The role of the calcaneal ‘heel’ as a propulsive lever in basal archosaurs and extant monitor lizards

ABSTRACT

The structurally complex ankles of Triassic archosaurs (Reptilia: Diapsida) have been repeatedly described, and tarsal characters have played an important role in analyses of archosaur phylogeny. Morphological variations in the ankle joint undoubtedly had a major impact on locomotion, but the functional implications of many ankle features remain poorly understood. This paper investigates the function of one such structure, the prominent and highly distinctive lateral to posterolateral calcaneal process that occurs throughout basal archosaurs and is homologous to the heel-like calcaneal tuber of pseudosuchians. A morphologically analogous lateral process occurs in extant varanid lizards, and X-ray rotoscopic analysis of the limb movements of savannah monitors (Varanus exanthematicus) indicates that the lateral process rotates into a vertical orientation as the limb retracts. As the varanid lateral process approaches the vertical, it becomes increasingly effective in adding to the propulsive moment arm of m. peroneus longus. A computer model of the ankle was compared to modified versions in which the lateral process and the associated proximal expansion of the fifth metatarsal were eliminated, and this virtual experiment demonstrated that the combination of the two structures increases the peak propulsive moment arm of the varanid m. peroneus longus by over 26%. The lateral calcaneal process of basal archosaurs probably acted similarly, enhancing the ability of the peroneus musculature to contribute to propulsion during walking. Like the calcaneal tuber of pseudosuchian archosaurs, the primitive lateral process was a lever that contributed to locomotion by allowing ankle plantarflexion to generate a greater propulsive moment.

ACKNOWLEDGMENTS

This study is based on part of a Ph.D. dissertation carried out under the valuable guidance of F. A. Jenkins, Jr. I am grateful to Prof. Jenkins and to the other members of my dissertation committee, G. V. Lauder, A. W. Crompton, and C. R. Marshall, for much advice, encouragement, and constructive criticism. D. B. Baier and S. M. Gatesy provided extensive practical help with the kinematic experiments described in this paper, and introduced me to Maya as a tool for kinematic analysis. Additional thanks are due to Prof. Gatesy for providing access to his article on SR methodology in advance of publication, and for his valuable comments on an earlier draft of this paper. J. Fröbisch also read the manuscript and offered helpful suggestions, and the paper subsequently benefited from the valuable comments of two anonymous referees. Alias generously supported my work for 2 years by providing a license for the use of Maya. The savannah monitors were kindly made available by T. Owerkowicz, and their pelvic and hind limb bones were μCT-scanned by R. Fajardo. I thank S. Chapman, P. Barrett, and J. Hooker (BMNH), J. Botha and E. Butler (NM), S. Kaal and R. M. H. Smith (SAM), F. A. Jenkins, Jr., J. Rosado, and C. R. Schaff (MCZ), J. Powell (PVL) and M. Carrano (USNM) for access to specimens and hospitality during my visits to their respective institutions. This research was financed by a Chapman Fellowship, by departmental funding from Harvard University, and by the Chinese Academy of Sciences.