Movement ecology, landscape connectivity, and hunting-mortality risk in a transboundary European brown bear (Ursus arctos) population

Conserving large mammalian carnivores remains one of the greatest challenges for global biodiversity, due to their high energetic and spatial demands and their increasing exposure to habitat fragmentation and human-caused mortality, such as hunting. In Europe, the overall successful recovery of the brown bear Ursus arctos illustrates the complex task of fostering coexistence within heavily human-modified landscapes, for instance, by maintaining landscape connectivity and mitigating the spatially heterogeneous risk of mortality. The transboundary Finnish-Russian Karelian brown bear population, managed through legal hunting and facing limited connectivity to the Scandinavian peninsula, provides a representative case study for exploring these ecological dynamics. This thesis aimed to identify the spatial and temporal scales at which direct and indirect forms of human disturbance may affect movement behaviour, space-use patterns, and landscape connectivity in this population. The research comprises four chapters integrating behavioural, movement, and spatial ecology, and draws on long-term (2002–2014) GPS telemetry and hunting-mortality datasets. Specifically, I employed advanced analytical tools, including step-selection analysis for movement and habitat selection, Circuitscape for modelling connectivity, and resource selection functions for assessing mortality risk. First, I found that mating requirements strongly influenced male brown bear movements, leading to increased daily displacements within high-disturbance areas and indicating a behavioural response to elevated risk perception. Second, when comparing movement rates across two regions differing in human impact, bears in the more disturbed region showed lower overall movement, suggesting proactive spatial adjustments to minimise risk exposure. However, the absence of a temporal avoidance signal suggested that individuals may be approaching the limits of behavioural plasticity. Third, spatial modelling revealed that roughly 44% of the potential corridor area in Finland overlapped with zones of high hunting-mortality risk, implying that such risk may constrain intra-population connectivity toward Scandinavia by acting as a functional barrier to bear movements. Fourth, among four alternative approaches tested for modelling hunting-mortality risk, I found the random-background method yielded the highest predictive performance, suggesting that harvest primarily reflects landscape accessibility and human-use patterns rather than animal space use alone. Overall, the findings from this thesis support the hypothesis that human-caused mortality risk, arising from multiple anthropogenic activities, may significantly undermine landscape connectivity — a process of broad relevance for large-carnivore conservation worldwide. Moreover, this research provides practical methodological insights for wildlife managers, helping to design effective harvest management strategies and foster long-term human–wildlife coexistence.