-driven habitat changes. Assessing the impact of anthropogenic disturbance requires data on distribution and activity patterns of target species in relation to human presence and infrastructures. Here, we used camera trap data to study the influence of anthropogenic disturbance on the community of medium-to-large mammals in a mountainous area in the eastern Italian Alps, with emphasis on the local population of brown bear (Ursus arctos). In 2015, we sampled a study area of 220 km2 with 60 camera trap locations adopting a systematic grid. Such design was inspired by the terrestrial vertebrate monitoring protocol developed by the TEAM Network, a pan-tropical biodiversity programme. Camera traps run for 30 days in each site and cumulated 1978 camera trapping days, yielding 1514 detection events of 12 species of mammals. For the 8 most recorded species, we used detection/non-detection data to model estimated occupancy and detection probability in relation to a suite of environmental and disturbance covariates. Our analysis revealed that human disturbance plays a significant role in influencing species-specific detection probability, while we found little evidence of significant relationship between occupancy and anthropogenic disturbance. For example, we found that brown bear's detectability was negatively correlated with capture rate of humans at sampling sites, and positively correlated with distance from settlements. We also assessed species-specific daily activity patterns and found that, for all species, the overlap with human diel pattern decreased significantly at sites with higher human presence. We also discuss the potential of our approach for cost-efficient and long-term monitoring of mammals.