Kinematics of treadmill galloping by cats: I. Steady-state coordination under aversive control

Abstract

A series of three studies was made of cats galloping on a treadmill. The purpose of the work was to furnish information on kinematic time constraints that could be used for biomechanical and neurophysiological interpretations. The data for this first report were obtained from nine cats trained to avoid shock or an air jet. They galloped at a constant velocity for a number of successive strides (cycles of movement by the four limbs). Cinematographic measurements were made and normalized of test runs in which footfall order was invariant, at speeds from 2.8 to 6.1 m/sec. Three previously defined types of galloping were seen at all speeds: the rotatory gallop, the transverse gallop and the half bound. Stride length increased markedly and linearly as forward velocity increased, and percentage of zero support (there was one flight phase per stride) increased slightly, but all other measures were quite insensitive to velocity changes. Movements of individual limbs, as measured by duty factor (stance or down time as a percentage of stride duration) varied little across the four limbs or three types of gallop. Averaged interlimb measures showed that across gallop types the forelimb touchdowns were always separated by approximately 25% of stride duration, but the hindlimbs could be separated by 11% (rotatory and transverse gallops) or else could touch down simultaneously (half bound). Overall, the most invariant hindlimb-to-forelimb interval was from the trailing (rear) hindlimb to the trailing forelimb. No evidence was gained to refute the assumption that selection of one of three types of gallop on the treadmill was arbitrary, at least for avoidance running.