CONICET | Buscador de Institutos y Recursos Humanos

Birdsong is a complex motor activity thatemerges from the interaction between the peripheral system, the central nervoussystem and the environment. The similarities to human speech, both inproduction and learning, have positioned songbirds as unique animal models forstudying the production and perception of this learned motor skill.In this work, I will present a low dimensionaldynamical system as a model of the avian neural network for song production. Wedeveloped a neural model in which the variables were the average activities ofdifferent neural populations within the nuclei of the song system. We performedelectrophysiological experiments to record neural activity from one of thesenuclei and found that the low dimensional model could reproduce the neuraldynamics observed. Also, this model could reproduce the respiratory motorpatterns used to generate song. We showed that sparse activity in one of theneural nuclei could drive a more complex activity downstream in the network. Thisinterdisciplinary work shows how low dimensional models can be a valuable toolfor studying the emergence of complex motor tasks.