Context dependence of elk (Cervus elaphus) vigilance and wolf (Canis lupus) predation risk

To assess the relationship between predation risk perceived by elk (Cervus elaphus L., 1758) as evidenced by vigilance, we conducted focal animal observations in elk winter range. We stratified our observations in Glacier National Park, Montana, USA, and Waterton Lakes National Park, Alberta, Canada, in valleys with three wolf (Canis lupus L., 1758) population levels (Saint Mary Valley: no wolf; Waterton Valley: moderate wolf; North Fork Valley: high wolf). Although the lowest elk vigilance occurred in Saint Mary and the highest in the North Fork, our analysis revealed a complex picture. Our model included distance to forest edge, group size, distance to road, social class, and impediments to detecting and escaping wolves. In Saint Mary, none of the variables were significant. In Waterton, vigilance decreased as elk group size increased (p < 0.00001) and increased as impediments increased (p = 0.0005). In the North Fork, vigilance increased as group size increased (p = 0.03), bulls were more vigilant (p = 0.02), and the interaction between group size and impediments was significant (p = 0.03). Where a high wolf population existed, elk did not exhibit uniform or expected response to predation risk factors. High wolf presence may necessitate adaptive elk behaviour that differs from response to moderate wolf presence.