A two-level problem: habitat selection in relation to prey abundance in an ambush predator, the speckled rattlesnake (Crotalus mitchellii)

Understanding how food distribution influences habitat choice is a central goal of behavioural ecology. Because animals select habitats at different scales, and because the selection criteria used may be scale-dependent, studies examining the significance of food in relation to habitat selection should ideally incorporate more than one spatial scale. We relied on spatial ecology data and prey abundance estimates to examine whether prey (rodent) abundance is a consistent predictor of habitat selection at two different spatial scales in an ambush predator, the speckled rattlesnake (Crotalus mitchellii), during the mating and post-mating seasons. At the macrohabitat level, selection and avoidance of particular macrohabitats by C. mitchellii positively correlated with rodent abundance during the post-mating season. Because most snake foraging activity occurs during the post-mating season, this finding suggests that C. mitchellii selects macrohabitats that increase prey encounter rates. In contrast, rodent abundance was relatively low at the microhabitats selected by the snakes. Therefore, the higher prey abundance in C. mitchellii’s preferred macrohabitat did not translate into increased access to prey at the snake microhabitats, indicating that prey distribution may not be a constant predictor of predator habitat choice across spatial scales. Indirect evidence suggests that lower prey abundance at the snake microhabitats was the result of changes in rodent behaviour (e.g., avoiding or reducing activity at snake locations) to decrease predation risk. Our study underscores the dynamics of predator–prey interactions in nature, and emphasizes the challenges that ambush predators may face when foraging for risk-sensitive prey.