© 1996 by The Society for Integrative and Comparative Biology
The Use of Gait Transition Speed in Comparative Studies of Fish Locomotion1
Museum of Comparative Zoology, Harvard University Cambridge, Massachusetts 02138
Physiological and biomechanical inquiries into the principles of vertebrate locomotion require comparison among animals of different size, habitat and phyletic association. In designing comparative studies of locomotion, a major challenge is to isolate the effects of experimentally imposed variation from the confounding effects of variation in animal activity level associated with differences in scale and life history. For swimming vertebrates, traditional measures of speed used for comparison, including sprint speed and critical swimming speed should in theory each elicit similar efforts from different animals but have practical shortcomings that can limit their usefulness. This paper presents an alternative approach, adapted from the work of mammalian physiologists, which controls for differences in relative activity level among swimming animals of different size and habitat through comparison at gait transition speeds. The method is illustrated with examples from study of the teleost fish family Embiotocidae, whose members exhibit a distinct transition from exclusively pectoral fin oscillation to combined pectoral and caudal fin propulsion with increasing swimming speed. The pectoral-caudal gait transition speed, or any percentage thereof, is shown to be 'biomechanically equivalent for swimmers of different size. When this performance limit is expressed in terms of body lengths traveled per unit time, a common normalization of swimming speed, it varies markedly across size and habitat within the family. This finding has the important implication that length specific speeds may not induce comparable degrees of exercise from different fishes, and thus kinematic and physiological comparisons at such speeds can yield misleading results. The comparative approach described for pectoral fin swimmers, and the limitations of length-specific speed, should be generally applicable to studies of other swimming vertebrates.