Needle in a Forest: Searching for Thresholds to Maximize Coexistence Between People and Wildlife in a Remote Wilderness Park.

Managers of protected areas are charged with providing public access and enjoyment while minimizing wildlife disturbance, a difficult feat given that wild animals often respond to humans as predators. Evaluating the effects of people on wildlife is critical to inform management actions, but designing methods to assess visitor impacts or test mitigations with free-ranging wildlife is challenging. In Glacier Bay National Park, Alaska, we tested the effects of vessel-based bear viewing on the behavior of brown bears using experimental approaches, and we tested the effects of human presence on wildlife activity at one frontcountry and four backcountry locations during times of normal (2017-2018) vs. low visitation (2020) using remote cameras. We found that vessel approaches within 100 m of brown bears led to increased stress behaviors. Detection rates for five species (black bear, brown bear, wolf, moose, and coyote) varied by site and year. For most species, we did not detect a threshold of human activity at which wildlife activity clearly declined. However, at the four backcountry sites, wolf activity dropped to zero as human activity exceeded 20 camera minutes/week. Furthermore, wolves were not detected at the frontcountry site during all years, suggesting spatial avoidance of a high-human-use area. Additional research is warranted to determine the strength of these associations and whether these observations are generalizable. Our study provides a framework for estimating thresholds of human disturbance to improve management of protected areas.