Temporal Patterning of Grooming in Three Lines of Mice: Some Factors Influencing Control Levels of a Complex Behaviour

Abstract

AbstractTemporal patterning of limb movements during grooming in mice was studied for the purpose of analyzing mechanisms of control of behaviour. Patterns of limb movement were determined from frame-by-frame analysis of grooming film taken at 32 or 64 frames per second under several different test situations. Grooming was filmed while mice were isolated in a 5 X 5 X 18 cm. chamber. The mice were (1) placed directly into the chamber from the home cage, or (2) put into the chamber after a one-minute forced swim, or (3) observed in the chamber after unilateral attachment of a weight to the right forelimb under light ether anaesthesia. Grooming was also filmed while mice were exploring a double T-maze. The cerebellar mutant mouse nr (nervous), and control littermates from three lines (inbred BALB, F2 (BALB X DBA)-hybrid, and the backcross of the F2-hybrid to BALB) were used as subjects. The effect of presence of the nr gene, strain differences, and test situation on grooming patterns was evaluated from film analysis. Parameters studied were: (1) trajectories of individual face-grooming limb movements, (2) sequential linkage of individual grooming movements, as measured by information analysis, (3) relative integrity and (4) patterns of connection of "unit patterns" consisting of observer-defined temporal clusters of individual face-grooming movements, and (5) the relative connectivity of face-grooming to body-grooming. Parameter (1) was evaluated only for the effect of the nr gene. Test situation had important effects on all other levels of control of face-grooming. Strain differences were most evident in parameter (5): hybrid mice linked face- and body-grooming together more often than did BALB mice. The nr gene had significant effects on individual stroke trajectories, and on the stability of unit patterns during grooming with a weighted limb. Stroke trajectories were more irregular, and pattern stability was weakened, in nr mice. The results of film study preclude analysis of control of face-grooming temporal patterns in terms of a first-order Markovian model. Also, neither a simple "reflex" peripheral control model nor a pure "central motor program" model are satisfactory explanations for control of temporal patterns of grooming. Peripheral feedback interacts in some unknown fashion with levels of organization, such as unit patterns, that are determined without direct reference to movement-to-movement linkages.