Bitter taste enhances predatory biases against aggregations of prey with warning coloration

Aposematic prey that possess chemical defenses advertise these to potential predators using conspicuous warning coloration. Aposematism is often associated with group living, which is hypothesized to enhance the protection of these species. Predators exhibit unlearned biases against foods with warning coloration, and the presentation of a novel sound or bitter-tasting toxin augments these biases. Whether these nonvisual signal components also cause naive predators to more strongly avoid aggregated prey, and whether biases against aggregations are restricted to situations where aggregated prey possess visual signals typically associated with aposematism, is unknown. We conducted an experiment in which naive domestic chicks (Gallus gallus domesticus) acted as predators and used artificially colored pastry prey. The experiment had a 2×2 design in which naive birds were offered a drop of either water or bitter-tasting chloroquine solution before being given the choice between solitary and aggregated prey that were either both red, a typical aposematic color, or both green (usually associated with crypsis and palatability). We found that birds were warier of red-aggregated prey and attacked significantly more solitary prey before aggregated prey compared with green. After sampling bitter-tasting chloroquine solution, the birds showed a bias in their attack decisions, attacking significantly fewer aggregated prey in total compared with those who had sampled water, but only when prey were red. Thus, exposure to a bitter-tasting toxin affected predatory preferences. We discuss our findings in relation to the mechanisms of bias, the benefits of group living, and the evolution of warning coloration and aggregation.This work was funded by Natural Environment Research Council grant NE/E016626/1 and a research fellowship to H.M.R. from the Wingate Foundation. H.M.R. is currently supported by a research fellowship from Churchill College, Cambridge.This is the author accepted manuscript. The final version is available from Oxford University Press via http://dx.doi.org/10.1093/beheco/art01