Gabriel Kreiman, Christof Koch, Walter Metzner, Fabrizio Gabbiani
Supported by NSF and ERC.
Top
P-type tuberous receptors in weakly electric gymnotiform fish encode changes in electric field amplitude. Single receptor afferents transmit high information rates about random amplitude modulated stimuli (Wessel et al. 1996, J. Neurophys. 75:2280-2293). We characterized the robustness of encoding and its relationship to the variability of P-receptor afferents spike trains. Random amplitude modulated electric fields were presented to curarized fish while recording extracellularly from P-receptor afferents. Reconstruction techniques were used to estimate the stimulus from the spikes. By adding random temporal jitter to the spikes, we explored the timing precision required for accurate encoding. The robustness of encoding depended on stimulus cut-off frequency (fc) and contrast. The amount of jitter required for a 50 % drop in coding fraction was 3 ms for high fc (88Hz) but it was higher than 20 ms for low fc (less than 10 Hz). Similar results were obtained by random deletions and additions of spikes. By repeating the same 'frozen' stimulus 10 times, we analyzed the time windows over which the spikes are reproducible using spike metrics (Victor et al. 1996, J.Neurophys. 76:1310-1326). Variability decreased for high cut-off frequencies but we did not observe spike alignment at the ms level in raster plots. In spite of this, the intertrial variance versus mean spike count clustered near the curves of minimum variance. Our results suggest that P-receptor afferents use an adaptive coding scheme. Accurate and robust information can be carried for low fc without spike time reliability at the ms level while at high fc the code is not robust to jittering at the 3 ms level.