Twenty-seven new specimens recovered from lower levels of the Sarmiento Formation (Casamayoran SALMA, middle Eocene) in the 'Cañad on Hondo' area, Chubut (Argentina), bring new information on the Palaeogene long-necked chelid Hydromedusa casamayorensis. The discovery of cranial and postcranial remains allows a complete description of the only known extinct species of the genus Hydromedusa recognizing diagnostic characters that revalidate the taxonomic status of this species. Hydromedusa casamayorensis is incorporated into a cladistic framework of chelid turtles. The unconstrained and constrained phylogenetic analyses recovered H. casamayorensis as a sister taxon of Hydromedusa maximiliani plus H. tectifera. A Danian age is proposed at least for the origin of the Hydromedusa clade, while a Late Cretaceous age is suggested for the split from the Yaminuechelys clade. Great differences in size are documented for H. casamayorensis based on the various specimens recovered from this study. The osteohistological study in H. casamayorensis and H. tectifera shows similarities in the internal cortex and cancellous bone of both taxa, but some differences are observed in the external cortex of H. casamayorensis (e.g. a parallel fibred bone on the more external surface of this cortex).

The study of bone microstructure of fossil vertebrates (i.e., paleohistology) has demonstrated to be a very important source of paleobiological information. Paleohistological studies are based on the standardized analysis of petrographic thin sections. Although the development of new technologies (e.g., microtomography) have provided non-destructive procedures for the study the fossil tissues, thin sections are still the main source of information in paleohistology. In this contribution, we provide a detailed protocol for sampling and thin-sectioning preparation of bone tissue from both fossil and extant vertebrates. We describe the most common procedures for sampling and also some particularities related to variations in equipment and sampling techniques. The main goal of this contribution is to offer an alternative protocol for research teams of recent formation and/or with limited funding.

No abstract