A simple rule for the costs of vigilance: empirical evidence from a social forager

It is commonly assumed that anti-predator vigilance by foraging animals is costly because it interrupts food searching and handling time, leading to a reduction in feeding rate. When food handling does not require visual attention, however, a forager may handle food while simultaneously searching for the next food item or scanning for predators. We present a simple model of this process, showing that when the length of such compatible handling time H. is long relative to search time S, specifically H-c/S > 1, it is possible to perform vigilance without a reduction in feeding rate. We test three predictions of this model regarding the relationships between feeding rate, vigilance and the H-c/S ratio, with data collected from a wild population of social foragers (samango monkeys, Cercopithecus mitis erythrarchus). These analyses consistently support our model, including our key prediction: as H-c/S increases, the negative relationship between feeding rate and the proportion of time spent scanning becomes progressively shallower. This pattern is more strongly driven by changes in median scan duration than scan frequency. Our study thus provides a simple rule that describes the extent to which vigilance can be expected to incur a feeding rate cost.