Ecology of fear alters behaviour of grizzly bears exposed to bear-viewing ecotourism

Humans are perceived as predators by many species and may generate landscapes of fear, influencing the spatiotemporal activity of wildlife. Additionally, wildlife might seek out human activity when faced with predation risks (human shield hypothesis). We used the Anthropause, a decrease in human activity resulting from the COVID-19 pandemic, to test the ecology of fear and human shield hypotheses and quantify the effects of bear-viewing ecotourism on grizzly bear (Ursus arctos) activity. We deployed camera traps in the Khutze watershed in Kitasoo Xai’xais Territory in the absence of humans in 2020 and with experimental treatments of variable human activity when ecotourism resumed in 2021. Daily bear detection rates decreased with more people present and increased with days since people were present. Human activity was also associated with more bear detections at forested sheltered sites, and less at exposed sites, likely due to the influence of habitat on bear perception of safety. The number of people negatively influenced adult male detection rates, but we found no influence on females with young detections, providing no evidence that females responded behaviourally to a human shield effect from reduced male activity. We also observed apparent trade-offs of risk avoidance and foraging. When salmon levels were moderate to high, detected bears were more likely to be females with young than adult males on days with more people present. Should managers want to minimize human impacts on bear activity and maintain baseline age-sex class composition at ecotourism sites, multi-day closures and daily occupancy limits may be effective. More broadly, this work revealed that antipredator responses can vary with the intensity of risk cues, habitat structure, and forage trade-offs, as well as manifest as the altered age-sex class composition of individuals using human-influenced areas, highlighting that wildlife avoids people across multiple spatiotemporal scales.

Methods
Data were collected through an array of 40 remote cameras in the Khutze watershed in Kitasoo Xai'xais Territory to assess how ecotourism in the form of bear viewing might influence grizzly bear activity. This was assessed through multiple modelling techniques, including an overall bear detection rate model, individual detection rate models for different age-sex classes of bears, a multinomial model to further assess age-sex class differences, and an activity pattern analysis to assess temporal shifts in grizzly bear activity.

We used data from 2020 when ecotourism was closed in Khutze, as well as 2021 when ecotourism returned. For the full methods, please see the open access paper, Short et al., 2024 "Ecology of fear alters behaviour of grizzly bears exposed to bear-viewing ecotourism".