The assembly of the gnathostome bodyplan constitutes a formative episode in vertebrate evolutionary history, an interval in which the mineralized skeleton and its canonical suite of cell and tissue types originated. Fossil jawless fishes, assigned to the gnathostome stem-lineage, provide an unparalleled insight into the origin and evolution of the skeleton, hindered only by uncertainty over the phylogenetic position and evolutionary significance of key clades. Chief among these are the jawless anaspids, whose skeletal composition, a rich source of phylogenetic information, is poorly characterized. Here we survey the histology of representatives spanning anaspid diversity and infer their generalized skeletal architecture. The anaspid dermal skeleton is composed of odontodes comprising spheritic dentine and enameloid, overlying a basal layer of acellular parallel fibre bone containing an extensive shallow canal network. A recoded and revised phylogenetic analysis using equal and implied weights parsimony resolves anaspids as monophyletic, nested among stem-gnathostomes. Our results suggest the anaspid dermal skeleton is a degenerate derivative of a histologically more complex ancestral vertebrate skeleton, rather than reflecting primitive simplicity. Hypotheses that anaspids are ancestral skeletonizing lampreys, or a derived lineage of jawless vertebrates with paired fins, are rejected.

No abstract

Osteostracans are the closest jawless relatives of jawed vertebrates, informing the gradual assembly of the vertebrate mineralised skeleton. Conflicting interpretations of their dermal skeletal histology arise from failure to account for topological variation, obscuring their significance in elucidating vertebrate skeletal evolution. To resolve this, we characterize the cranial and trunk dermal skeleton of a single individual of Tremataspis mammillata (Osteostraci, Thyestiida) at submicron resolution using synchrotron‐ and computed‐ tomography. Our results show that the architecture of the Tremataspis dermal skeleton is, for the most part, conserved over the skeleton and is broadly consistent with previous histological hypotheses based on 2D thin section study. We resolve debate over the homology of the basal layer, identifying it as osteogenic acellular isopedin rather than odontogenic elasmodine or metaplastic ossification of the stratum compactum of the dermis. We find topological variation between all dermal skeletal elements studied, and particularly between the cranial and postcranial dermal skeleton. This variation can be largely explained by reduction in differentiation due to geometric constraints imposed within smaller skeletal elements, such as scales. Our description of the dermal skeleton of Tremataspis mammillata provides a foundation for interpreting data from cursory topological samples of dermal skeletal diversity obtained in other osteostracans. This reveals general aspects of histological structure that must be ancestral for osteostracans and, likely, ancestral jawed vertebrates. Finally, we draw the distinction between hypotheses and descriptions in palaeohistology.