Long-term adaptive changes in primate vestibuloocular reflex. II. Electrophysiological observations on semicircular canal primary afferents.

I. Single-unit recordings were obtained from 551 fibers in the vestibular portion of the eighth nerve in awake rhesus monkeys. Slightly more than half of these recordings were from normal animals and the remainder from animals adapted to dove prism (leftright) reversing spectacles for 2 or more weeks. 2. Horizontal vestibuloocular reflex (VOR) gains were determined at intervals throughout the recording period (several weeks for each animal), using passive sinusoidal oscillations of the whole animal about the vertical axis at 0.2 Hz, 220”. Mean horizontal VOR gain for the normal animals was 0.94 (range, 0.91-0.98) and for the prism adapted, 0.15 (range, 0.05-0.2). The latter showed large phase lags relative to normal (mean, -93”; range, -15 to -160”). 3. On the basis of their responses to angular head oscillations and the dynamic and static phases of forward or backward head tilting, eighth nerve fibers were classified as otolith or semicircular canal afferents, and the latter assigned to one of the three canals. The proportions of the various fiber types were very similar in the normal and prism-adapted animals, and each included a small percentage of fibers that were unresponsive to any of the stimuli employed. 4. Canal fiber responses to sinusoidal angular oscillations were normalized to provide a best estimate of their sensitivity to head velocity (in spikes l Udeg l s-l) when the response plane of the canal in question coincided with the horizontal stimulus plane. 5. Nonparametric statistical tests applied to the population data for the canal afferents in normal and adapted animals failed to reveal any significant differences in their resting discharge rates or their responses to sinusoidal angular head oscillations (sensitivity or phase) with one minor exception: activity of horizontal canal units in the adapted animals was on average a few degrees more phase advanced than that in the normals. Since this difference was small and uncorrelated with the much larger phase lags seen in the vestibuloocular responses of the adapted animals, it is thought to result from biased sampling. Correlations between these measured parameters (resting discharge rate, sensitivity, phase) were also not significantly different in normal and adapted animals. 6. We conclude that the signals carried by the semicircular canal primary afferents in the monkey do not undergo any significant modification when the animal adapts to reversing prisms and the modifiable elements responsible for the observed changes in the vestibuloocular reflex must lie within the central nervous system.