Predicting risk of livestock depredation by wolves in southwestern Alberta

Wolves can potentially prey on all ungulates within their distributional range, including domestic livestock. The potential for conflict between wolves and humans therefore exists especially in rural areas where livestock production is a major economic activity, such as southwestern Alberta. Limited studies have examined factors that predispose livestock to depredation by wolves, and none occurred in southwestern Alberta. The purpose of this study was to determine the spatial relationships between habitat characteristics, human use and wolf depredation on livestock in southwestern Alberta. The goal is to use these characteristics as predictors for areas at risk. We used Geographic Information Systems (GIS) to examine the effects of vegetation productivity, geography, and proximity to roads, rivers, and cover on predicting livestock depredation by wolves. Binary logistic regression analyses, ranked using Akaike Information Criteria (AIC), were used to determine what variables were best at explaining depredation occurrence. On private lands, greenness and elevation were important variables in the best logistic regression model (y = -22.366 + 0.009(elev) + 0.024(green)). These variables were also significantly different between depredated and random sites (elevation (tcrit = 1.97, p = 0.0035), greenness (P tcrit = 1.97, p = 4.40E-08)). Our results indicate that ranches (and land within an 8-km buffer of them) in proximity to the Rocky Mountains and in areas of higher vegetation productivity are at risk of depredation by wolves.