Welcome

The Margoliash Lab

Neuroethology is an approach to studying the nervous system that relies on analysis of natural behavior to give insight into mechanisms of those behaviors.  It emphasizes the study of mechanisms and development of mechanisms considering behaviorally relevant stimuli and responses, often in highly specialized animals.  Ultimately, this approach evaluates the functional consequences of behavior, and is centered in evolutionary biology.
We take a neuroethology perspective to the study of vocal learning.  Our approach is comparative, studying several species of song birds and humans, with strong interactions between elements of the research program directly informing each other.  We adopt whatever technology is necessary to answer the question.  This include electrophysiology from brain slices to chronic recordings in freely-moving singing animals, from single units to multisite recordings, recordings and analyses of developing birds, operantly trained birds, and other behavioral approaches, a broad range of histological techniques, and computational and modeling approaches.  In the past but not currently, this has also included field work.
One focus of our research is the mechanisms of developmental song learning.  We investigate the representation of sensory memories in very young pre-singing birds, and how these interact with auditory feedback (during daytime singing) to shape behavior.  We have also demonstrated neuronal replay during sleep, how replay changes the structure of spike bursts during sleep in adult birds, and how replay begins to emerge at the onset of song learning, just prior to the first changes in objective singing behavior.  These studies motivate the hypothesis that during sleep, bursting spontaneous discharge carries sensory information, and that sleep modifies networks via an interaction between  sensory and motor representations.  Collaborative work in humans studies have helped to highlight the breadth of the role of sleep in learning.  This has stimulated us to turn back to the animal work, to establish a more general animal model for sleep research (song perceptual learning in adult starlings).  Our interest in state-dependent mechanisms of learning has also stimulated EEG studies of sleep structure in birds, and technical development of algorithms to analyze EEG signals, and motivates hypotheses regarding the evolution of vocal learning and the vertebrate forebrain.
We have demonstrated neuronal replay during sleep, how replay changes the structure of spike bursts during sleep in adult birds, and how replay begins to emerge at the onset of song learning, just prior to the first changes in objective singing behavior.  These studies motivate the hypothesis that during sleep, bursting spontaneous discharge carries sensory information, and that sleep modifies networks via an interaction between  sensory and motor representations.  Collaborative work in humans studies have helped to highlight the breadth of the role of sleep in learning.  This has stimulated us to turn back to the animal work, to establish a more general animal model for sleep research (song perceptual learning in adult starlings).  In related work on song learning, we also investigate the representation of sensory memories and how these interact with auditory feedback (during daytime singing) to shape behavior.  Our interest in state-dependent mechanisms of learning has also stimulated EEG studies of sleep structure in birds, and motivates hypotheses regarding the evolution of vocal learning and the vertebrate forebrain.
We also study auditory perceptual mechanisms in adult animals.  Here we focus more on European starlings, that have complex songs, establish substantial auditory memories that help guide social interactions, and adapt well to laboratory procedures including operant techniques.  In one set of studies we investigate forebrain auditory structures, examining the dynamics of memory representations at the level of single cells and populations of neurons, and modeling the receptive field properties of the neurons.  In a related set of studies, we use behavioral/operant techniques to explore the information in stimuli (typically songs) that birds perceive as they learn those stimuli.  One extension of these studies is our recent demonstration that starlings can perceive complex syntactic patterns drawn from types of grammars long hypothesized by some linguists to be uniquely available to humans.
Our work is also highly collaborative.  We enjoy extensive interactions with the laboratory of Howard Nusbaum (Psychology, UofC), working towards coordinated research programs on sleep/learning, sensorimotor feedback, and syntax studies in humans and birds.  We have recently established a substantial research effort with the laboratory of Henry Abarbanel (Physics, UCSD), examining parametric estimation of Hodgkin Huxley models of single song system neurons, with the goal of creating realistic, data-driven models of single neurons and of networks of these neurons.  The work is strongly computational and will involve massively parallel computation, probably on Argonne computers.  We have continuing interactions with Zhiyi Chi (Statistics, Univ. Conn.) on spike pattern recognition and receptive field modeling.
Finally, we are also exploring the potential of the birdsong system in translational research.  We have significant progress in establishing birdsong as a model for pediatric epilepsy, in collaboration with Wim VanDrongelen and Michael Kohrman (Pediatrics and Neurology, UofC).  We are also excited by our studies of singing in the presence of delayed auditory feedback, which present the opportunity for developing birdsong as an animal model for human stuttering.