The effects of frequency content on auditory streaming in budgerigars and zebra finches.

Abstract

Auditory streaming is a phenomenon that has been documented in a wide variety of animal species. Recently, space, intensity, time, and spectral composition were all found to be important factors for auditory streaming of birdsong by budgerigars and zebra finches, although the cues varied in importance. Those experiments were extended here to further examine the role of frequency characteristics on the auditory streaming of birdsong. The birds were initially trained using operant conditioning procedures to differentially peck keys in response to either a synthetic zebra finch song consisting of five syllables (whole song) or to the same song with the fourth syllable omitted (broken song). Correct responses were reinforced with millet, and incorrect responses were punished with a lights-off timeout period. Once the birds reached high-performance levels with the training stimuli, probe trials were inserted on a small portion of all trials. The probe stimuli contained either a white noise burst in the missing syllable’s location or narrowband pieces of the original fourth syllable. Results show that the different cues are differentially effective in eliciting streaming of birdsong by zebra finches and budgerigars, similar to the previous experiments and results from human speech experiments.