Auditory Processing in Songbirds

Auditory experience is essential for song learning and, in some species, for the maintenance of normal adult song patterns. The connectivity and functioning of brain nuclei controlling song learning, perception and production have been extensively examined. Together, these nuclei are called the song system. In the song system, auditory neurons show specialized sensitivity to the bird's own song. This specialized auditory tuning develops as the bird is learning to produce its song. The song system receives its primary sensory input from the auditory system, yet very little is known about how vocalizations are processed by the auditory system. My long-term goal is to characterize the processing of vocalizations used for communication in the songbird auditory system and to examine the role of this processing in song learning and maintenance. The proposed research will examine the processing of natural and synthetic sounds in the songbird auditory midbrain region, the mesencephalicus lateralis, pars dorsalis (MLd). This brain region is the avian homologue of the mammalian inferior colliculus (IC), where specialized responses to conspecific vocalizations are found in bats and mice. The first set of experiments will characterize the spectral temporal tuning of Mld neurons. Characterizations will be done both directly using conventional methods (Le by characterizing a neuron's frequency tuning curves, amplitude modulation and frequency modulation responses) and indirectly by first estimating the spectral temporal receptive fields (STRFs) of neurons with complex synthetic stimuli and then extracting the tuning parameters of the neurons from the STRF. Our second Specific Aim is to assess whether the spectral-temporal characteristics completely describe the response properties on Mld neurons or whether higher order structure (such as that found in natural sound) can also affect their responses. The third Aim will begin to address the issue of auditory tuning for conspecific vocalizations. We will characterize the responses of MLd neurons in two species of songbird, zebra finches and bengalese finches, to conspecific and heterospecific vocalizations. These experiments will test whether or not Mld neurons are specialized to extract the acoustic features of conspecific vocal signals. Taken together, the experiments are designed to provide the first examination of auditory processing related to vocal communication in the songbird midbrain.