Quantitative studies of stimulus coding in first-order vibrissa afferents of rats. 2. Adaptation and coding of stimulus parameters.

Mechanosensory neurons are often classified as either rapidly adapting or slowly adapting. We examined response decay (adaptation) during constant deflection of the vibrissae with quantitative, repeatable, ad hoc measures. We found that first-order vibrissa-activated neurons of the fifth ganglion exhibit a variety of adaptation rates that appear to be distributed continuously between the rapidly and slowly adapting extremes. Also, adaptation rate is influenced markedly by stimulus magnitude. We found no evidence for a dichotomy within the more slowly adapting neurons on the basis of discharge regularity. Threshold tuning curves were used to evaluate vibration sensitivity. Both the best frequencies and 1:1 discharge thresholds for sinusoidal stimulation ranged over two orders of magnitude and were continuously distributed. First-order vibrissa-activated afferents exhibit a broad variety of response patterns to constant-velocity stimulation. The pattern of discharge varied both as a function of time during constant-velocity (ramp) deflection and as a function of stimulus velocity. Although information about the parameters of a stimulus may be conveyed by any of several features of the response pattern, it appears that few if any neurons function as "pure" encoders of any particular stimulus parameter. We examined quantitatively the relationship between discharge rate and both velocity and amplitude of vibrissa deflection with the aid of a computer-based curve-fitting procedure. We found that about half the observed rate-level functions were best described by a power function; the remainder were best fit by a logarithmic function. The parameters of the best-fitting functions varied widely and continuously, emphasizing further the diversity of coding properties of the rat's vibrissa afferents. Rate-level curves for stimulus magnitude generally exhibited saturation; some were nonmonotonic. None were described adequately by either a logarithmic function or a power function.