What explains variation in the strength of behavioral responses to predation risk? A standardized test with large carnivore and ungulate guilds in three ecosystems

If access to food is affected by the risk of predation, then the number of individuals killed by predators is an incomplete measure of the limiting effect of predation. Nonetheless, it is often assumed that the costs of antipredator responses (risk effects) are either small enough to be ignored or positively correlated with direct predation, and thus unlikely to alter inferences based on predation rates. These assumptions are rarely tested. Here we studied five large carnivores and ten prey species in three Zambian ecosystems to test relationships between direct predation, antipredator vigilance and trade-offs with foraging. The presence of a predator caused vigilance to increase by a factor of 2.4, with substantial variation among prey species in the strength of this response. This was associated with a 28% decrease in the proportion of individuals foraging, a trade-off that was consistent across species. We detected no correlation between direct predation and the strength of antipredator responses, which undermines the gambit of ignoring risk effects. The strength of antipredator responses was uncorrelated with broad attributes of predators and environments, but was correlated with attributes of prey. Responses were stronger for small species and for browsers/mixed feeders relative to grazers. It has previously been noted that small ungulates face higher rates of direct predation. Building on this inference, our results suggest that carnivore loss/restoration will also have stronger behaviorally-mediated effects on small ungulates, particularly browsers and mixed feeders. If such species increase their representation where carnivores are depleted, then cascading effects on vegetation would be expected.