Endocast and Bony Labyrinth of a Devonian "Placoderm" Challenges Stem Gnathostome Phylogeny.

ABSTRACT

Our understanding of the earliest evolution of jawed vertebrates depends on a credible phylogenetic framework for the jawed stem gnathostomes collectively known as "placoderms".1 However, their relationships, and whether placoderms represent a single radiation or a paraphyletic array, remain contentious.2-13 This uncertainty is compounded by an uneven understanding of anatomy across the group, particularly of the phylogenetically informative braincase and brain cavity-endocast. Based on new tomographic data, we here describe the endocast and bony labyrinth of Brindabellaspis stensioi from the Early Devonian of New South Wales.14 The taxon was commonly recovered as branching near the base of placoderms.5-9,11,12,15-17 Previous studies of Brindabellaspis emphasized its resemblances with fossil jawless fishes in the braincase anatomy14 and endocast proportions1,18 and its distinctive features were interpreted as autapomorphies, such as the elongated premedian region.19 Although our three-dimensional data confirmed the resemblance of its endocast to those of jawless vertebrates, we discovered that the inner ear and endolymphatic complex display a repertoire of previously unrecognized characters close to modern or crown-group jawed vertebrates, including a pronounced sinus superior and a vertical duct that connects the endolymphatic sac and the labyrinth cavity. Both parsimony and Bayesian analyses suggest that prevailing hypotheses of placoderm relationships are unstable, with newly revealed anatomy pointing to a radical revision of early gnathostome evolution. Our results call into question the appropriateness of arthrodire-like placoderms as models of primitive gnathostome anatomy and raise questions of homology relating to key cranial features.