New Zealand's late Pleistocene and Holocene Haast's eagle (Harpagornis moorei) was the sole predator of the moa (Aves: Dinornithiformes) and the largest eagle so far described. Recent molecular evidence has hypothesised that it evolved from a much smaller Asian/Australian eagle witthhin tht past 1.8 my. Osteometrof the igh-reaolution spirol computed tomography of skeletal remains of Haast's eagle were used to determine morphological and functionally significant characteristics of the central and peripheral nervous system and sensory apparatuses and to test these against hypotheses about behavior and evolution based on molecular and external morphology. Compared with other accipitrids, Haast's eagle had a proportionally low endocranial volume for its body weight. There was no anatomical evidence to support the significant use of olfaction by Haast's eagle, as the olfactory bulb size and cross-sectional area of the olfactory nerve foramen were no larger than would be expected for Accipitridae of this body weight. Both the size of the neural canal at the cervicothoracic junction and the brachial spinal quotient (ratio of neural canal area at cervicothoracic and caudal thoracic levels) were low, suggesting that the spinal cord at brachial levels was also small. These data are consistent with the hypothesis that Harpagornis had evolved to large size from a smaller ancestor and that the ancestors of Haast's eagle appear to have undergone rapid expansion of body size and elements of the hindlimb somatic nervous system at the expense of enlargement of the brain and visual, olfactory, and vestibular apparatuses.

Apateodus Woodward (Teleostei: Aulopiformes) is a Late Cretaceous bony fish that is known by only a few cranial specimens. On the basis of a previously unreported partial skeleton of Apateodus from the Smoky Hill Chalk Member (lower Campanian portion) of the Niobrara Chalk in western Kansas, U.S.A., here, we emend the diagnosis for the genus and describe a new species, A. busseni. The specimen substantiates the presence of Apateodus in the North American Western Interior Seaway during the early Campanian and marks the fifty-fifth osteichthyan fish so far recognized from the Niobrara Chalk of Kansas. Like another species of Apateodus, A. striatus (Agassiz), Apateodus busseni sp. nov. was a medium-sized (ca. 40 cm TL), slender pelagic fish, that was likely capable of fast swimming.

Isolated skull and postcranial bones previously assigned to the monotypic batrachosauroidid salamander genus Mynbulakia Nesov, 1981 (Uzbekistan, Byssekty Formation, Turonian) are re-interpreted as a mixture of two salamander taxa. The holotype maxilla of Mynbulakia surgayi Nesov, 1981, and dentaries, parietals, and femora previously referred to that species belong to smaller and, presumably, younger individuals of the cryptobranchid Eoscapherpeton asiaticum Nesov, 1981. Atlantes and trunk vertebrae previously referred to M. surgayi, along with additional vertebrae collected more recently from the Byssekty Formation, belong to Nesovtriton mynbulakensis gen. et sp. nov.. The structure of the atlas, the absence of spinal nerve foramina in the trunk and anterior caudal vertebrae, and the presence of unicipital transverse processes on the trunk vertebrae suggest that N. mynbulakensis may be a member of the Cryptobranchoidea. Two salamander taxa (E. asiaticum and N. mynbulakensis) are currently known from the Bissekty Formation and no salamander families are shared between the Late Cretaceous tetrapod assemblages of Asia and Euramerica.

Detailed description of the skull morphology of Winterswijk pachypleurosaurs expands the diagnosis of Anarosaurus heterodontus greatly. A. heterodontus is characterized by a broad and roughly shield-shaped sculptured nasal. The maxilla is separated from the anterolateral frontal processes by the nasal and the largely protruded prefrontal. The parietal in A. heterodontus exhibits a very characteristic lateral process that forms most of the dorsal margin of the upper temporal opening. The squamosal and quadrate are distinctly large and massive. The posterior part of the lower jaw is unusually stout. The pterygoid has an ascending posterior process that is ventrally articulated with the quadrate, supporting the mandibular condyle. The comparison of skull morphology with known pachypleurosaurs shows that A. heterodontus is most similar to Anarosaurus pumilio and Dactylosaurus. On the basis of the massive construction of the posterior part of the skull and the lower jaw, it is hypothesized that Anarosaurus heterodontus was able to crush hard-shelled prey.

Based on diagnostic jaw and dental material from the Cenomanian—middle Paleocene of the western U.S.A., we erect a new fossil lizard taxon, Contogeniidae taxon nov., that contains the following species: Contogenys sloani Estes, 1969 (Maastrichtian—early Paleocene, Montana); Contogenys ekalakaensis sp. nov. (middle Paleocene, Montana); Palaeoscincosaurus middletoni Sullivan and Lucas, 1996 (early Paleocene, Colorado); Palaeoscincosaurus pharkidodon sp. nov. (Campanian, Utah); Utahgenys evansi gen. et sp. nov. (Turonian, Utah); and Utahgenys sp. indet. (Cenomanian, Utah). These taxa share unique characteristics of tooth and jaw morphologies (e.g., tooth crowns truncate and bearing longitudinal apical grooves; hypertrophied inferior alveolar canal). Phylogenetic analysis indicates that among squamates, Contogeniidae taxon nov. is most closely related to Xantusiidae.

Metriorhynchids have been interpreted as the only archosaurs entirely adapted to pelagic marine life, given the deep morphological modifications of their skeletons. The most conspicuous feature in the skull involves the “fenestra and fossa antorbitalis complex.” Exceptionally preserved natural cast of snout cavities of Geosaurus araucanensis, found in the Late Jurassic of the Neuquén Basin (northwestern Patagonia, Argentina), allow exploration of the rostral anatomy of this metriorhynchid. The presence of a paranasal sinus system, hypothetically reconstructed by other authors using EPB method approach, can now be confirmed based on direct morphological evidence. We propose that the openings classically identified in the literature as internal and external antorbital fenestrae of metriorhynchid have been misidentified; the preorbital opening of metriorhynchids is a neomorph associated with a novel salt gland that evolved independent of the antorbital fenestra, and that the true antorbital cavity of G. araucanensis, as well as this cavity in all other metriorhynchids, is internalized. Although this hypothesis could be considered as unorthodox, it is the one that requires the least ad-hoc assumptions to explain observations. Many phylogenetic studies depict the Thalattosuchia nested within Neosuchia. ln these trees, the successive outgroups of Thalattosuchia are Dyrosaurids/Pholidosaurids, followed by derived Neosuchians (Goniopholids, Bernissartia, Eusuchians). All these taxa have a closed antorbital fenestra. Within this phylogenetic scenario, the internalization of the antorbital cavity did not occur in the ancestor of thalattosuchians, but in the ancestor of a much more inclusive clade of neosuchians.

A broad survey of crocodyliform archosaurs and their outgroups was conducted to explore the evolutionary and morphological patterns of the orbitotemporal region, which is a highly apomorphic but poorly understood portion of the head. Data were gathered on the topological similarity and phylogenetic congruence of the epipterygoid, laterosphenoid, and temporal region as a whole, including relevant osteological correlates and such inferred soft tissues as the trigeminal nerves and jaw musculature. Despite the complete suturing of the palatocranial junction, the epipterygoid remained a consistent cranial element throughout crocodyliform evolution, only to be replaced by the topologically analogous, but developmentally neomorphic lateral bridge of the laterosphenoid during the early evolution of eusuchians. These changes led to a unique morphology of the region surrounding the exit of the trigeminal nerve. Mesoeucrocodylian taxa exhibit a diversity of epipterygoid morphologies including waisted (e.g., Araripesuchus), overlapping (e.g., Sarcosuchus), and isolated (e.g., Goniopholis, Leidyosuchus) forms. The isolated form, in which the epipterygoid is uncoupled from the pterygoid and does not to cover the cavum epiptericum laterally, represents a key transition to the extant condition of loss of the epipterygoid. Changes in the epipterygoid coincide with the migration of M. pseudotemporalis superficialis onto the laterosphenoid outside of the dorsotemporal fossa and the topological change in the intermuscular path of the maxillary nerve, both of which are apomorphies found in extant crocodylians. These data reflect a diverse and potentially homoplastic distribution of orbitotemporal morphologies among mesoeucrocodylians and indicate that the epipterygoid was only recently eliminated in crocodyliform evolution.

The long bone histology of the “hypsilophodontid” dinosaurs Orodromeus makelai, Dryosaurus altus, and Tenontosaurus tilletti, was examined from perinate to largest available ontogenetic stages. These ontogenies were compared with each other, and to those of other dinosaurs, notably hadrosaurids such as Maiasaura and Hypacrosaurus. Orodromeus is a small dinosaur, and its more moderate growth trajectory is consistent with those generally observed for relatively small dinosaurs and other small ornithosuchians. Tenontosaurus achieves a relatively larger adult size, and its bone histology through ontogeny is similar to those of hadrosaurs, although reflective of slightly lower growth rates. Dryosaurus is thought to be a small dinosaur, but in its largest recognized ontogenetic stages it does not display the “adult” histological features of other ornithopod adults, although it does suggest growth rates comparable to those of larger ornithopods, including hadrosaurs. Determinate growth is observed in Orodromeus, Tenontosaurus, and hadrosaurs, so we infer that the adult stage of Dryosaurus has not yet been recognized. Dinosaurs continued to grow slowly after reaching adult size, so the largest known specimens of most taxa are not maxima.

A new ornithischian dinosaur, Albalophosaurus yamaguchiorum gen. et sp. nov., is described on the basis of a partial skull and lower jaw recovered from the Lower Cretaceous Kuwajima Formation (Tetori Group) of central Japan. The specimen exhibits a unique suite of characters in the dentition, maxilla, and dentary. A phylogenetic analysis recovers Albalophosaurus as a basal member of Ceratopsia: however, this placement is weakly supported and no unambiguous ceratopsian synapomorphies can be identified in the material. This new taxon should, therefore, be regarded as Cerapoda incertae sedis until further material is discovered. Nevertheless, Albalophosaurus offers additional information on the diversity of east Asian dinosaur faunas during the Early Cretaceous and represents only the third valid dinosaur taxon to be described from Japan.

Considerable controversy surrounds the taxonomy and phylogenetic relationships of Maastrichtian basal neornithischian taxa from North America. Discovery of previously unrecognized cranial material from the paratype specimen of Thescelosaurus neglectus (USNM 7758), along with the examination of two new specimens that preserve nearly complete skulls and mandibles, allows for reevaluation of specimens previously referred to Thescelosaurus that preserve cranial material, including the holotypes of Bugenasaura infernalis (SDSM 7210) and Parksosaurus warreni (ROM 804). A phylogenetic analysis was conducted that included as terminals the holotypes of B. infernalis and P. warreni, the type series of T. neglectus, and six specimens previously referred to Thescelosaurus. This analysis is the first to recover a clade containing all basal neornithischian taxa from the Cretaceous of North America, within which P. warreni is recovered as the sister taxon to a Thescelosaurus clade whose monophyly is supported by five cranial autapomorphies. The results of this analysis support: (1) the synonymization of Bugenasaura with Thescelosaurus; (2) the removal of ROM 804 from Thescelosaurus; and (3) the validity of Thescelosaurus garbanii and its referral to Thescelosaurus despite its fragmentary nature. Currently, Parksosaurus and Thescelosaurus are the only basal neornithischian taxa definitively known from Maastrichtian age sediments of North America, while other basal neornithischian taxa proposed to demonstrate fossorial behavior (i.e., Orodromeus, Oryctodromeus, and Zephyrosaurus) are known from Campanian and older sediments. This temporal segregation may support previous hypotheses of an environmental or ecological shift during the latest Cretaceous in North America.

Colepiocephale lambei from the Foremost Formation of Alberta is a problematic pachycephalosaurid that has recently been hypothesized as (1) synonymous with Stegoceras validum, (2) a distinct species within the Stegoceras clade, and (3) an unusual taxon more derived than Stegoceras. Conflict among these hypotheses arises from different interpretations of the parietosquamosal region of the dome. Re-evaluation of the morphology of C. lambei based on all known material from Alberta indicates that the posteromedial extension of the parietal is broken off at its base in every specimen, and what was interpreted as the posterior margin of the dome actually is the anterior margins of large supratemporal fenestrae. The squamosals are not preserved in any specimen, but the form of the parietal suggests that they were present in the typical pachycephalosaurid configuration. Colepiocephale lambei is considered to be a distinct taxon based on several autapomorphies and stratigraphic segregation from S. validum. Phylogenetic analyses that include new anatomical data confirm that C. lambei occupies a basal position within domed pachycephalosaurids, but its precise systematic position with respect to S. validum remains equivocal. This general interpretation is more consistent with its stratigraphic position as one of the oldest known members of Pachycephalosauridae. Additionally, this study results in the recognition of the first material referable to this taxon from the Judith River Formation of Montana, which results in a geographic range extension for this taxon and extends its stratigraphic range into the equivalent of the lower Oldman Formation in Alberta.

Although the macronarian sauropod Brachiosaurus is one of the most iconic dinosaurs, its popular image is based almost entirely on the referred African species Brachiosaurus brancai rather than the North American type species Brachiosaurus altithorax. Reconsideration of Janensch's referral of the African species to the American genus shows that it was based on only four synapomorphies and would not be considered a convincing argument today. Detailed study of the bones of both species show that they are distinguished by at least 26 characters of the dorsal and caudal vertebrae, coracoids, humeri, ilia, and femora, with the dorsal vertebrae being particularly different between the two species. These animals must therefore be considered generically separate, and the genus name Giraffatitan Paul 1988 must be used for “Brachiosaurus” brancai, in the combination Giraffatitan brancai. A phylogenetic analysis treating the two species as separate OTUs nevertheless recovers them as sister taxa in all most parsimonious trees, reaffirming a monophyletic Brachiosauridae, although only one additional step is required for Giraffatitan to clade among somphospondylians to the exclusion of Brachiosaurus. The American Brachiosaurus is shown to be somewhat different from Giraffatitan in overall bodily proportions: it had a longer and deeper trunk and probably a longer and taller tail, carried a greater proportion of its mass on the forelimbs, and may have had somewhat sprawled forelimbs. Even though it was overall a larger animal than the Giraffatitan lectotype, the Brachiosaurus holotype was probably immature, as its coracoids were not fused to its scapulae.

Bones of Alamosaurus sanjuanensis were collected from the Upper Cretaceous Javelina and Black Peaks formations in Big Bend National Park, Texas, for use in the first extensive histological study of a titanosaurian sauropod dinosaur. Samples include cervical, dorsal, and caudal vertebrae; ribs; a scapula; humeri; an ulna; a metacarpal; an ischium; femora; a tibia; fibulae; and a metatarsal, representing a minimum of twenty two individuals between 35% and 100% maximum length. In elements of less than 60% maximum length, cortical tissue is primarily reticular or plexiform fibro-lamellar, often with modulations. Resorption in some specimens appears to have proceeded as outwardly expanding bands. In larger specimens (70–79% maximum length), primary tissue is only visible interstitially as either fibro-lamellar or parallel-fibered with low vascularity and annuli. Secondary osteons are in many places linearly arranged. Haversian systems are so pervasive in many of the largest specimens (80–100% maximum length) that virtually no primary tissue remains. However, one adult sized femur possesses laminar fibro-lamellar tissue in its outer cortex, suggesting that the adult size for Alamosaurus may have been underestimated. Camellae typical of titanosaur dorsal vertebrae are present in Alamosaurus but absent in caudal vertebrae. Although pneumatic dorsal ribs characterize Titanosauriformes, sections of ribs analyzed in this study show the cavities in Alamosaurus are much more extensive than most previously described. Overall, Alamosaurus bone histology combines characteristics found in less derived sauropods, such as an early onset of secondary reconstruction, with novel characters such as extensively pneumatic ribs and longer retention of juvenile tissue.

Three-dimensional finite element analyses were performed on the cranium of the horned theropod Carnotaurus sastrei to assess how it would have performed mechanically during biting and frontal butting. This technique proved to be an effective tool to provide a better understanding of the cranial functional morphology of C. sastrei. The analyses indicated that the jaw-closing musculature of C. sastrei would have played a key role in diminishing the stress level on the cranium during biting, particularly for the braincase. Force analyses show that C. sastrei would have had a comparatively weak (and faster than Allosaurus fragillis) muscle-driven bite. The finite element analyses also provide quantitative evidence to suggest that the cranium of C. sastrei could have withstood high velocity impacts to its teeth resulting from a hypothetical hatchet-like biting mode, but could not have endured severe, rapid frontal blows during agonistic encounters.

The neurocranial fragment from the Late Miocene (Huayquerian SALMA, 9,0 to 6,8 Ma) of Estancia La Pastosa new locality, Puerto Madryn Formation, Patagonia, Argentina, is described. This is the first available fossil skull of Accipitridae in South America. Features as the elongated-triangular shape of the cranial roof, vast separation between the processus postorbitalis and between both fossae temporalis, and narrow and acuminate outline of the fossa temporalis allow to assign the fossil to Accipitridae Vieilloit, 1816. The well-preserved cavum cranii allowed construction of an endocast that partially reflects brain anatomy. The latter is very similar to those of extant and fossil species of Accipitridae. This fossil eagle is the only predator recorded at this new Miocene locality thus far.

The upper Omingonde Formation of Namibia contains a diverse vertebrate fauna represented by amphibians, dicynodonts, archosaurs, therocephalians, and cynodonts, which is generally considered to be of Anisian age. Four cynodont taxa (Cynognathus, Diademodon, Trirachodon, and Titanogomphodon) are currently known from this fauna. Here we document four additional cynodonts, all of which were recovered from the highest levels of the formation: Luangwa, an indeterminate traversodontid, Aleodon, and Chiniquodon. These discoveries allow us to recognize the Omingonde Formation as preserving the most diverse fauna of Middle Triassic cynodonts in the world. Furthermore, we recognize that the formation is a biostratigraphic link among Middle Triassic faunas from South Africa, Tanzania, Zambia, Argentina, Brazil, and Antarctica. Aleodon is recorded here for the first time in Anisian faunas of southern Africa, and the unexpected record of Chiniquodon poses a biostratigraphic enigma because this taxon is known only from Ladinian—Carnian faunas of South America. We explore some possible scenarios related to the radiation of traversodontid cynodonts in Gondwana during the Anisian.

We report for the first time the presence of the cynodont Diademodon tetragonus Seeley, 1894 in the South American Triassic. The specimen, represented by a fragmented skull and lower jaws, was found in levels of the Rio Seco de la Quebrada Formation (Puesto Viejo Group), Mendoza Province, Argentina. It is assigned to D. tetragonus based on its overall skull shape (narrow and elongated snout with a concave outline in dorsal view), the morphology of the jugal that forms most of the dorsoventral depth of the zygoma and shows a well-excavated external auditory meatus, and a postcanine series including circular outlined anterior teeth, ovoid gomphodont teeth in the middle, and posterior sectorial teeth. The association of this taxon with the cynodonts Cynognathus crateronotus and Pascualgnathus polanskii, along with comparisons to African Triassic assemblages, suggest an Early to Late Anisian age for the Río Seco de la Quebrada Formation. These levels are most likely correlated to the subzones B and C of the South African Cynognathus Assemblage Zone, where both Cynognathus and Diademodon are known. This discovery represents the fourth report of shared cynodont genera between allegedly Lower to Middle Triassic African and South American terrestrial faunas.

The first Pliocene marsupial fossil trackways from Australia are described. The trackways, attributed to Euowenia grata (De Vis) (Diprotodontoidea, Marsupialia), occur in the middle Pliocene Tirari Formation, Warburton River, northern South Australia. The trackways were formed as the animals made their way across a soft claypan. Pad impressions, subsequently infilled by a gypsum-cemented clay, indicate how weight was distributed within the pes.

The megatherioid sloths from the Santa Cruz Formation (Santacrucian Age; early-middle Miocene, Patagonia, southernmost South America) occupy basal positions in the most recent phylogenetic schemes. The cranial morphology of Eucholaeops, particularly of the teeth, suggests interesting functional features that shed light on the type of food it was capable of processing, and thus on the diet. A detailed morphofunctional analysis of the jaw apparatus was performed, and the results briefly compared with other contemporary megatherioid sloths. Comprehensive descriptions of the teeth of Eucholaeops allow us to generate a nomenclature for describing the inferred occlusal pattern analogous to that applied to other mammals. Based on examination and mapping of occlusal wear facets, we reconstruct two distinct jaw movements during the power stroke. One corresponds to the basic therian pattern equivalent to Phase I: the working side mandibular corpus is moved dorsally, mainly orthally but also anteriorly and slightly medially; the result is puncturing, tearing and shearing of food. The second is a distinct and unrelated movement of the working side corpus dorsally, mainly orthally, but also posteriorly and slightly medially; the dominant result is to produce shearing of food. The analysis of the tooth wear facets, combined with the shape of the temporomandibular joint, the presence of a fused mandibular symphysis, and a well-developed temporalis muscle, indicates that the orthal component was predominant during mastication. Eucholaeops, and probably nearly all other Miocene megatherioids, were most likely leaf eaters and the primary method of food reduction must have been by shearing or cutting.

Pachyarmatherium is an enigmatic cingulate genus formerly recognized as Dasypodoidea, but tentatively assigned to the poorly known subfamily Glyptatelinae (Glyptodontidae) by later authors. A new species is described from the late Pleistocene of northeastern Brazil. The material comprises several isolated osteoderms, carapace fragments, and miscellaneous postcranial material. The new species, Pachyarmatherium brasiliense, differs from the North American Pachyarmatherium leiseyi in being larger and in having osteoderms with heptagonal shape, main figures with oblong, subrounded and polygonal outline (instead of only polygonal), and a larger number of peripheral figures. To clarify the affinities of Pachyarmatherium, a cladistic analysis was conducted of 48 morphological postcranial characters and scored for 11 cingulates (the new species, four dasypodids, one pampathere and five glyptodontids) and two pilosans used as outgroup taxa. In the most parsimonious topology obtained, Pachyarmatherium is the sister-group to a clade including Pampatheriidae and Glyptodontidae. This result contradicts the previous tentative allocation of Pachyarmatherium to the Glyptodontidae, and indicates that the glyptodont-like characters of the carapace and osteoderms of Pachyarmatherium are homoplasies shared with glyptodonts. The taxonomic validity of the genera Pachyarmatherium and Neoglyptatelus is recognized, but the morphological evidence from the external ornamentation of osteoderms does not offer support for the placement of Pachyarmatherium within Glyptatelinae.

Although many desmostylian genera, including Behemotops katsuiei, Desmostylus hesperus, Paleoparadoxia tabatai, P. weltoni, P. repennngi, and Ashoroa laticosta are known from skulls and postcrania, Cornwallius sookensis has previously been known only from isolated teeth. New material from the Emlong Collection of the Smithsonian Institution is described here. ThNew MaterCornwallius Sookensis skulls, four partial mandibles, four isolated teeth, an innominate and a tibia from the Yaquina Formation of coastal Oregon (Zemorrian, Oligocene). The diagnosis of this taxon is emended to include the development of a postorbital process of the jugal, a dorsal midsagittal keel of the rostrum, paraglossal crests of the maxillary diastema, and a sagittal crest. Dental characteristics, including identification of canines and cusp homologies, are defined. Though the skulls and mandibles described are nearly complete, some premolars are missing and alveoli are damaged, making interpretation difficult. The adult dental formula is tentatively interpreted as 2.1.2?.3/ 1.1.1?.3, although the deciduous dentition includes two pairs of lower incisors. Mandibular characteristics include a posteriorly sloped anterior edge of the coronoid process. Phylogenetic analysis incorporating this new information resulted in one most parsimonious tree for the Order Desmostylia. The families Paleoparadoxiidae (comprising three Paleoparadoxia species) and Desmostylidae (comprised of Ashoroa, Cornwallius, Desmostylus, and Vanderhoofius) are found to be distinct clades. The clade including both species of Behemotops (Behemotopsidae) is found to be the sister group to the clade comprising Paleoparadoxiidae and Desmostylidae.

On the basis of a partial odontocete (toothed whale) skeleton (complete skull with associated ear bones, mandible, teeth, hyoid bones, and cervical vertebrae) from the middle Miocene (ca. 14–12 Ma) levels of the Pisco Formation at Cerro los Quesos, Pisco-Ica desert, southern coast of Peru, a new genus and species of beaked whale (Ziphiidae), Nazcacetus urbinai, is described. This small ziphiid, the best documented fossil species to date, is characterized by a strong reduction of the upper and lower dentitions except for a pair of large apical mandibular teeth, a Tasmacetus-like vertex, a short mandibular symphysis, and the abrupt elevation of the dorsal margin of the mandible towards the coronoid process. A phylogenetic analysis indicates that Nazcacetus is a crown ziphiid, in a more crownward position than Berardiinae and Tasmacetus, but branching before the Hyperoodontinae and Ziphiinae lineages. Several morphological traits observed in Nazcacetus, including the reduction of teeth, the small temporal fossa, and the large hamular process, suggest that this taxon possessed the suction feeding capacities of Recent ziphiids, a specialization possibly related to the exploitation by the early ziphiids of a new ecological niche. The presence of large apical teeth on the mandible of Nazcacetus, a character highly dimorphic in recent ziphiids, might indicate that sexual dimorphism was already present as early in ziphiid evolutionary history as the middle Miocene, a hypothesis supported by the nested position of Nazcacetus in the ziphiid phylogenetic tree.

The sites of Kossom Bougoudi (KB) and Kollé (KL) are intermediate in age between the Sahelanthropus sites of Toros Menalla and the Australopithecus sites of Koro Toro, and their bovid faunas are also intermediate in composition and stage of evolution. Four new taxa are erected. The hippotragine Tchadotragus fanonei nov. sp. is more derived than the type-species of the genus, from Toros Menalla. Kobus ammolophi nov. sp. is also clearly related to the most common reduncine of the latter sites, rather than with other African forms. However, Jamous kolleensis nov. gen. nov. sp., a new bovine with extremely divergent, horizontal horn-cores, is unrelated to other Chadian Bovini. The KB and KL assemblages document the gradual replacement of hippotragines by alcelaphines in Northern Chad, but it is not necessarily linked with environmental change, as all Chadian assemblages virtually lack indicators of bush or woodland like Tragelaphini and Aepyceros. In spite of a significant endemic component, the KB and KL Bovidae compare best with some of those from Sahabi in Libya, confirming North-South connections.

The great North American Pleistocene pantherine felid Panthera atrox has had a turbulent phylogenetic history, and has been claimed to show affinities to both the jaguar and the tiger; currently, it is most often regarded as a subspecies of the extant lion. The cranial, mandibular, and dental morphology of Panthera atrox was compared with those of extant lions, jaguars, and tigers using bivariate, multivariate, and shape analyses. Results indicate that the skull of Panthera atrox shows lion affinities, but also deviates from lions in numerous aspects. Mandibular morphology is more similar to jaguars and tigers and, as with cranial morphology, the mandible shows a number of traits not present among extant pantherines. Multivariate analyses grouped Panthera atrox separately from other pantherines. Panthera atrox was no lion, and cannot be assigned to any of the extant pantherines; it constituted a separate species. A possible scenario for evolution of P. atrox is that it formed part of a pantherine lineage that entered the Americas in the mid-Pleistocene and gave rise to the extant jaguar and Panthera atrox in the late Pleistocene of North America. These studies suggest that previous models of lion biogeography are incorrect, and although lions may have been present in Beringia, they did not penetrate into the American mainland.