ABSTRACT
The origin of gnathostomes encompasses a major event in vertebrate evolution and development. It involved extensive anatomical reorganisation relative to the primitive conditions in earlier, jawless, vertebrates. A robust phylogeny of the earliest gnathostomes is an essential prerequisite for any research programme addressing this evolutionary episode. However, no such hypothesis yet exists, partly because the early evolution of the Chondrichthyes (sharks, rays and chimaeroids), one of the two primary divisions of living gnathostomes, is mostly unknown. This article summarises data from well preserved stethacanthids, bizarre, primitive shark-like fishes from the Lower Carboniferous (~320 mya), which are being used to reinvestigate early chondrichthyan interrelationships. These are incorporated into a database of 23 taxa and 86 cranial and postcranial characters, using non-additive binary coding. Phylogenetic analysis of these data yields equivocal results: the shortest tree (161 steps) displays an unconventional topology in which holocephalans, including chimaeroids, emerge from a paraphyletic clade of stethacanthids and symmoriids; a more conventional solution, with holocephalans as sister-group to all other chondrichthyans, is only four steps longer. Character partition tests indicate that postcranial data support the stethacanthid-holocephalan group, and that neurocranial characters alone are insufficient for resolving relationships between the majority of well preserved Palaeozoic chondrichthyans. The broad phylogenetic implications of the new tree topology are outlined, and it is noted that they conflict with recent and/ or revived, classic comparative anatomical scenarios of jaw and hyoid arch evolution. Furthermore, when non-chondrichthyan taxa are also taken into account, it becomes apparent that there is a substantial difference between molecular based and morphology/fossil record based estimates of the age of the gnathostome crown-group radiation.
