Files
Abstract
Ray-finned fishes (Actinopterygii) comprise over 32,500 extant species, represent more than half of modern vertebrate diversity, and trace their evolutionary history back at least 425 million years; however, the Paleozoic diversification of Actinopterygii is poorly understood. While molecular and morphological studies have clarified the branching order among extant actinopterygian clades, the interrelationships of early actinopterygians — especially taxa historically assigned to the paraphyletic "palaeonisciform" grade — remain unresolved. This phylogenetic challenge is associated with incomplete anatomical description of compressed specimens, historical reliance on external fossil features (e.g. scales, dermal bones) that are prone to homoplasy, and the use of outdated taxonomies. Recent advances in X-ray micro-computed tomography (µCT) imaging have unlocked the endoskeleton as a new source of phylogenetic signal. Delicate articulated structures can now be recovered from fossils, revealing internal anatomy rarely accessible through mechanical or chemical preparation. These µCT data have already challenged conventional phylogenetic hypotheses, revising the timing and membership of both stem and crown Actinopterygii, but this reshuffling has resulted in approximately 100 million years of "missing" crown actinopterygian taxa. The period of highest phylogenetic uncertainty immediately follows the Hangenberg extinction event, which is thought to have triggered the rapid proliferation of actinopterygian morphological diversity in the late Paleozoic. Scarcity of comparative endoskeletal anatomy from this interval likely contributes to the poor phylogenetic signal, which occludes the origin of modern actinopterygian biodiversity.
This dissertation aims to address this knowledge gap by describing exceptional Carboniferous specimens of five species using μCT, constructing a new morphological character set to accommodate the torrent of new μCT-derived endoskeletal data, and exploring the implications of resulting phylogenetic hypothesis to refine our understanding of the evolutionary radiation of Paleozoic Actinopterygii. I begin with a general introduction to the history of actinopterygian systematics and comparative anatomy in which I outline the major challenges hindering current research, highlight important gaps in understanding, and propose new strategies to address these issues using modern technology and integrative analyses. Across three data chapters, the project progressively broadens descriptive anatomical coverage, refines the character set, and evaluates evolutionary hypotheses – anchored at each state with comparison of homologous features in extant taxa. In Chapter 2 I introduce a new actinopterygian species from the Upper Carboniferous as part of a broader investigation into the phylogenetic signal from Paleozoic actinopterygian braincases. Comparative analysis of neurocrania reveals evolutionary trends in braincase morphology, identifies a cluster of Late Paleozoic taxa with similar features, and explores challenges inherent to the construction of morphological actinopterygian datasets. Next, in Chapter 3, I present a contrasting example: a Carboniferous braincase that is unlike the rest. Using μCT data from two specimens of Trawdenia planti, I reveal a new kind of Carboniferous braincase, demonstrate preservation of internal ventricular structures, provide evidence for a large brain snugly fitting within the neurocranium, and show that endocast morphology can capture phylogenetically and ecologically informative neuroanatomy across the actinopterygian tree. Comparison of soft tissue, endocast, and braincase morphology with that of extant and fossil taxa further suggests that Trawdenia might be a stem-Chondrostean and bridge the late Paleozoic taxon gap. In my final data chapter, I present new endoskeletal features of the suspensorium, neurocranium, gill arches, pectoral girdle, and axial column from three Carboniferous taxa with an accompanying phylogenetic analysis. I recover Trawdenia on the chondrostean stem, the majority of “palaeonisciform” taxa as stem-neopterygians, and Nematoptychius greenocki on the Actinopteran stem. This is a radical departure from prevailing hypotheses of early actinopterygian relationships, fills a major evolutionary gap, and highlights the importance of high-quality comparative data with principled character construction. Furthermore, the time-calibrated analysis infers the most likely divergence dates for the actinopterygian, actinopteran, and neopterygian crown nodes coincident with Silurian, end-Devonian, and end-Permian mass extinction events. This hypothesis recasts early actinopterygian evolution as a series of pulsed, crisis-mediated radiations associated with substantial ecological perturbation.
Overall, this dissertation demonstrates the transformative potential of integrating advanced imaging techniques with traditional paleontological approaches, proposes a new phylogenetic hypothesis, and sets the stage for future work focused on the neglected endoskeletal disparity of Paleozoic fishes.