Assessing locomotor behavior is a standard methodology to characterize the behavioral phenotype of a genetic manipulation. Typically, levels of locomotor activity are measured using various methods that are based on the frequency of photobeam breaks or distance traveled as assessed by video-tracking systems. Locomotor behavior, however, is multi-dimensional and reflects the combined influences of multiple processes. Here, we examine the number of independent dimensions of locomotor behavior in mice based on measures derived from a video-tracking system. In addition, we test the hypothesis that locomotor behavior varies substantially across mouse strains. 84 mice were tested for 30 min in a 41 x 41 cm enclosure. Based on previous investigations in rats, we also assessed the spatial and dynamical aspects of locomotor behavior using the spatial scaling exponent, d, and the dynamical entropy, h. A principal component analysis and a one-way repeated measure ANOVA were conducted. C57 mouse strains differ substantially from 129 mouse strains on almost all measures of locomotor behavior. The principal component analysis revealed that two independent factors influence this set of measures. The first factor reflects the amount or level of locomotor activity, the second factor quantifies the degree of spatial and dynamical organization of behavior. These strain differences and the existence of at least two independent dimensions when measuring locomotor behavior may help to parse the effects of gene manipulations relative to strain differences in mutant mice.
Copyright 1999 Elsevier Science B.V.