Winter use of kill and scavenging sites by wolves in human‐modified landscapes

Abstract Understanding feeding behaviour of large carnivores is crucial for unveiling how adaptations to human‐modified landscapes can alter their behaviour and ecological role. In this study, we investigated wolf feeding activity during winter through interpretative field surveys of 454 clusters of GPS locations obtained from 8 wolves in the Abruzzo, Lazio, and Molise national Park (central Italy, 2008–2011). Using generalized linear mixed models, we explored spatio‐temporal use of feeding sites (i.e. kill and scavenging sites) accounting for the effect of ecological and anthropogenic factors. We detected feeding activity in 18.1% of the investigated GPS clusters, with 51.5% of the inspected feeding sites indicating scavenging on domestic prey. Wolves used feeding sites for an average of 2.4 days and revisited them about 3 (±2.7) times before being abandoned. Prey type (wild vs. domestic) and wolf category (i.e. solitary floaters, newly established breeding pairs, pack members) affected both prey handling time and recurrency of feeding site use. Pack members (≥3 wolves) spent relatively more time at feeding sites, especially those featuring large prey, and visited them more frequently compared to solitary floaters and wolf pairs. Although wolves used feeding sites mostly during the night, nocturnality significantly decreased with increasing distance to roads but not to settlements. We also revealed that time of cluster formation, number of visits, and mean slope best predict the presence of a feeding site at a GPS cluster. Despite the inclusion of scavenging sites and domestic prey, and limited to prey ≥15 kg, our predictive model would have revealed 62% of the feeding sites in the GPS clusters we investigated, while reducing of about 59% the field time required for ground truthing GPS clusters.