The first detailed description of the fossil acanthomorph teleost †Heteronectes chaneti is presented based on a single acid-transfer prepared specimen from the lower Eocene (Ypresian, NP 14) of Bolca, Italy. This genus shows striking cranial asymmetry, with the left orbit displaced toward the top of the skull but failing to reach the dorsal midline. Outside of the orbital region, most paired cranial bones show no clear asymmetry. The postcranial skeleton of †Heteronectes is comparable to that of many ‘generalized’ percomorphs, with a vertebral column comprising 10 abdominal and 14 caudal vertebrae, and a caudal fin with a procurrent spur and 17 principal rays in a I-8/7-I pattern, a caudal-fin endoskeleton consisting of a parphyural, five hypurals, three epurals, and two uroneurals, and pelvic fins with one spine followed by five rays. Evidence for the phylogenetic position of †Heteronectes is reviewed, and a previous interpretation of this genus as a stem member of the percomorph clade Pleuronectiformes is corroborated. However, †Heteronectes retains many primitive percomorph features transformed in other flatfishes, indicating that orbital migration was among the earliest of the distinctive pleuronectiform features to arise. Comparisons between †Heteronectes and candidates for the living sister group of flatfishes reveal the greatest number of similarities with latids (Nile perches and allies). These characters, most of which are of ambiguous polarity, should be the target of future analyses to determine whether they represent symplesiomorphies, homoplasies, or derived features uniting flatfishes and latids.

A full description of a complete and articulated, three-dimensionally preserved, placoderm fish, Materpiscis attenboroughi (Ptyctodontida), from the Late Devonian Gogo Formation is presented. The jaw articulation is unique in being preserved in situ, providing new morphological information on the articulation of the quadrate, metapterygoid, and labial skeleton. In addition, separate hyoid and opercular cartilages are identified, thus confirming their presence for the first time within placoderms. Preserved orbital ossifications indicate an anteriomesial orientation of this braincase region. The complete scapulocoracoid provides additional morphological information on the pectoral fin articulation. A new specimen of Austroptyctodus gardineri preserving the anterior portion of the head and jaws allows for revised descriptions of the endocranium and visceral skeleton, including the most complete endocranial ossifications recovered to date. As well, elements of the labial skeleton not known previously in ptyctodonts are preserved. Small amounts of soft tissues are recovered from Materpiscis, and extensive postcranial body muscle from Austroptyctodus. This new material resolves several questions regarding cranial and postcranial morphology of the Ptyctodontida. Moreover, similarities in placoderm and chondrichthyan pelvic fins and reproductive structures (claspers) suggest early parallel expression of HoxD and sonic hedgehog genes (shh), potentially representing the first evidence for expression of these genes in placoderms.

A broad phylogenetic analysis of extant and fossil fishes led previous authors to conclude that the occipital region of primitive actinopterygians was formed by the fusion of three segments. They included the dipnoan Griphognathus whitei from the Devonian of Western Australia in their study and assumed that all sarcopterygians also had the occipital ring made of three segments. In reaching this conclusion, they interpreted the foramina in the walls of the occipital rings in that species as carrying the spino-occipital nerves, and the furrows in the wall as being for myocommata. We now have access to three small juveniles of this species with the occipital rings isolated from the cranium, and three other young adults smaller than those described previously. There is no evidence of three occipital segments in any of these specimens. The isolated occipital rings have been examined by serial tomography and they have a single spino-occipital nerve that divides into dorsal and ventral branches. The attachment of the occipital segment to the cranium is described for the first time. The structure of the hyomandibula is described from a growth series, and its pattern is different from that of other sarcopterygians. The assumption that the occipital region in dipnoans is representative for all sarcopterygian fishes is shown to be incorrect.

The first detailed description of Furo muensteri is provided based on new well-preserved material from the Upper Jurassic of Ettling (Bavaria), the first record of Furo from this locality. Previously unknown features (circumorbital series, skull roof, maxilla, dorsal fin) are described. The corrected Latinized spelling replaces the original specific epithet “münsteri,” in accordance with Article 32.5 of the International Congress of Zoological Nomenclature code. Furo muensteri is diagnosed, among other features, by: pectoral fins as long as head; head length approximately 19% of standard length; 12 dorsal fin rays; 61 vertical scale rows from supracleithrum to hinge line; shape of ventral suborbital and position of its pit line; and morphology and arrangement of the maxilla and supramaxilla (gracile maxilla with deep posterodorsal angle and strongly tapered anteriorly; supramaxilla approximately 70% maxilla length, of uniform dorsoventral thickness, and with elongate anterior margin tapered to a point). Comparison with other Furo taxa reveals that F. muensteri is morphologically indistinguishable from F. praelongus (upper Kimmeridgian, Cerin, France). The two species are here synonymized, with F. praelongus becoming the junior synonym of F. muensteri (which has priority). Furo muensteri is also compared to the type species, F. orthostomus, and is currently retained in Furo pending a more detailed analysis of the genus. Presence of a sensory canal on the maxilla provides further evidence that Furo is sister to all other ophiopsids. Furo muensteri is currently known only from upper Kimmeridgian localities of Bavaria and France, except for Ettling whose exact geological age remains unknown.

The ability to determine relationships of Cenozoic fossil fishes relies heavily on having osteological information from their extant relatives with which to compare the fossil remains. For many higher teleost fishes, these osteological data do not exist. For example, †Parachanna fayumensis, from Eocene and Oligocene deposits of Egypt, was placed in the Recent snakehead genus Parachanna, but in the absence of data from extant members of the genus, this hypothesis could not be tested. Three other fossil channids were each given their own new genera, without an analysis of their relationships to one another or to the living genera. In order to properly test the relationships of these fossil species, the osteological information for living species is needed. This paper documents the osteology of the African snakehead, Parachanna obscura, for the first time. The documentation of the osteology forms the basis of a phylogenetic study of the relationships of the African fossil, †Parachanna fayumensis, with living African snakeheads. This study supports the generic placement of †P. fayumensis as being correct, as well as reciprocal monophyly of the two living snakehead genera and monophyly of the family. Monophyly of the Asian genus Channa and the relationships of the other fossil species need to be further assessed using a broader range of species, but this preliminary study is an essential first step. Apomorphic osteological characters are given for the family and genera, as well as a discussion of the biogeographical relationships of species of Parachanna.

The nearly complete skeleton of a mud turtle, Kinosternon skullridgescens, n. sp., is described from Santa Fe County, New Mexico. The specimen was collected from the early middle Miocene, early Barstovian, Skull Ridge Member of the Tesuque Formation (16.0–14.9 Ma) and is one of the oldest and most complete fossil representatives of the genus Kinosternon. Phylogenetically, the new taxon is placed on the stem of the flavescens group in conjunction with the extinct Kinosternon pojoaque from the late Barstovian of New Mexico. In addition to the flavescens group, K. skullridgescens shares morphological similarities with members of the subrubrum group (of North America), as well as with Kinosternon herrerai (from gulf drainages of eastern Mexico). Close proximity of the fossils K. skullridgescens and K. pojoaque to Mexico suggests that a small-plastroned ancestor of the extant taxa K. herrerai, Kinosternon angustipons (from lower Central America), and Kinosternon dunni (from northwestern South America) could have dispersed from North America into Central America by the middle Miocene (Barstovian). Phylogenetic results based on morphology suggest that the taxon Kinosternon steindachneri (formerly Kinosternon subrubrum steindachneri) be reinstated at the species level.

Three species of the genus Pseudopus are recognized in the Cenozoic of Eurasia: P. laurillardi (early—middle Miocene of Europe), P. pannonicus (late Miocene—middle Pleistocene of Central and Eastern Europe), and P. apodus (late Pleistocene—Recent, from Eastern Europe to Central Asia). Here, a new and oldest species of the genus Pseudopus, P. ahnikoviensis, sp. nov., is described from the early Miocene (MN 3) locality Merkur in Northwest Bohemia in the Czech Republic. The new species is distinguished from the other three species of the genus Pseudopus by five possible autapomorphies: length of sulcus between frontal and frontoparietal shields equals length of sulcus between frontal and interfrontal shields; ornamented surface of parietal consisting of densely arranged and very fine ridges, grooves, and pits; ventrolateral ridge of supratemporal process of parietal coincides with lateral margin of supratemporal process, hence no free ventrolateral surface of supratemporal process present; presence of surangular spine of dentary; and apices of crowns of robust bulbous teeth smooth. Pseudopus ahnikoviensis is most similar to P. laurillardi and represents the smallest species in the genus Pseudopus.

Here, we present information regarding the braincase anatomy of Chanaresuchus ischigualastensis, from the Ischigualasto Formation (Carnian) of Argentina. The braincase is very similar to that previously described for Chanaresuchus bonapartei, but bears some differences: rostral contour of the basal tubera nearly transversely oriented in ventral view; basal tubera with rounded distal tips in ventral view; paroccipital processes with a slight distal widening in caudal view; sulcus on the caudal surface of each paroccipital process restricted to the proximal two-thirds of the latter; and parabasisphenoidal ventral surface transversely concave. A phylogenetic analysis was performed using exclusively neurocranial characters, including a few new characters, to investigate phylogenetic hypotheses suggested by the neurocranial evidence. The neurocranial data support a sister-group relationship between Chanaresuchus ischigualastensis and C. bonapartei, on the basis of the presence of an intertuberal plate with a concave border. Our analysis also suggests the monophyly of a group formed by both species of Chanaresuchus and Proterochampsa. This group was supported by one apomorphy: the paroccipital processes are elongated relative to their height, which did not show homoplasy in our results. Our results show that Doswellia is the sister group of the clade formed by these proterochampsids, supporting inclusion of the former within the Proterochampsia, and this clade is recovered as more closely related to archosaurs than to Euparkeria.

We redescribe and present newly excavated sauropod material from the Lower Cretaceous Cloverly Formation of Wyoming that we refer to the titanosauriform Sauroposeidon proteles. In contrast to previous hypotheses that it was a brachiosaurid, we assert that Sauroposeidon is a member of the Somphospondyli on the basis of numerous features. Thus, the mid-Cretaceous disappearance of sauropods from the North American fossil record concerned both brachiosaurids and somphospondylans. We find claims for titanosaurs in the Early Cretaceous of North America to be unsubstantiated. The latest register of Sauroposeidon and other Early Cretaceous North American sauropods (before the ‘sauropod hiatus’) occurs in or below the coastal units marking transgression of the Western Interior Seaway, whereas many ecologically disparate dinosaur groups are present both below and above this boundary in the same geologic units that sauropods are found in. The presence of these through-ranging groups with sauropods before and after sauropod absence suggests that appropriate sauropod-bearing environments were present into the Late Cretaceous, implying that the disappearance of sauropods is not attributable to taphonomic or sampling bias. Furthermore, field observations of the Cloverly Formation indicate that Cretaceous pre-hiatus sauropods inhabited near-coastal environments, which were abundant in the western United States well after the start of the hiatus. The start of the sauropod hiatus is interpreted as the result of a genuine continent-wide extinction, coincident with the appearance of (and perhaps attributable to competition with) advanced ornithischian herbivores, decrease in habitat due to the incursion of the Western Interior Seaway, or both.

The pachycephalosaurian squamosal is one of the most diagnostic bones in this enigmatic group of dinosaurs, but little is known about variation in its morphology. Despite this, features of squamosal morphology are often used in diagnoses and phylogenetic studies. The recently proposed hypothesis of an ontogenetic transition from Dracorex to Pachycephalosaurus implies a large amount of ontogenetic variation in squamosal ornamentation that has not been well documented due to the small number of squamosals for these taxa. Stegoceras validum provides an important model for examining variation in squamosal morphology due to the large number of squamosals available. Here we examine the morphology of 14 S. validum squamosals both qualitatively and quantitatively. We find that although both node number and shape vary considerably, there is no clear ontogenetic pattern, which suggests a high degree of individual variation. The ornamental pattern, however, is maintained throughout the sample. Increase in node size is isometric for the enlarged medial node and normal posterior nodes, but negatively allometric for the vertex node. The squamosal bar is taller laterally than medially in nearly all specimens and increases in height with positive allometry at the same rate medially and laterally. We found no correlation between the width of the supratemporal fenestra on the squamosal and the size of the squamosal itself. Rather, this feature appears to exhibit a large degree of individual variation. Overall, these findings have implications for previous hypotheses of ontogenetic growth and function of the ornamentation, as well as the construction of phylogenetic characters.

We describe new avian remains from the middle Eocene of eastern Ukraine. The material includes well-preserved bones of a small species of Pelagornithidae (bony-toothed birds), which was tentatively identified as Odontopteryx toliapica in an earlier study. The Ukrainian pelagornithid is, however, distinguished from this species and other early Eocene pelagornithids in several features, and we describe it as Lutetodontopteryx tethyensis, gen. et sp. nov. The new material includes the most complete pseudodentition of a small Paleogene pelagornithid, and details of pseudotooth morphology strengthen previous assumptions that the pseudodentition of pelagornithids derives from tooth-specific developmental programs and is homologous to true avian teeth on a molecular level. Most major postcranial bones of L. tethyensis are represented by well-preserved bones, and in some derived morphological features the Ukrainian species agrees with Neogene pelagornithids but differs from the early Eocene Dasornis species. We also report a partial sternum of a much larger bony-toothed bird, which resembles the sternum of the middle Eocene Gigantornis eaglesomei. We further describe previously unknown skeletal elements of the loon Colymbiculus udovichenkoi (Gaviiformes), which show that this species substantially differs, in both wing and leg morphology, from the well-known gaviiform stem-group taxon Colymboides.

A fragmentary coracoid as well as isolated ulnae, carpometacarpi, and tarsometatarsi from the Varswater Formation at Langebaanweg, South Africa (early Pliocene), can be assigned to a new genus and species of true woodpecker (Picidae, Picinae), Australopicus nelsonmandelai, gen. et sp. nov. The new taxon is the first documented pre-Pleistocene record of woodpeckers from the entire African continent and it is clearly distinct from the three extant lineages of Picinae that are endemic to sub-Saharan Africa, i.e., Campethera-Geocolaptes, Dendropicos, and Dendrocopos obsoletus. Our phylogenetic analysis shows that the new taxon forms a clade with the extant woodpecker genera Celeus and Dryocopus, which do not occur in Africa, but in the Americas and Eurasia. The new taxon represents a previously unknown fourth lineage of African woodpeckers of Eurasian origin that probably became isolated on the African continent as a result of environmental changes during the Miocene. Evidence for an arboreal true woodpecker in the fossil record strongly supports previous hypotheses regarding the presence of riverine forests at Langebaanweg during the early Pliocene.

How is the bone tissue in skeletal supports of a neonatal elephant organized, and how does this histological structure differ among the neonates of modern species, mammoths, and insular dwarfs? We used synchrotron X-ray microtomography (SR-µCT) to obtain high-resolution image-‘slices’ noninvasively, from the femoral and tibial diapbyses of neonatal African elephants, a young juvenile Asian elephant, Columbian mammoths, and California Channel Island pygmy mammoths. The results compared favorably in level of detail with histological sectioning, but without the shrinkage, distortion, or loss of tissue inevitable with histology. From the tomography images we were able to rank by ontogenetic stage specimens of taxa that are otherwise difficult to categorize because they vary greatly in size; from these images we observed that laminar fibrolamellar bone predominated and were able to quantify vascular patterns. Bones of the Columbian mammoth typically had the thickest and largest number of laminae, whereas the insular dwarfs were notable in their variability. A distinct change in tissue microstructure marks the boundary between prenatal and postnatal periosteal bone deposition. Qualitatively, patterns of early bone growth of elephantids resemble those in many young tetrapods that grow into large adults, including sauropod dinosaurs.

Gnatho-dental specimens of the anthracotheres (Mammalia; Artiodactyla) from the four Neogene localities of central Myanmar are described. Four species of anthracotheres are recognized in the Neogene of central Myanmar: Microbunodon silistrensis and a small bothriodontine from the middle Miocene; and Microbunodon milaensis and Merycopotamus dissimilis from the latest Miocene to Pliocene. This discovery extends the temporal range of Microbunodon up to the Pliocene. The co-occurrence of forest-dwelling Microbunodon and grass-eating and semi-aquatic Me. dissimilis reinforces that central Myanmar was less arid and had a wider range of habitats than the northern Indian Subcontinent during the Pliocene. This implies the possibility that Pliocene Southeast Asia might have been a refugium for some late Miocene forest-dwelling ungulates.