The process underlying vocal communication acquisition in humans, the most complex form of vocal communication, has been aptly named “Vocal Production Learning”. While human language remains unrivaled in complexity, a multitude of species have evolved individual hallmark features to approximate its intricacies and allow them to be ranked on a heterogenic spectrum. The zebra finch (Taeniopygia guttata) has been studied extensively and serves as an ideal model organism to investigate some of the characteristic features of vocal production learning: The auditory processing and integration of conspecific vocalizations and the potential for temporal and spectral adjustments of vocalizations during the critical developmental period. While behavioral approaches have provided insights into song learning, the neuronal mechanisms underlying this process remain poorly understood. The cortical vocal premotor nucleus HVC (proper name) is an integral part of the song system. In addition to receiving input from multiple upstream auditory nuclei, HVC innervates the downstream motor pathway, triggering song production, and sends efference copies of the motor program to the anterior forebrain pathway. To understand how representation of prominent temporal and spectral song features develops in the neuronal activity patterns of excitatory glutaminergic HVCRA/X projection- and local inhibitory GABAergic interneurons, I investigated their membrane potential during singing and listening to song, employing intracellular single- and extracellular multiunit recordings in awake juvenile and adult birds. During playback experiments, excitatory projection neurons of adult animals did not respond with consistent action potentials to either intact bird’s own song or to a pitch shifted or syllable swapped version thereof. In juvenile birds, however, precisely timed, highly robust response patterns temporally locked to individual syllables were elicited. These patterns occurred independent of the syllables position in the song. Furthermore, firing rates were altered in response to spectral shifts. Inhibitory interneurons in both adult and juvenile animals exhibited activity patterns precisely locked to the temporal aspects of the song while spectral song alterations only seemed to elicit limited responses in juvenile birds. In an additional set of experiments, I was able to provide evidence for a less efficient, less sparse representation of the premotor output program responsible for the elicitation of song production during singing in juvenile birds. These results indicate that the neuronal network in HVC undergoes a complex refinement process during song learning and maturation. The development of the inhibitory network is hypothesized to be responsible for the suppression of excitatory activity and ultimately the protection of the already learned temporal and spectral song features.
