On the origin of the sauropods and the classification of the Saurischia

Abstract

Existing classifications of the Saurischia are briefly reviewed. The sauropods of the later Mesozoic are generally considered to have originated from the prosauropods of the Upper Trias–the Thecodontosauridae, Plateosauridae and Melanorosauridae. The Melanoro-sauridae have been rightly recognized as the family which is most like the sauropods, and have therefore been considered transitional between the better known Thecodontosauridae and Plateosauridae on the one hand and the sauropods on the other. Re-examination of melanorosaurid material in South Africa shows that none is properly associated and that the proportions cannot be determined. Information is given on undes-cribed skeletons and trackways, all from the lower part of the Stormberg Series; the most interesting are the quadrupedal ‘Soebeng trackways’, just like those of Jurassic sauropods. The Stormberg sauropodomorphs are listed. Present concepts of sauropodomorph phylogeny, embodying the evolutionary series pseudosuchian → thecodontosaurid → plateosaurid → melanorosaurid → sauropod, are incompatible with the stratigraphical position of the Melanorosauridae below the Thecodontosauridae (in South Africa). There is no evidence to suggest that the sauropod line of evolution was other than purely quadrupedal, i.e. the sauropods were not quadrupedal reversions from bipedal forbears. The various prosauropods were partly bipedal offshoots from this main quadrupedal line, undergoing adaptive radiation and extinction in the Trias. The similarity of their teeth to those of sauropods is only partial and superficial, and probably has no phylogenetic significance. Triassic carnosaurs, hitherto placed in the Palaeosauridae and Teratosauridae, differ considerably from the later carnosaurs; it is only the nature of their teeth (where known) which has led to their classification as carnosaurs, the post-cranial skeletons of palaeo-saurids and teratosaurids being virtually indistinguishable from those of the contemporary thecodontosaurids and plateosaurids respectively. Only because these Triassic forms were so like the prosauropods was it thought that all Carnosauria and the Sauropodomorpha were of common origin. But in the Teratosauridae there is no proven association of teeth with post-cranial skeleton, and in the Palaeosauridae (here called Gryponychidae) the teeth are unknown; further, even if these prosauropod-like skeletons are correctly associated with carnosaur-like teeth, it could mean no more than that the prosauropods included both carnivores and herbivores. Some post-cranial ‘teratosaurid’ material from China and England is referred to the Melanorosauridae, greatly extending the geographical and strati-graphical ranges of that family. The Gryponychidae and all other supposedly teratosaurid post-cranial remains are also transferred from the Carnosauria to the Prosauropoda. The respective stratigraphical-geographical ranges confirm that it is not easy to separate either the Gryponychidae from the Thecodontosauridae or the post-cranial remains of ‘Teratosauridae’ from the Plateosauridae. The members of the first-named family in each pair are therefore transferred to the second. The primary sub-division of the Saurischia is into the sub-orders Sauropodomorpha and Theropoda; the former is further divided into the infra-orders Prosauropoda and Sauropoda, the latter into the Coelurosauria and Carnosauria. This new classification is formally summarized. The Saurischia originated, probably polyphyletically, from large quadrupedal Thecodontia. In the late Middle and early Upper Trias the Saurischia underwent a great radiation into a variety of forms, both quadrupedal and bipedal. Their distribution by families is briefly discussed.