Winter habitat use of roe deer (Capreolus capreolus) in the presence of lynx (Lynx lynx)

Population sizes of large carnivores have increased in Scandinavia during the last century. Increases in predator population sizes can affect prey populations not only through increased mortality, but also through behavioural responses as prey redevelop anti-predator behaviours. In this study, the habitat use of roe deer (Capreolus capreolus) in an area with lynx (Lynx lynx) was investigated to assess the relative importance of predators and other factors in shaping their use of habitat and cover. Bed sites had more cover than foraging sites; a difference that was also reflected in the different use of open habitats vs. forests for bedding and foraging. Activity type was not the only factor contributing to differences in habitat use. Local weather (snow depth, temperature and wind speed) explained much of the variation in canopy cover and distances between beds and foraging sites. Roe deer used sites with more canopy cover when temperatures were low and windspeed high. They also walked shorter from beds to foraging sites when snow was deep. As the winter progressed, fat reserves will deplete and the energy budget becomes tighter. That effect was expressed by the selection for more canopy cover at foraging sites by night, less by day, decreasing distance between beds and foraging sites, and from beds to humans, as the season progressed. Data fits the hypothesis of tighter energy budgets for families (females with young at heel). They had higher canopy cover over foraging sites and walked shorter distances from beds to foraging sites. Males used artificial feeding sites less often, and beded further from humans than females. These indications might suggest that males are more cautious towards humans compared to females, possibliy because of a higher mortality due to hunting. Few clear differences between the current study and earlier studies from areas without lynx were found. Two non-exclusive explanations are suggested. Despite the presence of resident lynx, few individual roe deer are ever attacked by lynx in this area, thus it will be highly adaptive to adjust trade-offs between predator avoidance and other factors like climatic stress and available forage in favour of the latter when predation risk is low. In addition, weather was quite severe in the winter when the study was conducted, further increasing the importance of avoiding climatic stress. To assess whether roe deer adjust their level of predator avoidance to the immediate predation pressure will require further studies on responses to predator presence