Through computer feedback control, muscle-unit tension was maintained by altering the stimulation rate of a functionally isolated motor axon. The required stimulation patterns and fatigue properties of motor units from the flexor carpi radialis (FCR), flexor digitorum longus (FDL), and medial gastrocnemius (MG) muscles of the cat were studied when tension was maintained or "clamped" at a constant average level (25% of maximum tetanic tension). 2. In each muscle, two distinct stimulation patterns were observed during constant-tension contractions, one associated with slow-twitch units and the other with fast-twitch units. Once target tension was reached, slow-twitch units required fairly constant rates in order to maintain a constant force, whereas fast-twitch units displayed a marked decline in rate during the early phases of the contraction, averaging between 42 and 54% for the three muscles. The decline in rate most likely represented potentiation of the contractile response and slowing of contractile speed. In general, slow-twitch units responded with lower mean rates (approximately 14 pps less), averaged over the course of the contraction, than fast-twitch units. 3. For fast-twitch units of each muscle, resistance to fatigue varied continuously and over a wide range. The duration that tension could be maintained at 25% of maximum, defined as endurance time, ranged between 16 and 2063 s. No categorization of fast-twitch units into groups could be made on the basis of endurance time. Of the 5 slow-twitch units followed beyond 2700 s, only one failed to maintain tension during the observation period. 4. For hindlimb fast-twitch units, endurance was independent of the stimulation rate needed to maintain tension during the contraction. By contrast, there was a significant tendency for an inverse relation between endurance time and mean stimulation rate for FCR fast-twitch units. 5. Recovery of maximum tension was evaluated at 30 s, 1 min, 2 min, and 5 min following a constant-tension contraction. After a 5-min rest, fast-twitch units were able to produce an average of 80-85% of their maximum tetanic tension. By using the median endurance time (approximately 100 s) to divide the fast-twitch population into "low" and "high" endurance groups, recovery of tension was found not to be uniform among the two groups. High endurance units were able to recover a greater percentage of their original maximum tetanic tension. No difference was found between force recovery for low and high endurance units at 30 s.(ABSTRACT TRUNCATED AT 400 WORDS)
