Vertebrate bone is composed of three main cell types: osteoblasts, osteoclasts and osteocytes, the latter being by far the most numerous. Osteocytes are thought to play a fundamental role in bone physiology and homeostasis, however they are entirely absent in most extant species of teleosts, a group that comprises the vast majority of bony 'fishes', and approximately half of vertebrates. Understanding how this acellular (anosteocytic) bone appeared and was maintained in such an important vertebrate group has important implications for our understanding of the function and evolution of osteocytes. Nevertheless, although it is clear that cellular bone is ancestral for teleosts, it has not been clear in which specific subgroup the osteocytes were lost. This review aims to clarify the phylogenetic distribution of cellular and acellular bone in teleosts, to identify its precise origin, reversals to cellularity, and their implications. We surveyed the bone type for more than 600 fossil and extant ray-finned fish species and optimised the results on recent large-scale molecular phylogenetic trees, estimating ancestral states. We find that acellular bone is a probable synapomorphy of Euteleostei, a group uniting approximately two-thirds of teleost species. We also confirm homoplasy in these traits: acellular bone occurs in some non-euteleosts (although rarely), and cellular bone was reacquired several times independently within euteleosts, in salmons and relatives, tunas and the opah (Lampris sp.). The occurrence of peculiar ecological (e.g. anadromous migration) and physiological (e.g. red-muscle endothermy) strategies in these lineages might explain the reacquisition of osteocytes. Our review supports that the main contribution of osteocytes in teleost bone is to mineral homeostasis (via osteocytic osteolysis) and not to strain detection or bone remodelling, helping to clarify their role in bone physiology.

Osteocytes, cells embedded within the bone mineral matrix, inform on key aspects of vertebrate biology. In particular, a relationship between volumes of the osteocytes and bone growth and/or genome size has been proposed for several tetrapod lineages. However, the variation in osteocyte volume across different scales is poorly characterized and mostly relies on incomplete, two‐dimensional information. In this study, we characterize the variation of osteocyte volumes in ray‐finned fishes (Actinopterygii), a clade including more than half of modern vertebrate species in which osteocyte biology is poorly known. We use X‐ray synchrotron micro‐computed tomography (SRµCT) to achieve a three‐dimensional visualization of osteocyte lacunae and direct measurement of their size (volumes). Our specimen sample is designed to characterize variation in osteocyte lacuna morphology at three scales: within a bone, among the bones of one individual and among species. At the intra‐bone scale, we find that osteocyte lacunae vary noticeably in size between zones of organized and woven bone (being up to six times larger in woven bone), and across cyclical bone deposition. This is probably explained by differences in bone deposition rate, with larger osteocyte lacunae contained in bone that deposits faster. Osteocyte lacuna volumes vary 3.5‐fold among the bones of an individual, and this cannot readily be explained by variation in bone growth rate or other currently observable factors. Finally, we find that genome size provides the best explanation of variation in osteocyte lacuna volume among species: actinopterygian taxa with larger genomes (polyploid taxa in particular) have larger osteocyte lacunae (with a ninefold variation in median osteocyte volume being measured). Our findings corroborate previous two‐dimensional studies in tetrapods that also observed similar patterns of intra‐individual variation and found a correlation with genome size. This opens new perspectives for further studies on bone evolution, physiology and palaeogenomics in actinopterygians, and vertebrates as a whole.

Teleost fishes comprise one-half of all vertebrate species and possess a duplicated genome. This whole-genome duplication (WGD) occurred on the teleost stem lineage in an ancient common ancestor of all living teleosts and is hypothesized as a trigger of their exceptional evolutionary radiation. Genomic and phylogenetic data indicate that WGD occurred in the Mesozoic after the divergence of teleosts from their closest living relatives but before the origin of the extant teleost groups. However, these approaches cannot pinpoint WGD among the many extinct groups that populate this 50- to 100-million-y lineage, preventing tests of the evolutionary effects of WGD. We infer patterns of genome size evolution in fossil stem-group teleosts using high-resolution synchrotron X-ray tomography to measure the bone cell volumes, which correlate with genome size in living species. Our findings indicate that WGD occurred very early on the teleost stem lineage and that all extinct stem-group teleosts known so far possessed duplicated genomes. WGD therefore predates both the origin of proposed key innovations of the teleost skeleton and the onset of substantial morphological diversification in the clade. Moreover, the early occurrence of WGD allowed considerable time for postduplication reorganization prior to the origin of the teleost crown group. This suggests at most an indirect link between WGD and evolutionary success, with broad implications for the relationship between genomic architecture and large-scale evolutionary patterns in the vertebrate Tree of Life.

Salamanders are an important group of living amphibians and model organisms for understanding locomotion, development, regeneration, feeding, and toxicity in tetrapods. However, their origin and early radiation remain poorly understood, with early fossil stem-salamanders so far represented by larval or incompletely known taxa. This poor record also limits understanding of the origin of Lissamphibia (i.e., frogs, salamanders, and caecilians). We report fossils from the Middle Jurassic of Scotland representing almost the entire skeleton of the enigmatic stem-salamander Marmorerpeton . We use computed tomography to visualize high-resolution three-dimensional anatomy, describing morphologies that were poorly characterized in early salamanders, including the braincase, scapulocoracoid, and lower jaw. We use these data in the context of a phylogenetic analysis intended to resolve the relationships of early and stem-salamanders, including representation of important outgroups alongside data from high-resolution imaging of extant species. Marmorerpeton is united with Karaurus , Kokartus , and others from the Middle Jurassic–Lower Cretaceous of Asia, providing evidence for an early radiation of robustly built neotenous stem-salamanders. These taxa display morphological specializations similar to the extant cryptobranchid “giant” salamanders. Our analysis also demonstrates stem-group affinities for a larger sample of Jurassic species than previously recognized, highlighting an unappreciated diversity of stem-salamanders and cautioning against the use of single species (e.g., Karaurus) as exemplars for stem-salamander anatomy. These phylogenetic findings, combined with knowledge of the near-complete skeletal anatomy of Mamorerpeton, advance our understanding of evolutionary changes on the salamander stem-lineage and provide important data on early salamanders and the origins of Batrachia and Lissamphibia.