Representation of a Species-Specific Vocalization in the Primary Auditory Cortex of the Common Marmoset: Temporal and Spectral Characteristics. Wang, Xiaoqin, Michael M. Merzenich, Ralph Beitel and Christoph E. Schreiner. Coleman Laboratory and W.M.Keck Center for Integrative Neuroscience, University of California at San Francisco, Box 0732, San Francisco, CA 94143- 0732, U.S.A.
APStracts 2:0220N, 1995.
SUMMARY AND CONCLUSIONS
(1) The temporal and spectral characteristics of neural representations of a behaviorally important species-specific vocalization were studied in neuronal populations of the primary auditory cortex (A1) of barbiturate anesthetized adult common marmosets (Callithrix jacchus), using both natural and synthetic vocalizations. The natural vocalizations used in electrophysiological experiments were recorded from the animals under study, or from their conspecifics. These calls were frequently produced in vocal exchanges between members of our marmoset colony and are part of the well-defined and highly stereotyped vocal repertoire of this species. (2) Spectrotemporal discharge pattern of spatially distributed neuron populations in cortical field A1 was found to be correlated with spectrotemporal acoustic pattern of a complex natural vocalization. However, the A1 discharge pattern was not a faithful replication of the acoustic parameters of a vocalization stimulus, but had been transformed into a more abstract representation than that in the auditory periphery. (3) Subpopulations of A1 neurons were found to respond selectively to natural vocalizations as compared with synthetic variations that had the same spectral but different temporal characteristic. A subpopulation responding selectively to a given monkey's call shared some but not all of its neuronal memberships with other individual-call-specific neuronal subpopulations. (4) In the time domain, responses of individual A1 units were phase-locked to the envelope of a portion of a complex vocalization, which was centered around a unit's characteristic frequency (CF). As a whole, discharges of A1 neuronal populations were phase-locked to discrete stimulus events but not to their rapidly-changing spectral contents. The consequence was a reduction in temporal complexity and an increase in cross-population response synchronization. (5) In the frequency domain, major features of the stimulus spectrum were reflected in rate-CF profiles. The spectral features of a natural call were equally or more strongly represented by a subpopulation of A1 neurons that responded selectively to that call, as compared with the entire responding A1 population. (6) Neuronal responses to a complex call were distributed very widely across cortical field A1. At the same time, the responses evoked by a vocalization scattered in discrete cortical patches were strongly synchronized to stimulus events and to each other. As a result, at any given time during the course of a vocalization, a coherent representation of the integrated spectrotemporal characteristics of a particular vocalization was present in a specific neuronal population. (7) These results suggest that the representation of behaviorally important and spectrotemporally complex species-specific vocalizations in A1 is a) temporally integrated and b) spectrally distributed in nature, and the representation is carried by spatially dispersed and synchronized cortical cell assemblies that correspond to each individual's vocalizations in a specific and abstracted way. 

Received 30 January 1995; accepted in final form 24 July 1995.
APS Manuscript Number J65-5.
Article publication pending J. Neurophysiol.
ISSN 1080-4757 Copyright 1995 The American Physiological Society.
Published in APStracts on 10 August 1995.