Archosaur ankles and the relationships of the thecodontian and dinosaurian reptiles

Abstract

Several attempts have been made to outline the early evolution and radiation of the archosaurs through a study of the morphology and function of their ankle joints^1–6 (Figs 1a, 2). Here we present new interpretations of the evidence which show that several aspects of archosaur ankle anatomy have been misunderstood, and we re-state the case for the ‘advanced mesotarsal’ (AM) ankle of the dinosaurs being derived from an intermediate stage and not directly from the ‘primitive mesotarsal’ ankle of a proterosuchid thecodontian (Fig. 1b). Two modifications to the original scheme² are required. First, that Euparkeria-like ankles are basically ‘proterosuchian’, digitigrade and unlikely to be ancestral to any other form; the term ‘modified primitive mesotarsal’ is proposed to describe this pattern of ankle. The second modification is to recognize that no ‘advanced mesotarsal (reversed)’ ankle² has been described to date, and that previous allocations of forms to this type have been in error. Therefore, all forms with the AM ankle (Fig. 2) probably share a common ancestry. The ‘advanced mesotarsal (normal)’ ankle seems to be present in all known dinosaurs. The ‘crocodile reversed’ ankle is at present known only in the Ornithosuchidae². This supports earlier suggestions that the dinosaurs are a monophyletic group^7–9, a view not generally accepted⁵. The character state used here to define the dinosaurs is acquisition of the ‘fully improved’ gait, exemplified by the AM ankle¹⁰.