The stratigraphically earliest and the most primitive examples of vertebrate skeletal mineralization belong to lineages that are entirely extinct. Therefore, palaeontology offers a singular opportunity to address the patterns and mechanisms of evolution in the vertebrate mineralized skeleton. We test the two leading hypotheses for the emergence of the four skeletal tissue types (bone, dentine, enamel, cartilage) that define the present state of skeletal tissue diversity in vertebrates. Although primitive vertebrate skeletons demonstrate a broad range of tissues that are difficult to classify, the first hypothesis maintains that the four skeletal tissue types emerged early in vertebrate phylogeny and that the full spectrum of vertebrate skeletal tissue diversity is explained by the traditional classification system. The opposing hypothesis suggests that the early evolution of the mineralized vertebrate skeleton was a time of plasticity and that the four tissue types did not emerge until later. On the basis of a considerable, and expanding, palaeontological dataset, we track the stratigraphic and phylogenetic histories of vertebrate skeletal tissues. With a cladistic perspective, we present findings that differ substantially from long-standing models of tissue evolution. Despite a greater diversity of skeletal tissues early in vertebrate phylogeny, our synthesis finds that bone, dentine, enamel and cartilage do appear to account for the full extent of this variation and do appear to be fundamentally distinct from their first inceptions, although why a higher diversity of tissue structural grades exists within these types early in vertebrate phylogeny is a question that remains to be addressed. Citing recent evidence that presents a correlation between duplication events in secretory calcium-binding phosphoproteins (SCPPs) and the structural complexity of mineralized tissues, we suggest that the high diversity of skeletal tissues early in vertebrate phylogeny may result from a low diversity of SCPPs and a corresponding lack of constraints on the mineralization of these tissues.

Among the morphological changes that occurred during the 'fish-to-tetrapod' transition was a marked reorganization of the cranial endoskeleton. Details of this transition, including the sequence of character acquisition, have not been evident from the fossil record. Here we describe the braincase, palatoquadrate and branchial skeleton of Tiktaalik roseae, the Late Devonian sarcopterygian fish most closely related to tetrapods. Although retaining a primitive configuration in many respects, the cranial endoskeleton of T. roseae shares derived features with tetrapods such as a large basal articulation and a flat, horizontally oriented entopterygoid. Other features in T. roseae, like the short, straight hyomandibula, show morphology intermediate between the condition observed in more primitive fish and that observed in tetrapods. The combination of characters in T. roseae helps to resolve the relative timing of modifications in the cranial endoskeleton. The sequence of modifications suggests changes in head mobility and intracranial kinesis that have ramifications for the origin of vertebrate terrestriality.

We used light microscopy and scanning electron microscopy to compile a complete histological description of the dermal skeleton of the antiarch placoderm, Bothriolepis canadensis. Placodermi is most often cited as the sister group of crown group Gnathostomata, but some recent authors propose that placoderms instead represent a paraphyly of forms leading to the crown. In either phylogenetic scenario, comparative analysis of placoderm and gnathostome histological data allows us to address the primitive condition of both the gnathostome skeleton and the jawed vertebrate skeleton. The results of this work support the interpretation that the external skeleton of Bothriolepis canadensis is comprised exclusively of cellular dermal bone tissue. The unique stratification of the antiarch thoracic skeleton that has led to controversial interpretations in the past is explained by the nature of the articulations between adjacent elements. Skeletal features long thought to be gnathostome innovations are instead discovered to arise along the gnathostome stem. These innovations include secondary osteons, the systematic reconstruction of the skeleton in response to growth, and unfused, overlapping joints that enable marginal growth while maximizing the area of the articulation surface. The extensive evidence for spheritic mineralization agrees with a model of the skeleton as one capable of a high growth rate and active remodeling. Dermal skeletal development in both placoderms and osteichthyans is primarily skeletogenetic with only a minor odontogenetic contribution in some taxa. This demonstrates the problem inherent with assuming a broad application for those hypotheses of dermal skeletal evolution that are based on a chondrichthyan model. Our results highlight the importance of anatomical and ontogenetic context in the interpretation of fossil tissues.

No abstract