Uintatherium

Uintatherium, from Uinta Mountains, and Ancient Greek θηρίον (theríon), meaning "beast", is an extinct genus of herbivorous dinoceratan mammal that lived during the Eocene epoch. Two species are currently recognized: U. anceps from the United States during the Early to Middle Eocene (50.5-39.7 million years ago) and U. insperatus of Middle to Late Eocene (48–37 million years ago) China. The first fossils of Uintatherium were recovered in the Fort Bridger Basin, and were initially believed to belong to a new species of brontothere. Despite other generic names being assigned, such as Edward Drinker Cope's Loxolophodon and Othniel Charles Marsh's Tinoceras, and an assortment of attempts at naming new species, Uintatherium anceps has since come to encompass all of these.

The phylogeny of Uintatherium and other dinoceratans has long been debated. Originally, they were assigned to the now-invalid order Amblypoda, which united various basal ungulates from the Palaeogene. Ambylpoda has since fallen out of use. Since then, various hypotheses of dinoceratan phylogeny have been proposed. The most widespread is that they are related to the South American xenungulates, together forming a mirorder called Uintatheriamorpha. If this is correct, dinoceratans, and thus Uintatherium, may not be ungulates at all. However, it has been noted that traits shared between the two groups may be the result of convergent evolution. Within Dinocerata itself, Uintatherium belongs to the family Uintatheriidae, and is one of two members of Uintatheriinae; the other two are Eobasileus and Tetheopsis.

Uintatherium was a very large animal, with U. anceps having a shoulder height of Template:Cvt and a body mass of Template:Convert. The largest Uintatherium skulls known, originally assigned to Loxolophodon, measure Template:Cvt in length. It is overall similar to the other two uintatheriine genera, though it had a broader skull. Like them, UintatheriumTemplate:'s skull bears a series of bony, skin-covered protrusions: one pair on the tip of the snout, one pair above the gap between the canine and cheek teeth, and one pair toward the back of the skull. EobasileusTemplate:' skull was quite similar, though the middle pair of protrusions sat further back, directly above the cheek teeth. The canines of Uintatherium were very large, and were supported by a pair of bony flanges extending from the lower jaw. They were likely sexually dimorphic, and may have been used in display or for defense. Behind the skull, the skeleton of Uintatherium bears a combination of characteristics often associated with proboscideans (elephants and relatives) and rhinocerotids.

Uintatherium evolved during the Paleocene-Eocene thermal maximum, a period which saw some of the highest global temperatures in Earth's history. Most of the North American continent was covered in closed-canopy forests, with the Bridger Formation, one of the localities U. anceps is best known from, consisting of an inland lake surrounded by birch, elm and redwood trees. The depositional environment of the later Uinta Formation was interspersed by open savannahs, resulting from a global cooling event which resulted in the gradual aridification of North America. The Chinese U. insperatus lived in a brackish environment mixed with a semi-arid steppe.

Taxonomy

Early history

File:Rs-11415 1024x1024.jpg

Restoration of Edward Cope's proboscidean Loxolophodon theory from 1873

In September 1870, a fragmentary skeleton (cataloged under YPM 11030) of Uintatherium was unearthed by US army Lieutenant W. N. Wann in the Bridger Basin of Wyoming. These sediments of the Bridger Basin come from the Eocene-aged Bridger Formation. This incomplete skeleton was then sent to paleontologist Othniel Charles Marsh who described it in 1871 as a new species of the brontothere Titanotherium, Titanotherium anceps. This was the first mention of a uintathere in scientific literature. The following year, Marsh and Joseph Leidy collected in the Eocene Beds near Fort Bridger while Edward Drinker Cope, Marsh's competitor, excavated in the Washakie Basin. In August 1872, Leidy named Uintatherium robustum based on an incomplete skull and partial mandibles (ANSP 12607). Another specimen discovered by Leidy's crews consisting of a canine was named Uintamastix atrox and was thought to have been a saber-toothed and carnivorous.

Eighteen days after the description of Uintatherium, Cope and Marsh both named new genera of Uinta dinoceratans, Cope naming Loxolophodon in his "garbled" telegram and Marsh dubbed Tinoceras. Due to Uintatherium being named first, Cope and Marsh's genera are synonymous with Uintatherium. Cope described two genera in his telegram, Loxolophodon and Eobasileus; the latter is currently considered separate from Uintatherium. Tinoceras was a new genus made for Titanotherium anceps by Marsh. Several days later, Marsh erected the genus Dinoceras. Dinoceras and Tinoceras would receive several additional species by Marsh throughout the 1870s and 1880s, many based on fragmentary material. Several complete skulls were found by Cope and Marsh crews, leading to theories like Cope's proboscidean assessment. Because of Cope and Marsh's rivalry, the two would often publish scathing criticisms of each other's work, stating their respective genera were valid. The trio would name 25 species now considered synonymous with Marsh's original species, Titanotherium anceps, which was placed in Leidy's genus, Uintatherium. In 1876, William Henry Flower, Hunterian Professor of Comparative Anatomy, wrote a letter in Nature wherein he formally suggested incorporating all of Cope's, Leidy's, and Marsh's taxa into Uintatherium, due to it being named first (which would make it a senior synonym), and a lack of convincing evidence for their separation.

File:Uintatherium insperatus Holotype IVPP.jpg

Holotype skull (IVPP V6379) of U. insperatus, Paleozoological Museum of China

Many additional discoveries of Uintatherium have since occurred, making it one of the best-known and popular American fossil mammals. Princeton University launched expeditions to the Eocene beds of Wyoming in the 1870s and 1880s, discovering several partial since skulls and naming several species of uintatheres that are now considered synonyms of U. anceps. Major reassessment came in the 1960s by Walter Wheeler, who synonymized and redescribed many of the Uintatherium fossils discovered during the 19th century A cast of a Uintatherium skeleton is on display at the Utah Field House of Natural History State Park. A skeleton of Uintatherium is also on display at the Smithsonian National Museum of Natural History in Washington, DC.

Fossils assigned tentatively to Uintatherium have been described from parts of Asia since 1962, when Zhou Mingzhen and Y. S. Zhou reported teeth (the third upper molar and two upper canines) closely resembling those of the genus from Xintai, Shandong, China. In 1977, Gabounia reported fossils possibly referrable to Uintatherium had been recovered from Tschaibulak, near Zaisan, Kazakhstan. These were both referred to an indeterminate position within Uintatheriidae, not to Uintatherium itself, though the former was documented as cf. Uintatherium sp. In November 1978, the first unambiguous Asian specimen of Uintatherium was recovered. Wang Daning, Tong Shuisheng, and Wang Chuanqiao, working at strata from the lower part of the Lushi Formation (Henan Province, China), recovered an almost intact skull. Aside from damage to the nasal bone and zygomatic arches, it was essentially complete. The latter two wrote that the skull likely belonged to an elderly individual due to the condition of the teeth, which were severely worn. In 1981, the specimen was described. It was assigned to a new species of Uintatherium, U. insperatus.

Classification

Uintatherium was initially regarded by Marsh as a brontothere. However, similarities to proboscideans (relatives of elephants), noted by various authors, lead Cope to classify it as a member of that group. While he acknowledged Marsh's reasoning, he nonetheless believed that it stemmed from "unusual sources", and that: "The absence of incisor teeth no more relates these animals to the Artiodactyla than it relates the sloth to the same order [...] the presence of paired horns no more constitutes affinity to the ruminants than it does in the case of the 'horned-toad'." It has since been recognised that similarities to proboscideans are likely the product of convergent evolution. Uintatherium was reclassified by Henry Fairfield Osborn in 1881 as part of the order Dinocerata. At the time, dinocerates were believed to be part of Amblypoda, a group uniting an assortment of basal ungulates from the Palaeogene, and were sometimes referred to simply as "dinoceratous amblypods".

The group Amblypoda has since fallen out of use, and is generally regarded as polyphyletic, meaning that it was an unnatural group consisting of an assortment of distantly related clades. Dinocerata, however, has persisted, though the precise relationships of the order have been the subject of debate. Relationships with South American native ungulates (SANUs), specifically xenungulates, have been suggested, due in part to perceived similarities to Carodnia, with Spencer G. Lucas and Robert M. Schoch in 1998 supporting the complete removal of both clades from Ungulata. If dinoceratans and xenungulates are indeed related, they may constitute the mirorder Uintatheriamorpha. Lucas and Schoch, in 1985, noted dental similarities between uintatheriamorphs and the "anagalid" Pseudictops, which might in turn be related to modern lagomorphs (rabbits, hares, and pikas), and would thus, as the same authors commented in 1998, be "[...] tantamount to identifying uintatheres as giant horned bunnies [...]". It has since been asserted that no strong evidence for this relationship exists, and that similarities observed may simply be the result of convergence, in no small part because of how small and specialised anagalids are in relation. In 1997, Malcolm McKenna regarded Uintatheriamorpha as a synonym of Dinocerata, though did not elaborate. Bruce J. Shockey and Federico Anaya Daza, in 2003, rejected the use of the term Uintatheriamorpha altogether, considering the supporting data too weak. A 2015 analysis by Benjamin R. Burger, presented to the Society of Vertebrate Palaeontology as a conference abstract, not only recovered a monophyletic Uintatheriamorpha (consisting of Carodnia + Dinocerata), but recovered them immediately basal to the artiodactyl/perissodactyl split.

Donald R. Prothero, Earl M. Manning, and M. S. Fischer, in 1988, suggested that dinoceratans and pyrotheres were part of Paenungulata (now consisting solely of hyracoid and tethythere afrotheres), which by their definition also included perissodactyls. Regardless, a phylogenetic analysis published in 2019 by Thomas Halliday et al. recovered Uintatherium (the only dinoceratan included in the dataset) within a clade consisting entirely of SANUs, as the most basal branch of a clade otherwise consisting of Astraponotus, Carodnia, Parastrapotherium, and Pyrotherium.

A cladogram showing the phylogenetic position of Uintatherium, after Halliday et al. (2019), is as follows:Template:Clade

File:Uintatheriin skull comparison.png

Diagram of the skulls of Eobasileus cornutus, Uintatherium anceps, and U. insperatus

Dinocerata has historically been divided into two families, Prodinoceratidae and Uintatheriidae, though some authors use only one family. Assuming two families exist, Uintatheriidae consists of the majority of dinoceratans, and has itself been divided into Gobiatheriinae and Uintatheriinae; occasionally, the latter has been divided even further, down to tribe level (Bathyopsini and Uintatheriini). The most basal uintatheriid was Bathyopsis. Walter H. Wheeler suggested in 1961 that the taxa now classed as uintatheriines formed a primarily anagenetic lineage, and that Uintatherium was one of few diverging genera, possibly evolving from Bathyopsis middleswarti (which he believed to be ancestral to both Uintatherium and later dinocerates). Robert M. Shoch and Spencer G. Lucas, in 1985, offered a cladistic hypothesis of Dinocerata, in which Uintatherium is the sister taxon to a clade consisting of Eobasileus and Tetheopsis, slightly more derived than Bathyopsis. William D. Turnbull, however, suggested in 2002 that both Tetheopsis species could be lumped into Eobasileus, and that Uintatheriini might thus consist exclusively of Eobasileus and Uintatherium.

Description

File:Uintatherium DB.jpg

Life restoration of U. anceps

Uintatherium was a large, graviportal animal, with short and robust limb bones. It appears to have exhibited strong sexual dimorphism: males had larger canines, larger flanges on the lower jaws, larger sagittal crests, larger horns, and an overall larger body size. As it was a fairly large mammal which lived mostly in temperate environments, William Berryman Scott suggested that Uintatherium may have been predominantly hairless, though noted that there is no direct evidence. Its thick, barrel-shaped ribcage has led some to suggest that it may have practised hindgut fermentation, like modern horses and sea cows.

Skull

File:Dinoceras mirabile Marsh MNHN.jpg

Cast of U. anceps skull, French National Museum of Natural History, Paris

The skull of Uintatherium is roughly three times longer than it is wide. Most U. anceps skulls range from Template:Convert in length, whereas the only known U. insperatus skull measures Template:Convert. Some specimens have skulls which, when measured at the zygomatic arches, are roughly Template:Convert wide, suggesting a very large overall skull size. Furthermore, some specimens initially referred to Loxolophodon have skull lengths of up to Template:Cvt, nearly a third larger than most others. The skull of U. anceps can be distinguished from those of other uintatheriins (if that clade exists) by its broadness, while that of U. insperatus was slenderer. Eobasileus and Tetheopsis have skulls which are relatively longer and slenderer than U. ancepsTemplate:'.

The nasal bones of Uintatherium are very long, comprising roughly half of the total length of the skull. They project far enough that they completely overhang the external nares. While an elephant-like trunk or proboscis was suggested early on, based on alleged affinities to proboscideans, the structure of the ethmoturbinal bones of the nasal passage and the structure of the olfactory nerves suggest that no such structure existed. In its place, there may have been a flexible upper lip, similar to that of modern rhinocerotids. The frontal bones were large, almost as wide as long, though considerably shorter than the nasals. Uintatherium had large zygomatic arches, of which the maxilla comprised the anterior (front) portion, similar to proboscideans. Like other dinoceratans, the skull of Uintatherium lacked a postorbital process. At the back of UintatheriumTemplate:'s skull was a very large occipital crest, extending posteriorly (rearward) further than the occipital condyles. To either side of the occipital crest sat a pair of very large parasagittal crests. In some specimens, the lacrimal, the bone anterior to and above the orbits (eye sockets), was distally (outwardly) expanded, overhanging the zygomatic arches; in others, those formerly referred to Loxolophodon, the zygomatic arches projected beyond them. The parietal bones of Uintatherium, like those of its close relatives, were tightly fused, and they bear a distinct transverse ridge which strengthens that overall portion of the skull. The occiput, overhung by a large occipital crest, was rectangular in outline (though was subject to a degree of individual variation), and bore deep concavities where powerful neck muscles and ligaments would have attached.

Much like other dinoceratans, UintatheriumTemplate:'s skull was adorned with a series of well-developed outgrowths, sometimes called horns, three pairs in total. The first pair sits at the front of each nasal, and differs in form between specimens: in some, these protrusions are small and deflected upward and outward, while in others, they are larger and more horizontal. In U. insperatus, it is slightly longer and more triangular, and the portion of the nasal anterior to it is slightly longer. The second pair, above the maxillae, sits directly above the diastema (gap) separating the canines and premolars. The last, the so-called parietal horns, sits far anterior to (in front of) the occipital bone, on the parasagittal crests. This differs from the related Eobasileus and Tetheopsis, in which the parietal horns are closer to the occipital. Furthermore, in those two genera, the maxillary set of horns sits above the premolars, meaning the portion of the snout anterior to the maxillary horns is far longer; in Uintatherium, the portion of the snout anterior to the maxillary horns is fairly short. In U. anceps, the maxillary and parietal horns projected outward slightly, while in U. insperatus, they were essentially erect. Despite their description as horns, it is unlikely that any of these outgrowths were cornified (reinforced by keratin), as there is no evidence of the vascularization necessary for a keratinous covering. It is likely that they were covered only by skin. Nevertheless, Othniel Charles Marsh noted damage to several Uintatherium horn cores, likely inflicted while the animals were still alive, suggesting that they used their horns in agonistic behaviors. Like other animals with extensive cranial ornamentation, UintatheriumTemplate:'s skull was lightened by well-developed sinuses, though not to the same extent.

File:Dinocerata - a monograph of an extinct order of gigantic mammals (1886) (20768915928).jpg

Mandible of Uintatherium, minus the lower incisors

Projecting from the anteroventral (towards the front and at the bottom) portion of UintatheriumTemplate:'s mandibles (lower jaws) are a pair of large flanges. In most specimens, these would have provided support to the large upper canines, though specimens formerly referred to Loxolophodon had smaller flanges which did not extend as far. It has been suggested that the observed difference in flange size is the result of sexual dimorphism, with larger-flanged jaws belonging to males. Similar structures are observed in the related Bathyopsis. Flanges aside, the lower jaw of Uintatherium is fairly slender. Unlike most other ungulates, the condyles are deflected posteriorly, likely to accommodate the large upper tusks: without such a modification, the jaws would be unable to fully open. This condition is otherwise only seen in some marsupials and members of the former order Insectivora. The mandible's coronoid process is large, curves posteriorly, and is pointed dorsally (at the top). The mandibular condyles are small and convex, and sit slightly above the level of the cheek teeth. Below the condyles, the posterior border of the mandible is very rough, due tothe attachment of the pterygoid muscles.

Endocast anatomy

UintatheriumTemplate:'s brain was the smallest, proportionally, of any mammal which Othniel Charles Marsh was aware of, such that he stated that "it could apparently have been drawn through the neural canal of all the pre-sacral vertebrae [those preceding the sacrum]". Harry J. Jerison, in 1979, estimated its weight as Template:Convert based on the size of its endocast. The olfactory bulbs, the parts of the brain dedicated to processing smells, were very large. Dorsally, the two cerebral hemispheres are only weakly differentiated.

Dentition

File:Dinocerata (Pl. XVIII) (7158996244).jpg

Upper cheek teeth of Uintatherium, from below (left) and left lateral view (right)

Uintatherium has a dental formula of Template:DentalFormula,Template:Efn though one early record provided a dental formula of Template:DentalFormula. Analysis of its tooth enamel has demonstrated the presence of oblique ("zigzag") lines, similar to those observed in many other Paleogene herbivores, including Coryphodon, as well as more omnivorous and carnivorous taxa such as entelodontids and hyenas; most modern eutherian mammals have enamel reinforced by Hunter-Schrerger bands instead. Uintatheriids in general lack upper incisors, and Uintatherium is no exception. The loss of the upper incisors likely indicates the presence of a firm elastic pad on the ventral portion of the premaxilla, similar to that of ruminants. The lower incisors are bilobate, bearing crowns which are split into two distinctive cusps. The lower canines were somewhat incisiform, meaning that they resemble conventional incisors, while the upper canines are large and have been compared to sabres. Eobasileus and Tetheopsis have similar canines. The size of the canines, as with their supporting flanges, appears to have been sexually dimorphic, and they may have served a display function or been used in defense. Between the canines and cheek teeth, there is a large gap, the diastema. Behind the diastema on both upper and lower jaws are three premolars and three molars, all of which were fairly smalland brachyodont, meaning they have short crowns and well-developed roots; Horace Elmer Wood, in 1923, described them as "inadequate-appearing". The first upper premolar appears to have completely disappeared, with only the occasional preservation of the alveolus (tooth socket); reduced first premolars, on both upper and lower jaws, are a diagnostic trait of dinoceratans. Whether or not the first lower premolar is retained in Uintatherium is uncertain, as some sources report it as present, while others report it as absent. The third lower molar is very short, with reduced ectoconid and hypoconulid crests. The paraconids and paracristids of all teeth from the third upper premolar to the second upper molar are greatly reduced. As a whole, it has been noted that UintatheriumTemplate:'s dentition is intermediate between that of Bathyopsis and Eobasileus: the former taxon has smaller upper canines, less incisiform lower canines, and less strongly bilophodont cheek teeth than Uintatherium, while the latter has more extreme developments of those traits. This is part of the reason why an evolutionary sequence between the three genera has been proposed.

Vertebral column

File:Dinocerata (Pl. XXVI) (7159056536).jpg

Multi-angle rendition of the first (1–5) and last (6–10) lumbar vertebrae of Uintatherium

Uintatheriines as a whole are characterised by their heavy and robust skeletons, often historically compared to proboscideans, though compared by Turnbull to hippopotamuses. With the exception of parts of the skull, the known parts of UintatheriumTemplate:'s skeletal were solid, a condition known as pachyostosis. UintatheriumTemplate:'s neck was overall quite similar to proboscideans. It was also similar to that of Eobasileus, though in that genus, the neck was considerably shorter. The first cervical (neck) vertebra, the atlas, and the second cervical vertebra, the axis, are particularly proboscidean-like. The atlas in particular is massive, while the axis is short and robust. The rest of the cervical ceries is more elongated than in proboscideans, though still short in relation to the axis. The vertebral centra are taller than they are long, and are in turn wider than they are tall. All of the dorsal (back) vertebrae are opishthocoelous, convex anteriorly and concave posteriorly; the same condition is seen in proboscideans, but in them, it is more extreme. The first thoracic vertebra has a fairly small neural spine and short transverse processes. Further back in the thoracic column, the vertebrae are much larger, and have bigger neural spines. The lumbar vertebrae have wedge-shaped centra and weak, laterally-compressed neural spines, with thin transverse processes. Four vertebrae were present in the sacrum. Only four of UintatheriumTemplate:'s caudal (tail) vertebrae are known. They were bore long, narrow centra, decreasing in size the more posterior they are. Despite their relative slenderness, the caudal vertebrae are quite broad in comparison to those of proboscideans. The ribs of Uintatherium also resembled proboscideans, and have been compared to mastodons specifically, while the sternum more closely resembles certain artiodactyls.

Limbs

As with much of the postcranial skeleton, UintatheriumTemplate:'s forelimbs and hind limbs, and the pectoral and pelvic girdles respectively, were very convergent with proboscideans. Like in other terminal dinoceratans, the long bones (i.e. humeri and femora) were abnormally thick and dense, a condition known as pachyostosis.

Front limbs

File:Dinocerata (Pl. LIV) (7159107224).jpg

Comparison between the forefoot (below) and hindfoot (above) of Uintatherium.

The scapula (shoulder blade) of Uintatherium resembles that of proboscideans, though is less developed above the glenoid fossa. The humerus (upper arm bone) is fairly short and massively built. Its great tuberosity is slightly compressed and does not extend above the humeral head. The lower portion of the humerus resembles that of rhinocerotids. The radius and ulna are essentially equal in size. The ulna has a small face where it articulates with the lunate, again similar to proboscideans. Where UintatheriumTemplate:'s forelimbs differ the most from proboscideans are the manus (forefeet), which each bear five digits. There are eight carpal (wrist) bones, which interlock, similar to perissodactyls. UintatheriumTemplate:'s scaphoid bone is somewhat like elephants, though is shorter and stouter, and has a rounded proximal (near) end. The smallest bone of the carpus is the trapezoid. Unlike elephants and proboscideans, the unciform bone articulates with both the cuneiform and lunar bones. The phalanges (digit bones) are short, and grew increasingly rugose distally (away from the centre of the body). Overall, UintatheriumTemplate:'s manus anatomy somewhat resembled that of the pantodont Coryphodon. In life, it is likely that all four of UintatheriumTemplate:'s appendages bore fleshy pads like those of elephants, and were somewhat columnar in shape.

Hind limbs

UintatheriumTemplate:'s pelvis is very large, with a sub-oval outline, only superficially resembling that of proboscideans. Its width suggests that it supported a greatly enlarged hindgut. The femur (thighbone) is fairly short, lacked a pit to accommodate the round ligament, and had a great trochanter which was flat and recurved. Distally, the femur was more strongly laterally compressed than in a proboscidean. It has femur condyles around the same size. In life, Uintatherium would have held its hind leg essentially straight, as in elephants and humans. The patella (kneecap) is oval-shaped. The fibula is slender, with prominent articular faces for the elements of the tarsus (ankle and foot). The astragalus, or talus, is more like perissodactyls than proboscideans, in that its anterior portion has articular faces for both the cuboid and navicular bones. UintatheriumTemplate:'s pes (hind foot) has four well-developed digits, and a fifth which is smaller and less well-developed. Though smaller, the pedal anatomy is otherwise similar to the manus. Similarities to proboscideans have been noted, though, unlike their semi-plantigrade gait, UintatheriumTemplate:'s pedal anatomy is fully digitigrade.

Size

Uintatherium anceps was stated by Marsh to have stood roughly four-fifths the height of Eobasileus, so about Template:Cvt at the shoulder. In 1979, Harry J. Jerison provided a body length of Template:Convert, while in 2002, an average body length of Template:Convert, based on three mounted specimens, was provided by William D. Turnbull. A plethora of body mass estimates have been proposed for the genus over the decades. In a 1963 work, Harry J. Jerison provided various mass estimates for a multitude of Palaeogene taxa. An average of two estimates resulted in a mass of Template:Convert, while the use of scale models resulted in a range of Template:Convert. John Damuth, using head–body length and data from teeth recovered considerably a smaller body mass of Template:Convert. Using Jerison's methods and additional data provided by Damuth, in 2002, Turnbull proposed an estimate of Template:Convert. He recovered larger masses in other analyses, though expressed his belief that these were overestimates due to the methodologies applied. Nevertheless, in 1998, Spencer G. Lucas and Robert M. Shoch provided an even larger body mass of Template:Convert for U. anceps. The size of U. insperatus is not certain, though it is believed to have been smaller.

Paleoecology

Diet and lifestyle

Like other uintatheriids, the molars of Uintatherium were bilophodont (two-ridged). Cheek teeth with this morphology often belong to browsing (feeding on leaves, shoots and twigs of relatively high-growing plants) animals. It has therefore been suggested that Uintatherium adopted a similar lifestyle. However, in 2002, Turnbull suggested that it, and other late-stage dinoceratans, were more ecologically analogous to hippopotamuses, citing traits such as pachyostosis, short legs, and a barrel-shaped ribcage as supporting evidence. As C₄ grasses, on which hippopotamuses often feed, became widespread only fairly recently, and dinoceratan teeth were not suited for grazing, he noted that they likely fed quite differently to hippopotamuses. Whereas most modern ungulates ferment plant matter in their foregut, Turnbull suggested based on pelvic anatomy that Uintatherium was instead a hindgut fermenter, similar to proboscideans and perissodactyls. He further proposed that late-stage dinoceratans had digestive systems analogous to sirenians (sea cows). If this model is accurate, the processing of food would have occurred primarily in the hindgut, reducing demands on the cheek teeth and resulting in the "inadequate appearance" observed by Wood.

Paleoenvironment

File:Early Eocene proxy ensemble data from fossil localities showing (a) MAT and (b) MAP estimates.png

Map of the Northern Hemisphere during the Eocene showing mean annual temperature and precipitation of various locations

Uintatherium evolved during a period in Earth's climatic history called the Paleocene-Eocene thermal maximum. This period saw some of the highest average temperatures in Earth's history with temperatures in Colorado (where Uintatherium fossils have been found) reaching an annual average of Template:Convert—much higher than today where the mean annual temperature in Colorado is only around Template:Convert. Although global average temperatures declined throughout the Eocene, the average temperatures in North America remained relatively consistent for the first half of the period, and only cooled slightly towards the end of the Eocene. North America did see considerable climatic developments dutring the course of the Eocene in spite of the relatively constant regional average temperatures. The uplifting of the Rocky Mountains and their associated volcanism lad to considerable drying in the North American interior. The arid scrublands which characterize the western United States today (as exemplified by Arizona, Nevada, and New Mexico) began to emerge during this period.

When Uintatherium first appeared in North America, most of the continent was covered primarily closed-canopy forests. This environment is exemplified by the Bridger Formation, which consisted of inland lakes surrounded by dense forests. This is inferred by the abundance of plant fossils and the presence of a great diversity of primate fossils, which are predominantly arboreal. Fossils of redwoods, elms, and birch trees are known from throughout North America during this period, suggesting that the amount of precipitation did not vary considerably across latitudes. Most of North America was likely covered by temperate forests and temperate rainforests. Even organisms more typically adapted to low-latitude environments, such as palm trees and crocodylians have fossils preserved as far North as Alaska and Ellesmere Island, exemplifying the extreme climatic conditions of the early and middle Eocene.

By the time of the Uinta Formation, the landscape had changed considerably. The large lakes emblematic of the earlier Eocene had shrunk, and the majority of deposition was the product of low-volume streams. Insectivorous and frugivorous mammals (especially primates) declined in diversity alongside a rise of folivorous artiodactyls, which is interpreted as reflecting an increase in more open habitats resulting in a gradual decline in tree cover. Considerable forests existed, likely alongside the numerous waterways, but these were probably interspersed by open savannah environments. This trend towards aridifcation was facilitated by a general decline in the amount of precipitation in North America while average annual temperatures remained high. It would not be until the later parts of the Eocene that the global cooling began to affect North American ecosystems, by which point, Uintatherium was already extinct.

U. inseparatus appeared in Asia during the middle part of the Eocene. Its fossils are known from the Lushi Basin in China, which consisted of large, deep lakes that preserve fossils of bivalves and gastropods. These lakes were surrounded by forests and swamps and were interspersed by semi-arid steppe. Variations in sea-levels and intermittent flooding at the time also produced brackish lakes and swamps. The inland lakes varied in size over the course of the middle Eocene before eventually disappearing completely and being replaced by rivers and floodplains.

Contemporary fauna

North America

File:A History of Land Mammals in the Western Hemisphere Fig. 48.jpg

A map of North America during the Eocene, with modern borders shown

Uintatherium anceps is known from various strata from the Bridgerian and Uintan North American land mammal ages. This corresponds to the interval between 50.5 and 39.7 million years ago—a span of just over 10 million years within the Eocene. The oldest remains confidently assigned to this species are from the faunal zone "BR3" of the Bridger Formation, which is at the end of the Bridgerian land mammal age.

In the Bridger Formation, U. anceps coexisted with a variety of primitive ungulates including helohyids, homacodontids, brontotheriids, amynodontids, and hyopsodontids. The environment was also host to some of the ancestors of modern perissodactyl groups including Hyrachyus (a primitive relative of rhinos), Helaletes (an early relative of tapirs), and several species of Orohippus (a primitive horse). North America at the time also had a diverse assemblage of early primates including Microsyops, Notharctus, Smilodectes, and the members of Omomyidae (relatives of modern tarsiers). Mammalian predators of the region included mesonychids like Mesonyx and Harpagolestes, hyaenodontids like Limnocyon and Sinopa, oxyaenids like Patriofelis and Machaeroides, and early carnivoran-relatives like Miacis and Vulpavus. A variety of more enigmatic mammal forms were also present including members of Tillodontia, Stylinodontidae, and Pantolestidae and the small insectivorous Apatemys and Metacheiromys. Primitive sciuromorph rodents, leptictids, and eulypotyphlans coexisted with the metatherians Herpetotherium and Peradectes.

Reptiles were also abundant in this environment. Fossils from turtles including softshelled turtles, tortoises, terrapins, and baenids lived alongside anguids, varanids, teiids, and boids as well as crocodilians like Boverisuchus and Borealosuchus. Remains of primitive owls and cranes have also been found.

File:Uintatherium and Orohippus by Knight.jpg

A pair of Uintatherium depicted with the contemporary equid Orohippus

In the transition from the Bridgerian to the Uintan, several of these animals became extinct and new forms emerged. The oxyaenids and phenacodontids disappeared during this transition and new groups like the oromerycids and the earliest chalicotheres (the eomoropids). This transition is followed by the appearance of several medium and large ungulate genera including Protylopus, Amynodon, and Eobasileus. This faunal subinterval is represented by the Devil's Graveyard Formation and has been argued to be a distinct land mammal sub-age (the "Shoshonian" or "UI1b biochronological zone"), although this is not universally accepted. This transition also saw a marked decline in primate diversity in North America, which would continue throughout the Eocene until primates eventually became extinct in North America.

The middle-Uintan land mammal age (sometimes called "UI2" biochronological zone) is the most recent interval from which fossils of U. anceps are known. This corresponds to the eponymous Uinta Formation. This interval saw the diversification of brontotheres, helohyids, and rhinocerotoids as well as the emergence of the first protoceratids, agriochoerids, and camelids. It also saw the extinction of North American cimolestans and leptictids as well as most of the remaining North American primates, with only the omomyids remaining extant. Primitive carnivoramorphs like Miocyon also emerged. The end of this interval saw the final extinction of Uintatherium in North America alongside other long-lived genera such as Mesonyx and Hyrachyus.

Asia

File:Mollweide Paleographic Map of Earth, 45 Ma (Lutetian Age).png

A map of the world during the Middle Eocene, with modern borders shown

The second species of Uintatherium, U. inseperatus, lived in the Lushi Formation what is now Henan, China during the Middle Eocene. The precise age of the fossils assigned to this species are uncertain, but they have been estimated to be between 48 and 37 million years ago, which is roughly contemporaneous with the existence of U. anceps in North America. This corresponds to the Sharamurunian Asian land mammal age, which lasted for about the same length of time. Remains assigned to U. inseperatus have also been found in the similarly-aged Uqbulak Formation in the Junggar Basin.

The composition of Asian land mammal assemblages was similar in several ways to the contemporary assemblages in North America, although the precise timing of faunal turnover is not as well studied with respect to Eocene ecosystems in Asia. The carnivorous mammals of the continent were generally similar, with mesonychids, haplodectids, hyaenodontids, and the carnivoramorphan Miacis being the most abundant predators. However, several endemic carnivores coexisted with these including Eusmilus (an early nimravid), Cynodictis (a primitive amphicyonid), and the controversial carnivorous ungulate Andrewsarchus. Prey for these animals included a diverse array of terrestrial ungulates including late surviving members of Paleocene lineages such as the coryphodont Eudinoceras, dichobunids, tillodontians, and taeniodontans. Ungulate groups common in North America were also represented, including Hyrachyus as well as the helohyids, brontotheriids, helaletiids, and amynodontids. They were accompanied by a diverse array of perissodactyls, which underwent a radiation in Asia during the Middle Eocene. These new groups included the paraceratheriids, hyracodontids, chalicotheriids, and deperetellids. The artiodactyl anthracotheres also first evolved in Asia during this period.

Further reading

• Academy of Natural Sciences
• National Park Service
• Wood, Horace Elmer 1923, The problem of the Uintatherium molars, Bulletin of the American Museum of Natural History; v. 48, article 18
• Fossil Evidence – Eocene: Smithsonian National Museum of Natural History