Supplemental feeding for ungulates is a widespread practice in many human‐dominated landscapes across Europe and North America, mainly intended to seasonally support populations. Surprisingly, little consideration was given so far to the effect of supplemental feeding on ungulate spatial ecology at a large scale, in management and conservation studies. Analyses of the main ecological drivers influencing the use of supplemental feeding sites by ungulates across a gradient of abiotic and biotic factors are currently lacking. We conducted a large‐scale assessment of ecological and management drivers of use of feeding station sites in roe deer (Capreolus capreolus), a small cervid widely distributed across Europe that is particularly sensitive to winter severity. We tested four competing hypotheses by comparing the time spent at feeding station sites by 180 individual Global Positioning System‐collared roe deer from nine populations spanning a wide latitudinal and altitudinal gradient. We found that roe deer used feeding station sites highly opportunistically in response to winter severity across its range. The harshest weather conditions at the northern range limit or the highest elevations provoked an intense use of feeding station sites, which typically peaked at the end of winter, in accordance with the adverse weather and nutritional condition hypotheses. Consistently, milder winters corresponded to a reduced and/or more homogeneous use of supplemental feeding. In general, intensively used feeding station sites heavily conditioned spatial behavior of roe deer. Importantly, biotic factors such as the presence of competitors decreased roe deer use of supplemental feeding station sites. Our results emphasize the importance of this human‐induced alteration to resource distribution, especially in the context of the rapidly occurring climate change that is modifying resource availability for ungulate populations.
Human-induced environmental alterations in the Alps may importantly affect small mammal species, but evidence in this sense is limited. We live-trapped small rodents in the Central-Eastern Italian Alps in three close-by habitat types (rocky scree, alpine grassland, and heath) at 2100 m a.s.l. during summer-fall, in 1997 and 2016. We compared small rodent assemblages through a Redundancy Detrended Analysis (RDA). In both surveys, we detected two specialist species, i.e., the common vole (Microtus arvalis) and the snow vole (Chionomys nivalis), and, unexpectedly, the forest generalist bank vole (Myodes glareolus). In 1997, grassland was mainly occupied by the common vole, while the bank vole and the snow vole were sympatric in the other habitats. In 2016, the snow vole was detected only in the scree, while other species did not show distribution changes. We discuss a series of hypotheses that might have driven the differences observed across decades, among which is a species-specific response to abiotic and biotic environmental alterations, with the alpine habitat specialist moving out of sub-optimal habitats. We encourage further research on this topic, e.g., via long-term longitudinal studies.