Comparative Assessment of Food Preferences and Aversions Acquired by Blackbirds via Observational Learning

Abstract

Red-winged Blackbirds (Agelaius phoeniceus) can acquire food preferences and aversions merely by observing conspecifics. In Experiment 1, red-wings were trained to prefer or avoid food paired with yellow, as conspecifics watched. After training, all birds were given two-choice tests between food paired with yellow or green for 12 days. Trainers were tested in visual isolation, whereas watchers were tested either in visual isolation or in visual contact with birds who had observed the opposite behavior during training. Food aversions were more resistant to extinction than food preferences (P < 0.05), and, among watchers, social cues facilitated avoidance (P < 0.05). In Experiment 2, red-wings were trained to avoid food paired with yellow, as Common Grackles (Quiscalus quiscula) and red-wings watched, or vice versa, and then two-choice tests were given between yellow and green. Both grackles and red-wings exhibited observational learning, regardless of the training species (P's < 0.05).Social interactions among avian predators may influence how predator abundance affects the Batesian model-mimic complex. We speculate that avoidance learning, which occurs when a predator observes the ingestion of a model, is stronger than preference learning, which occurs when a mimic is ingested. Relatively few models would be needed for the model-mimic complex to operate successfully, and the number of mimics could exceed the number of models without jeopardizing the mimetic advantage.