Vocal conditioning with playback of two template sounds in budgerigars

Abstract

We examined a capability of budgerigars to produce a similar vocal pattern to a sound stimulus presented immediately just before. For this purpose, we trained birds using an operant conditioning procedure. In the training, two types of the birds’ own call were used as the auditory stimuli. Then, they were tested by probe stimuli (another vocal pattern of the subjects’ own, other birds’ vocalization). At test trials, the birds vocalized not any similar sounds in response to the probe stimuli but one of the vocal patterns which was produced at the training trials. Then, we attempted to train the birds to produce vocal patterns following playback sounds which were slightly changing as the trials went. 24-step intermediate sounds between two birds’ own vocal patterns were synthesized. Those intermediate sounds were shifting step-by-step from one to the other at each trial in a single session. Eventually, a bird created some novel sounds which were similar to those intermediate stimuli and had been never produced at the training trials. Taken together, the birds did not use the playback sounds as the vocal reference under the operant procedure. However, birds might store those playback sounds as the potential vocal repertoire.