A Passive Dynamic Model Predicts All Gaits of Horses

Starting in 1965, M. Hildebrand developed criteria to classify quadrupedal gaits by two numbers, the lag of footfall of ipsilateral feet and the duty factor. He compared over 150 different species of animals using this method [1, 2]. The gaits were categorized into two groups: symmetrical gaits and asymmetrical gaits, and analyzed separately. This method allows a graphical classification of gaits in a so-called gait graph, and Hildebrand results suggest that the range of possible gaits in this graph can be represented as a continuum, rather than a set of isolated regions. For horses, for example, the regions in the gait graph form a three pronged figure. While certain components of this shape, have been explained [3] , the reasons for its exact shape remain unclear. Why do animals use only certain footfall patterns and avoid others? Moreover, does there exist a gradual transition from one gait to another or are different gaits just overlapping regions in the gait graph.