Rodent foraging is affected by indirect, but not by direct, cues of predation risk

We used foraging trays to determine whether oldfield mice, Peromyscus polionotus, altered foraging in response to direct cues of predation risk (urine of native and nonnative predators) and indirect cues of predation risk (foraging microhabitat, precipitation, and moon illumination). The proportion of seeds remaining in each tray (a measure of the giving-up density [GUD]) was used to measure risk perceived by mice. Mice did not alter their GUD when presented with cues of native predators (bobcats, Lynx rufus, and red foxes, Vulpes vulpes), recently introduced predators (coyotes, Canis latrans), nonnative predators (ocelots, Leopardus pardalis), a native herbivore (white-tailed deer, Odocoileus virginianus), or a water control. Rather, GUD was related to microhabitat: rodents removed more seeds from foraging trays sheltered beneath vegetative cover compared with exposed trays outside of cover. Rodents also removed more seeds during nights with precipitation and when moon illumination was low. Our results suggest that P. polionotus used indirect cues rather than direct cues to assess risk of vertebrate predation. Indirect cues may be more reliable than are direct scent cues for estimating risk from multiple vertebrate predators that present the most risk in open environments.