Resource Selection and Spatial Separation of Mule Deer and Elk during Spring

Abstract

Elk (Cerous elaphus) are sympatric with mule deer (Odocoileus hemionus) across large areas of western North America, where populations of mule deer have declined while elk populations increased. Insight into the interspecific relationships of elk and mule deer requires a clear understanding of the patterns of resource selection and spatial separation between the 2 species on shared range. Accordingly, we monitored distributions of mule deer and elk during spring, 1993-96 at the U.S. Forest Service Starkey Experimental Forest and Range (Starkey), in northeast Oregon. We recorded animal movements with an automated telemetry system (ATS) that provided locations of each radiocollared animal about every 1-4 hr, 24 hr/day. We linked animal locations to a geographic information system (GIS) of physical, vegetation, and human-activity variables identified from the literature as being important predictors of distributions of elk or mule deer, and used logistic regression to identify which variables best predicted resource selection. We validated models of resource selection by comparing predicted and observed selection in 2 study areas with 5 data sets of animal locations collected during spring, 1993-96. Of the 8 variables significant for either deer or elk, 4 variables were significant for 1 but not the other species, and coefficients (β) for 3 other variables differed in sign between the 2 species, demonstrating strong spatial separation between elk and mule deer. The resource selection function (RSF) for elk was a stronger predictor of mule deer selection than was the mule deer RSF as a predictor of elk selection. Mule deer distribution was inversely related to elk resource selection, but elk distribution was not related to mule deer resource selection. In 4 vegetation types that dominate Starkey, mule deer electivity index [(used - availability)/(used + availability)] declined as elk resource selection increased, while elk electivity index was independent of resource selection of mule deer. The dichotomy in resource selection between the 2 species, combined with the inverse relationship between mule deer locations and elk RSF, indicated that mule deer avoided areas used by elk. Removal experiments are needed under controlled conditions to explicitly test whether mule deer avoidance of elk results in decreased population or individual performance of deer.