Factors associated with elk (Cervus canadensis) distributions during rifle season and individual responses to harvest risk in a prairie environment

Abstract

Hunting pressure alters habitat selection of elk and understanding responses to pressure is important for effective population and habitat management. While elk responses to hunting are well-studied in forested and mountainous environments in the western U.S., little is known about elk habitat selection in more open prairie landscapes. Our objectives were to evaluate the effects of landscape and environmental factors on elk habitat selection during the rifle season, assess individual variability, and investigate relationships between selection and harvest risk for male and female elk in two populations: the Custer Forest and Missouri River Breaks in eastern Montana, USA. We also provided management recommendations for habitat characteristics that allowed elk to mitigate exposure to harvest risk (i.e., security habitat) based on where most elk use occurred. We used resource selection function modeling with a use-available design and added random effects to estimate selection patterns for individual elk. Resource selection coefficients indicated that elk generally selected areas with restricted hunter access, rugged terrain, and greater distances from motorized routes with a few notable differences for elk in the Missouri River Breaks. In particular, canopy cover was consistently associated with large increases in relative probability of use. Estimates of individual random coefficients indicated that while elk typically shared the same direction of selection for a given covariate, individuals varied in the strength of these relationships, likely due, in part, to varying exposure to risk. Individual elk increased selection for habitat features that provided security when faced with higher harvest risk (i.e., the proportion of used locations that fell on publicly accessible lands). Our results indicated canopy cover was a particularly important factor associated with elk selection during the rifle season in our prairie landscapes where available cover is relatively limited. The relative importance of other factors varied depending on study area and sex. Based on where most elk use occurred, we recommend managing for security and preferred security areas with canopy cover > or = 28% and > or = 37% in the Custer Forest and > or = 3% and > or = 5% in the Missouri River Breaks, respectively, during the rifle season in these areas and prairie landscapes with similar habitat attributes.