Quantitative analysis of relative contribution of central and peripheral neurons to gill-withdrawal reflex in Aplysia californica.

1. There is general agreement that the gill-withdrawal reflex elicited by weak tactile stimuli (less than 2 g to the siphon skin) is mediated almost entirely by the central nervous system (CNS) (13, 15, 18, 23). However, there was disagreement concerning the effects of moderate intensity (2--4 g) stimuli. Kupfermann et al. (18) found that the CNS mediates approximately 94% of the reflex elicited by moderate-intensity stimuli, whereas Peretz et al. (23) found that in this stimulus range, the amplitude of the reflex was, on average, unaltered when the CNS was removed. 2. To resolve this difference we first carried out pilot experiments in collaboration with B. Peretz, J. W. Jacklet, and K. Lukowiak using isolated mantle preparations, and then performed a systematic study using both the isolated mantle and intact animals. The main difference that seemed to account for the discrepancy in the results was the magnitude of the reflex response that was selected for study. Previous studies from this laboratory used a minimum-response criterion whereby only brisk and clearly observable responses of at least 35% maximum were examined. By contrast, Peretz et al. (23) examined all responses, even those that were extremely small. In addition, the two groups used different methods of stimulation so that stimulus intensities could not really be compared. 3. By comparing the effects of moderate-intensity stimuli in experiments with and without a response criterion in isolated mantle preparations, we found that when a minimal response-amplitude criterion is imposed the CNS mediates 90--95% of the gill-withdrawal reflex, whether it is elicited by the "tapper" stimulus used by Peretz et al. (23) or by the servo-controlled probe previously used in this laboratory. On the other hand, when no minimal response criterion is used and small responses are also examined, the response to the probe is still significantly reduced by 85% when the CNS is removed, whereas the reflex response to the tapper is more variable, sometimes increasing and other times decreasing with deganglionation. 4. We have also tested, in intact animals, the role of the CNS in mediating gill withdrawal. Water-jet stimuli were delivered to the siphon to elicit the reflex, as in previous behavioral studies. As was the case with probe stimulation in the isolated mantle preparation, in intact animals the CNS mediates 90% of the reflex evoked by moderate-intensity stimuli when a minimal response-amplitude criterion is imposed. 5. Our experiments indicate that, using response criteria and methods of stimulation, one can reliably study the reflex as predominantly mediated by the CNS.