Study of memory traces in higher animals requires experimental models possessing well-localized and technically accessible memory traces--plasticity responsible for behavioral change, not dependent on control from elsewhere, and open to detailed investigation. Our purpose has been to develop such a model based on the wholly spinal, largely monosynaptic path of the spinal stretch reflex. Previous studies described operant conditioning of this reflex and of its electrical analog, the H-reflex. In this study, we sought to determine whether conditioning causes changes in the spinal cord that affect the reflex and are not dependent on continued supraspinal influence, and thus qualify as memory traces. 2. Sixteen monkeys underwent chronic conditioning of the triceps surae H-reflex. Eight were rewarded for increasing H-reflex amplitude (HR increases mode), and eight were rewarded for decreasing it (HR decreases mode). In each animal, the other leg was an internal control. Over several months of conditioning, H-reflex amplitude in the conditioned leg rose in HR increases animals and fell in HR decreases animals. H-reflex amplitude in the control leg changed little. 3. After HR increases or HR decreases conditioning, each animal was deeply anesthetized and surgically prepared. The reflex response to supramaximal dorsal root stimulation was measured from the triceps surae nerve as percent of response to supramaximal ventral root stimulation, which was the maximum possible response. Data from both legs were collected before and for up to 3 days after thoracic (T9-10) cord transection. The animal remained deeply anesthetized throughout and was killed by overdose. 4. The reflex asymmetries produced by conditioning were still present several days after transection removed supraspinal influence: reflexes of HR increases animals were significantly larger in HR increases legs than in control legs and reflexes of HR decreases animals were significantly smaller in HR decreases legs than in control legs. 5. Reflex amplitude was much greater in the control legs of anesthetized HR decreases animals than in the control legs of anesthetized HR increases animals. 6. Chronic conditioning had at least two effects on the spinal cord. The first effect, task-appropriate reflex asymmetry, was evident both in the awake behaving animal and in the anesthetized transected animal. The second effect, larger control leg reflexes in HR decreases than in HR increases animals, was evident only in the anesthetized animal. By removing supraspinal control, anesthesia and transection revealed a previously hidden effect of conditioning.(ABSTRACT TRUNCATED AT 400 WORDS)
