Neural control of limb coordination

Summary

This study examines the effect of removing input descending from the brain on the production of the distinctive leg motor patterns of walking and hatching by spinal circuitry of 0- to 3-day old posthatching chicks. Transection of the cervical spinal cord was performed and chicks were tested between 2 and 28 h after surgery. Walking with good weight support could be elicited from many spinal chicks when placed on a moving treadmill belt. In some cases, sensory stimulation resulting from tail and/or wing pinch was also used. Placing spinal chicks in the hatching position in glass eggs was sufficient to elicit hatching-like leg movements in some animals. Wing pinch was used to elicit more or longer episodes of leg movements. Quantitative analyses of EMG recordings from 6 leg muscles were used to evaluate the changes in motor patterns after cervical spinal transection. Most of the characteristic features of walking and hatching are maintained after descending input from the brain is eliminated. Each muscle is activated in the double or single bursting pattern typical of the normal behavior. Characteristic phase relationships are also preserved. In addition, burst duration versus cycle period relationships seen during the normal behaviors are maintained in the spinal animals. This shows that circuitry located in the spinal cord can produce these distinctive aspects of the hatching and walking motor patterns in the absence of brain input. While many features of walking and hatching patterns were maintained in spinal animals, some changes were noted. For example, although interlimb patterns of coordination are unchanged during spinal hatching and in many sequences of spinal walking, they do become more variable during some sequences of spinal walking. This suggests that descending input from the brain normally plays a role in stabilizing interlimb alternation during walking. An increase in cycle period is seen during walking in spinal chicks. In addition, the relative durations of hip and ankle flexor bursts (expressed as percent of cycle period) increase in both behaviors. Because these changes are also seen in chicks with lumbosacral deafferentations, they may represent the response of the pattern generating circuitry to any reduction in input. In some animals, cervical spinal transection was combined with lumbosacral deafferentation so that both brain input and sensory input from the legs were removed from the spinal cord. Results from these animals were similar to those from animals with deafferentation alone in that the walking and hatching patterns converged, but did not become identical (Bekoff et al. 1987). The remaining differences may be due to the use of some unique elements of pattern generating circuitry in each behavior, or to differences in the remaining input to the same pattern generating circuitry. A possible source for such differences in input is the undeafferented neck region.