Using learned calls to study sensory-motor integration in songbirds.

Communicating songbirds produce calls as well as song and some of these are learned. One of these-the long call in zebra finches-is used by both sexes in similar behavioral contexts, but is learned in males and not in females. The male long call includes learned spectral and temporal features. In several studies, the learned long call has been used as a tool to study sensory-motor integration and vocal learning in a way that complements the use of song. Lesion studies showed that production of the male-typical call features requires an intact nucleus RA, the sexually dimorphic source of the telencephalic projection to brainstem vocal effectors. Behavioral studies that quantified zebra finch calling in response to long call playbacks showed that intact adult males have a categorical preference, absent in females, for the long calls of females over those of males. By using synthetic call stimuli, it was found that males use both spectral and temporal information to classify long call stimuli by gender, but that females use only temporal information. In juvenile males, the emergence of categorical preference occurs during the same period when RA matures anatomically (40-50 days) and the first male-typical vocalizations are produced. Adult males with RA lesions lost the categorical preference for female long calls, suggesting that RA could also play a role in long call discrimination. Preliminary analysis of recordings from neurons in NCM-a telencephalic auditory area (see Mello and colleagues, this volume)-suggests a pattern of responses to the spectral features of synthetic call stimuli that parallels the behavioral responses they elicit.