Energy absorption during running by leg muscles in a cockroach.

Biologists have traditionally focused on a muscle's ability to generate power. By determining muscle length, strain and activation pattern in the cockroach Blaberus discoidalis, we discovered leg extensor muscles that operate as active dampers that only absorb energy during running. Data from running animals were compared with measurements of force and power production of isolated muscles studied over a range of stimulus conditions and muscle length changes.We studied the trochanter-femoral extensor muscles 137 and 179, homologous leg muscles of the mesothoracic and metathoracic legs, respectively. Because each of these muscles is innervated by a single excitatory motor axon, the activation pattern of the muscle could be defined precisely. Work loop studies using sinusoidal strains at 8 Hz showed these trochanter-femoral extensor muscles to be quite capable actuators, able to generate a maximum of 19-25 W kg-1 (at 25 degreesC). The optimal conditions for power output were four stimuli per cycle (interstimulus interval 11 ms), a strain of approximately 4 %, and a stimulation phase such that the onset of the stimulus burst came approximately half-way through the lengthening phase of the cycle. High-speed video analysis indicated that the actual muscle strain during running was 12 % in the mesothoracic muscles and 16 % in the metathoracic ones. Myographic recordings during running showed on average 3-4 muscle action potentials per cycle, with the timing of the action potentials such that the burst usually began shortly after the onset of shortening. Imposing upon the muscle in vitro the strain, stimulus number and stimulus phase characteristic of running generated work loops in which energy was absorbed (-25 W kg-1) rather than produced. Simulations exploring a wide parameter space revealed that the dominant parameter that determines function during running is the magnitude of strain. Strains required for the maximum power output by the trochanter-femoral extensor muscles simply do not occur during constant, average-speed running. Joint angle ranges of the coxa-trochanter-femur joint during running were 3-4 times greater than the changes necessary to produce maximum power output. None of the simulated patterns of stimulation or phase resulted in power production when strain magnitude was greater than 5 %. The trochanter-femoral extensor muscles 137/179 of a cockroach running at its preferred speed of 20 cm s-1 do not operate under conditions which maximize either power output or efficiency. In vitro measurements, however, demonstrate that these muscles absorb energy, probably to provide control of leg flexion and to aid in its reversal.