Motor modularity in voluntary frog behaviors and its neural basis

Tanuj Gulati

doi:10.17918/00000829

Modular control architectures in spinal cord allow rapid complex spinal motor output despite musculoskeletal redundancy and neural complexity. Understanding this modularity better would advance knowledge of motor control and is relevant to rehabilitation strategies. Modular organization is explored in voluntary frog behaviors in this study. Motor primitives/ modules are here defined and explored as very precise compositional elements: physiological pulsatile premotor drives in the spinal cord that activate fixed muscle weightings. Modularity was identified in voluntary bull frog behaviors of prey-strike and avoidance jumps by using independent components analysis (ICA). Twenty six muscles' electromyograms from either-side hindlimb and trunk muscles underwent ICA. In a unilateral analysis, seven components captured approximately 85% of variance in the thirteen left-side muscles, while six components did so on the remaining right-side muscles. In bilateral analysis of all 26 electromyograms together, components and premotor drive pulses were found to be local-to-a-leg and to have structurally similar weight structure on the left- and right-hand side of the spinal cord. Using cosine packet analyses to examine the pulse structure of these drives, we found narrower time scales between 60-200 msec in intact and decerebrate frogs compared to the more stereotyped 275 msec bursts seen in spinal frogs. In order to understand the basis for 'local-to-a-leg' primitives in the 'hopping' frog, we assessed component activation peak amplitudes and lags in left- and right-components in the turning lunges (p<0.05, MANOVA) and found some components were significantly different in amplitude or lag in right or left jumps while they were not significantly different in straight jumps. Additionally, using ground reaction forces, we observed that R/L lags in time to peak between lateral and rostral force (translational) force components were significantly greater for turning lunges, whereas these lags were ~0 in straight jumps (p<0.05, ANOVA) and this difference was manifested early in development. Finally, we also recorded neural activity in the spinal cord of lunging and turning frogs. Peri-stimulus time histograms and analysis of inner product projection measures showed that neural activity largely matches spinal motor primitive patterns in these behaviors and thus, there is a neural implementation of discrete modules.

Motor modularity in voluntary frog behaviors and its neural basis

Files and links (1)

Abstract

Metrics

Details

Motor modularity in voluntary frog behaviors and its neural basis

Files and links (1)

Abstract

Metrics

Details

Drexel University Social media