Selecting substandard prey in spatial dimensions

Background/Question/Methods

Predators are perceived to be highly selective in their pursuit of prey.  This selectivity theoretically results in a disproportionately higher take of substandard prey when compared to healthy prey.  The ecological theory describing this selectivity, however, has been developed without explicit consideration of spatial dimensions.  Considering space as a vital component to understanding the interactions of predator and prey, we assessed predator selection of substandard prey.  We evaluated whether wolves (Canis lupus) killed pathologic moose (Alces alces) in distinct areas as characterized by differences in landscape structure when compared to areas where non-pathologic moose were killed.  We endeavored to answer two questions, 1) are pathologic prey killed in different geographical areas when compared to non-pathologic prey and, if so, 2) do differences in landscape structure between these areas exist?  Differences between kill areas would suggest that the functional role of habitats differed for pathologic versus non-pathologic moose, where increased vulnerability of pathologic moose likely influenced habitat selection.  We developed a multi-level model to examine landscape structure at individual kill-sites (spatial linear mixed regression) and within core concentrations of kill sites (kill zones; multiple linear survey regression) for pathologic and non-pathologic moose.   

Results/Conclusions

Our results demonstrate that wolves killed moose of differing body condition within distinct kill sites and kill zones.  At kill sites we found that the likelihood of wolves killing pathologic moose was greater in sites with higher elevation, less overstory canopy cover, and more percent conifer when compared to sites where non-pathologic moose were wolf-killed. Within kill zones pathologic moose were more likely to be killed in zones characterized by higher elevations that were farther away from frozen inland lakes, and farther away from the frozen coastline of Lake Superior when compared to wolf-killed non-pathologic moose.  This analysis demonstrates that wolves kill substandard prey in specific spatial dimensions and that landscape structure of these areas functions differently depending on moose body condition.  Our findings complement and enhance the current ecological theory on the selectivity of predators for substandard prey.