Anti-snake behavior and snake discrimination in vervet monkeys

Abstract

Anti-predator behavior allows to investigate how animals classify potential threats in their environment and which cognitive mechanisms might be involved in risk assessment. Snakes are common predators for many primate species, yet most of our knowledge on primate anti-snake behavior stems from predator model experiments. Few studies have investigated natural predator-prey interactions. Here, we combine an observational study on anti-snake behavior in free-ranging vervet monkeys, Chlorocebus pygerythrus, with an experimental study of snake classification, to achieve an integrated understanding of the monkeys responses to and classification of snakes. Over 25 months, we gathered data from over 200 individuals in five groups and recorded natural encounters between vervet monkeys and five different species of snakes. We assessed whether the time that monkeys spent inspecting snakes from close by varied with the encountered species. We further examined whether age and sex classes differed in their propensity to inspect snakes or alarm call. Pythons were associated with higher average inspection durations and were more likely to elicit inspection responses. While adult males were less likely to approach and inspect snakes, we found no effect of age or sex on alarm calling probability. Alarm calling appeared to be restricted to individuals in close proximity to snakes, suggesting that recruitment of nearby group members is an essential function of this call type. In the experiments, we tested whether vervet monkeys discriminated snake models by length and/or diameter, but found no effect of model size on the intensity of anti-snake behavior. While the experimental data showed the same trends as data from natural encounters, responses towards model pythons were much stronger than responses towards real pythons. These results point to a potential sampling bias in predator model experiments that needs to be taken into account when assessing data from experiments alone.