Kinematics of treadmill galloping by cats: II. Steady-state coordination under positive reinforcement control

Abstract

This report, the second of three, describes treadmill galloping by cats that were trained solely by food reward. Four young adult animals provided data at various velocities from 2.3 to 4.1 m/sec. Another cat galloped at 2.0 m/sec. The overall speed range was lower than in the first study, probably because the animals were not punished for failure to run and, also, because they were slightly smaller and younger cats. At any one velocity, several successive strides were available whose footfall pattern was steady state, or invariant. The same three galloping patterns were seen that had been found previously in an aversive test situation: rotatory galloping, transverse galloping, and the half bound. The rotatory gallop was most frequently used. In general, the findings for movements of a single limb, and also the interlimb timings (with some exceptions during slow transverse galloping), were similar to those seen in the first study, as were such overall measures as stride length and the percentage of zero support. For one cat that provided a substantial number of test runs at increments between 3.1 and 4.1 m/sec, it was possible to show a linear trend of stride duration as a function of velocity, although the total change was only approximately 30 msec. The low-speed transverse gallop was of special interest, because its footfall patterns showed that a continuum of interlimb coordination could occur from trotting to this type of gallop. It was concluded that the overall similarities of food-reinforced galloping and galloping under aversive control far outweighed the differences, to imply economy of their neural control.