The capacity for vocal recognition of individual conspecifics is well documented in many species, but the perceptual mechanisms that underlie this ability in oscines are less well understood. Using operant conditioning, we trained three groups of European starlings on a baseline task to discriminate the songs of one male starling from those of four others. Each subject heard songs from the same five singers, but the to-be-recognized individual varied among birds. We grouped the subjects according to sex and their degree of previous exposure to the songs used as stimuli in this experiment. The first group (N=5 males) identified their own songs from those of four familiar males. The second group (N=5 males) was familiar with the song stimuli, but none of the songs was their own. The third group (N=4 females) was unfamiliar with the songs. After learning the baseline discrimination, the subjects were exposed to new natural and synthetic stimuli. The subjects maintained the ability to identify correctly an individual on the basis of novel song bouts, and showed differential responding on the basis of the sequence of song types in song bouts that were modelled using Markov chains. Based upon patterns of responding to these different stimuli, we conclude that European starlings are capable of individual vocal recognition, and that this process is mediated by mechanisms involving the memorization of individually specific song types, the sequential ordering of song types within different bouts of an individual, and perhaps by individually specific spectral (or voice) characteristics that generalize across song types.
Mendeley saves you time finding and organizing research
Choose a citation style from the tabs below