Coincidence detection in auditory neurons: a possible mechanism to enhance stimulus specificity in the grassfrog.

It is still a matter of debate whether neurons in the higher central nervous system of anurans become progressively more sharply tuned to sounds that have a behavioral importance or that such coding is performed by (small) groups of neurons. The approach we have taken to investigate this matter comprises simultaneous single-unit recording using two microelectrodes in the auditory midbrain of the grassfrog. The present study deals with 96 pairs of units responding to an ensemble of natural and synthetic mating calls in which carrier frequency and pulse-repetition rate were varied. This ensemble was presented without noise and also with background noise of increasing intensity. The spike trains were analysed for correlations between their firings. In 34 pairs (35%) a functional connection, mostly common input, was present. By selecting one of the units of a pair as a trigger it was investigated which window for a coincidence analysis would result in enhanced specificity for the unit pair. Such an analysis based on a logical AND operation could be a model for the action of a neuron on which both units under study would converge, and which would then show an enhanced specificity in their response to a stimulus ensemble. It was found that in 20 pairs (21%) the logical AND operation was more selective than each of the component neurons. The largest time window for which the selectivity was found was evenly distributed over the values 8 ms, 32 ms and 128 ms, in one case selectivity was found only for a window of 2 ms. There was neither preference for selective pairs to be found for recordings with one electrode (45 cases) or dual electrodes (51), nor for independent (62) versus functionally connected (34) pairs. In some cases selectivity resulted in a preference for one specific call, in other cases it resulted in a loss of responsiveness to the masking noise effectively resulting in an enhanced signal-to-noise ratio. The analysis stresses the importance of spatiotemporal patterns of nervous activity for the representation of sounds in the auditory midbrain of anurans.