Using multiple scales of movement to highlight risk-reward strategies of coyotes (Canis latrans) in mixed-use landscapes

Many wildlife species vary habitat selection across space, time, and behavior to maximize rewards and minimize risk. Multi-scale research approaches that identify variation in wildlife habitat selection can highlight habitat preferences and risk tolerance, but also movement strategies that afford coexistence or cause conflict with humans. Here, we examined how anthropogenic and natural features influenced coyote (Canis latrans) habitat selection in a mixed-use, agricultural landscape in Mendocino County, California, USA. We used resource selection functions and hidden Markov models to test whether coyote selection for anthropogenic and natural features varied by time of day or by behavioral state (resting, foraging, and traveling). We found that coyotes avoided development, but, contrary to our expectations, coyotes selected for roads, agriculture, and areas used by rifle hunters regardless of diel period or behavioral state. While traveling, coyotes increased selection for roads and av..., , R scripts, hidden Markov model, and data used to estimate the influence of natural and anthropogenic features on the behavioral and temporal dynamics of coyotes in an agricultural mixed-use landscape in the Sanel Valley, Mendocino County, CA, USA.