When animals are faced with extraordinary energy-consuming events, like hibernation, finding abundant, energy-rich food resources becomes particularly important. The profitability of food resources can vary spatially, depending on occurrence, quality, and local abundance. Here, we used the brown bear (Ursus arctos) as a model species to quantify selective foraging on berries in different habitats during hyperphagia in autumn prior to hibernation. During the peak berry season in August and September, we sampled berry occurrence, abundance, and sugar content, a proxy for quality, at locations selected by bears for foraging and at random locations in the landscape. The factors determining selection of berries were species specific across the different habitats. Compared to random locations, bears selected locations with a higher probability of occurrence and higher abundance of bilberries (Vaccinium myrtillus) and a higher probability of occurrence, but not abundance, of lingonberries (Vaccinium vitis-idaea). Crowberries (Empetrum hermaphroditum) were least available and least used. Sugar content affected the selection of lingonberries, but not of bilberries. Abundance of bilberries at random locations decreased and abundance of lingonberries increased during fall, but bears did not adjust their foraging strategy by increasing selection for lingonberries. Forestry practices had a large effect on berry occurrence and abundance, and brown bears responded by foraging most selectively in mature forests and on clearcuts. This study shows that bears are successful in navigating human-shaped forest landscapes by using areas of higher than average berry abundance in a period when abundant food intake is particularly important to increase body mass prior to hibernation.Significance statementFood resources heterogeneity, caused by spatial and temporal variation of specific foods, poses a challenge to foragers, particularly when faced with extraordinary energy-demanding events, like hibernation. Brown bears in Sweden inhabit a landscape shaped by forestry practices. Bilberries and lingonberries, the bears’ main food resources in autumn prior to hibernation, show different temporal and habitat-specific ripening patterns. We quantified the bears’ selective foraging on these berry species on clearcuts, bogs, young, and mature forests compared to random locations. Despite a temporal decline of ripe bilberries, bears used locations with a greater occurrence and abundance of bilberries, but not lingonberries. We conclude that bears successfully navigated in this heavily human-shaped landscape by selectively foraging in high-return habitats for bilberries, but did not compensate for the decline in bilberries by eating more lingonberries.Electronic supplementary materialThe online version of this article (doi:10.1007/s00265-016-2106-2) contains supplementary material, which is available to authorized users.

Animals adapt their foraging behavior to variations in food availability and predation risk. In Sweden, brown bears (Ursus arctos) depend on a nearly continuous intake of berries, especially bilberries (Vaccinium myrtillus) during late summer and early autumn to fatten up prior to hibernation. This overlaps with the bear hunting season that starts on 21 August. Bilberry occurrence varies across space, as does human-induced mortality risk. Here, we hypothesize that brown bears select for areas with a high probability of bilberry occurrence across a boreal forest ecosystem in Sweden (H1), and that human-induced mortality risk reduces bear selection for bilberries (H2). In addition, we hypothesized that bears that survived the hunting season avoided bilberry areas associated with high risk, whereas bears that were later killed selected more strongly for berries and less against risk prior to the hunting season (H3). To evaluate our hypotheses, we used resource selection functions to contrast bear GPS relocation data (N = 35, 2012-2015) and random positions within the bears´ home range with generalized linear mixed effect models against two focal variables: a map predicting bilberry occurrence and a map predicting human-induced mortality risk. We found that bears selected for areas with a high probability of bilberry occurrence (supporting H1), but avoided these areas if they were associated with and high risk of hunting mortality (supporting H2). The killed and surviving bears did not differ in their selection for bilberries, but they did differ in their selection against risk (partially supporting H3). Surviving bears strongly avoided high risk areas, whereas killed bears responded less to risk and selected for high-risk areas with a low probability of bilberry occurrence. This suggests that killed bears selected for other food sources than berries in high risk areas, which exposed them to human hunters. We conclude that bears respond to a landscape of fear during the berry season and that different foraging strategies may have a direct impact on individual mortality during the hunting season.