Biomechanical determinants of transverse and rotary gallop in cursorial mammals

Transverse and rotary gallop differ in the placement of the leading hindfeet and forefeet: ipsilateral in the former gait, contralateral in the latter. We analysed 351 filmed sequences to assess the gallop type of 89 investigated mammalian species belonging to Carnivora, Artiodactyla and Perissodactyla orders. Twenty-three biometrical, ecological and physiological parameters were collected for each species both from literature data and from animal specimens. Most of the species showed only one kind of gallop: transverse (42%) or rotary (39%), while some species performed rotary gallop only at high speed (19%). In a factorial analysis, the first principal component (PC), which accounted for 40% of the total variance, was positively correlated to the relative speed and negatively correlated to size and body mass. The second PC was correlated to the ratio between distal and proximal limb segments. Large size and longer proximal limb segments were associated with transverse gallop, while rotary and speed-dependent species showed higher metacarpus/humerus and metatarsus/femur length ratio and faster relative speeds. The resulting limb excursion angles were proportional to the square-root of the Froude number, and significantly higher in rotary gallopers. The gait pattern analysis indicated significant differences between transverse and rotary gallop in forelimb and hindlimb duty factor (t-test; P