97th ESA Annual Meeting (August 5 -- 10, 2012)

PS 36-182 - Human activity modulates mule deer risk assessment

Tuesday, August 7, 2012
Exhibit Hall, Oregon Convention Center
Evelyn H. Strombom, Biology, University of Pennsylvania, Philadelphia, PA, Mary V. Price, Rocky Mountain Biological Laboratory, Crested Butte, CO and Daniel T. Blumstein, Department of Ecology and Evolutionary Biology, University of California, Los Angeles, CA
Background/Question/Methods

Behavioral ecologists quantify anti-predator behavior via measurement of flight initiation distance (FID), the distance at which an animal moves away from an approaching threat.  Current theory uses cost-benefit analysis to evaluate the many factors that influence escape behavior; however, the factors determining the economic decision differ among species. The goal of this study was to determine the geographic, social and anthropogenic factors that influence mule deer (Odocoileus hemionus) flight behavior around the Rocky Mountain Biological Laboratory in Gothic, CO.  Mule deer at this site are important primary consumers. As such, their browsing patterns, in part determined by predator-prey relationships, have important consequences for the ecosystem. Focal individuals were observed in July and August, during dawn and dusk when deer are most active. An observer approached the focal individual at a measured pace of 1m/sec, noting Alert Distance, the distance at which the deer directs its head toward an approaching human, and FID.  Geographical features (type of habitat, proximity to cover and type of cover), social features (group size and presence of fawns) and distance from the center of town, a proxy for human activity level, also were noted.  ArcMap 10 was used to map deer locations and calculate distances from the town center. Bivariate and multivariate analyses were used to determine the optimal model for prediction of FID and were conducted in DataDesk. 

Results/Conclusions

Bivariate analyses revealed that FID increased significantly with Alert Distance (R2=41.7, p<0.0001), and distance from the center of town (R2= 19.6,  p<0.0034).  Geographic and social features did not have a significant effect on mule deer FID.  In the multivariate model, only Alert Distance (p<0.0001) and distance from town center (p<0.0288) were significant. Stepwise backwards exclusion and best subsets regressions were used for comparison. The change in behavior with distance from the town center suggests habituation to human activity. Reduced threat perception of humans has implications for mule deer browsing patterns and, subsequently, the communities built upon the resources they consume.