Amynodon

Description of A. advenus

Different fossils interpreted as type specimens of A. advenus

YPM 11763, an isolated third upper molar

YPM 11453, a skull

Amynodon is the best represented amynodont in the fossil record.^[10] The first fossils of Amynodon were briefly reported by Othniel Charles Marsh in 1875 from the Uinta Basin (Uinta Formation) in Utah.^[10] Marsh recorded several fragmentary fossils, most notably an isolated upper P₃ (third upper molar;^[2] YPM 11763^[10]) and a corresponding lower molar.^[2] Marsh believed that the fossils came from a true rhinoceros and named the new species Diceratherium advenum based on the material. The species was mainly based on the upper molar, for which Marsh provided measurements.^[2]

In 1877, Marsh named the new genus Amynodon based on a nearly complete skull^[1] (YPM 11453^[10]), also from the Uinta Formation in Utah.^[10][11] The type species was A. advenus, "which was provisionally referred to the genus Diceratherium when first described."^[1] Marsh believed the animal to be the earliest rhinoceros yet discovered, with a skull intermediate between rhinoceroses and tapirs but with molars entirely like those of rhinoceroses.^[1] Marsh's description was wrong on several anatomical details since the skull had not been removed from its matrix (surrounding rock) or properly put together before it was described.^[12]

Marsh did not designate a type specimen, a problem compounded by the use of different fossils to diagnose Diceratherium advenum in 1875 (YPM 11763) and Amynodon advenus in 1877 (YPM 11453).^[10] In 1890, Henry Fairfield Osborn interpreted YPM 11453 as the type specimen of A. advenus.^[12] In 1921, Edward Leffingwell Troxell instead interpreted the upper molar YPM 11763 as the type specimen.^[11] Later authors generally follow Troxell's assessment.^[10][13] In 1955, Jean Hough considered YPM 11453 to be a paratype specimen.^[14]

Marsh did not provide etymologies for the species name advenus^[2] or the genus name Amynodon.^[1] The American Museum of Natural History traces the etymology of Amynodon to the Greek ἀμύνω (amynos, to ward off or threaten) and ὀδούς (odontos, tooth) and interprets the name as a reference to the large, tusk-like canines.^[15] The species name advenus is Latin and can refer to a foreigner or stranger, something that is foreign or strange, or someone who is a stranger to a specific thing (i.e. inexperienced).^[16]

Description of Orthocynodon

In 1882, William Berryman Scott and Osborn described the new genus and species Orthocynodon antiquus, based on an associated skull and lower jaw^[17] (PU 10047^[10]), and another partial skull,^[17] found in the Washakie Formation^[10][18] in Wyoming.^[17] Scott and Osborn distinguished the fossils from Amynodon due to several features of the teeth; Orthocynodon was described to have more erect canines, a small postero-internal (internal, towards the back) cusp on the upper P₂ and P₃ (second and third premolar in the upper jaw), and generally more molariform (molar-like) premolars than Amynodon.^[17] Orthocynodon was also Bridgerian in age, and thus predated Marsh's fossils. Scott and Osborn presented it as the new oldest known rhinoceros.^[17] In 1883, Scott and Osborn described O. antiquus in more detail and established the new family Amynodontidae to contain Orthocynodon and Amynodon.^[18]

YPM 11453 was eventually prepared out of its matrix, and Marsh allowed Scott and Osborn to examine the skull and an associated lower jaw.^[12] Scott and Osborn found several features which they had believed made Orthocynodon distinct, such as erect canines.^[11] They designated Orthocynodon as a probable synonym of Amynodon in 1887^[19] and in 1890, Osborn treated O. antiquus as a species of Amynodon (A. antiquus).^[12] The only remaining features believed to distinguish A. antiquus as a species were in the premolars; according to Osborn, A. antiquus retains the lower P₁ (the first premolar), a tooth lost in A. advenus, and has a more molariform P₄.^[10]

Additional fossils and species

In 1890, Osborn described a third partial Amynodon skull^[12] (PU 10309^[10]), from the Uinta Formation, referred to a new species, A. intermedius.^[12] Osborn distinguished A. intermedius from A. advenus by its larger size, the retention of the upper P₁, and the procumbent position of the canines.^[12] Osborn believed that the three described Amynodon species formed an evolutionary sequence, with A. antiquus evolving into A. advenus, which then evolved into A. intermedius.^[12] In 1895, Osborn reported the discovery of a relatively complete Amynodon skeleton in the Uinta Formation, which he referred to A. intermedius.^[20] Osborn appears to never have figured these fossils.^[13]

Earl Douglass described a poorly preserved jaw (CM 734^[13]) in 1903, from Sage Creek in Montana.^[21] Douglass referred the jaw to Metamynodon^[21] but Horace Elmer Wood identified it as Amynodon, possibly a new species, in 1934.^[22] According to Wood, CM 734 is more advanced than A. antiquus and similar in size to A. intermedius, but distinguished by having one less pair of upper incisors.^[22] In 1910, Walter W. Granger reported a new Amynodon skull (AMNH 14601^[13]) from Beaver Divide in Wyoming.^[23] Walter initially referred the skull to A. antiquus^[23] but later changed his mind to A. advenus.^[13] Elmer S. Riggs reported new Amynodon fossils from the Uinta Basin in 1912 but did not describe them.^[24] The fossils found by Riggs included a skull (FMNH 12184) and a lower jaw (FMNH 12191), which were later referred to A. advenus.^[13] In 1919, Olof August Peterson [sv] reported Amynodon fossils from the Uinta Basin, from several individuals. Peterson referred the fossils to both A. advenus and A. intermedius, almost entirely based on size.^[25]

In 1921, Troxell suggested that YPM 11453 was distinct from YPM 11763 on the species level. All of the features that Troxell considered to be distinguishing are in the upper P₃ since YPM 11763 preserves nothing else. These features included that the P₃ of YPM 11453 is smaller, that both its posterior and anterior cingula are stronger, that the postsinus (gap behind the posterior cingulum) is deeper, and that the medisinus (gap between the two cingula) is much narrower.^[11] Since Troxell interpreted YPM 11763 as the type specimen of A. advenus, he moved YPM 11453 to a new species, A. erectus.^[11] Troxell distinguished A. erectus from A. antiquus by A. erectus having lost P₁ in both jaws and its smaller size, though noted that both species had similar erect canines; the retention of P₁ in A. antiquus was considered so significant by Troxell that it might justify Orthocynodon as a subgenus.^[11] A. erectus was distinguished from A. intermedius by A. intermedius's larger size and more procumbent canines.^[11]

The description of the new genus Amynodontopsis by Chester Stock in 1933 was the first record of amynodonts from California. Stock proposed that A. intermedius could belong to Amynodontopsis because A. intermedius shared some features with the new species, notably the depth and backward extent of the fossae (grooves in front of the eye orbits).^[26] In 1939, Stock described the new species Amynodon reedi, also from California.^[7] The type specimen of A. reedi is a fragmentary left maxilla, LACM (CIT) 314/2529,^[10] which preserves the upper P₄–M₃.^[7] Stock stated that the fossil was from "Poway conglomerate" in San Diego, California;^[7] modern stratigraphic revisions of the region place this in the Friars Formation.^[13] Stock compared the maxilla to a small Amynodon fossil found in the Uinta Formation, AMNH 1936A;^[13][7] some later authors have misinterpreted this as A. reedi also being present in the Uinta Formation.^[13] A. reedi was distinguished from the other species by Stock through its small size, and by P₄ being narrow, with a metaloph at a right angle to the tooth's external border.^[7] Stock also reported an Amynodon tooth, possibly A. advenus, from the Sespe Formation in California.^[7] David J. Golz and Jason A. Lillegraven reported additional Californian Amynodon fossils in 1977, from the Santiago Formation and referred these to A. intermedius.^[27]

Revisions

In 1981, John Andrew Wilson and Judith Schiebout described a large number of Amynodon fossils from Uintan-age strata at Pruett Tuff in the Buck Hill Group of Brewster County, Texas. Wilson and Schiebout referred the fossils to A. advenus.^[13] As the largest published sample of Amynodon from a single population, the assemblage allowed Wilson and Schiebout to study the degree of individual variation in a single Amynodon species. High variability was observed in several features: overall body size varied considerably, as did canine position (erect/procumbent/oblique), strength of the cingula, and the depth of the fossae.^[13] Some of the previously described Amynodon material, attributed to A. antiquus, A. erectus, A. advenus, and A. intermedius, was examined for comparison and was determined to show less variability than the single-population sample from Texas, despite being from different sites and slightly different stratigraphic levels (early/late Uintan).^[13] Wilson and Schiebout thus suggested that A. antiquus, A. erectus, and A. intermedius were all synonyms of A. advenus.^[13]

William P. Wall revised Amynodon in 1982.^[10] Wall recognized only A. advenus and A. reedi as valid, and agreed with Wilson and Schiebout's assessment that A. antiquus, A. erectus, and A. intermedius are junior synonyms of A. advenus.^[10] Wall pointed out that Wilson and Schiebout had made some mathematical errors in their analysis; contrary to their calculations, the Texan Amynodon sample shows less variability than the total sample.^[10] The species other than A. advenus were reassessed by Wall individually, which resulted in the designation of all except A. reedi as synonyms:

• The distinguishing features of A. antiquus, the retention of P₁ and the molariform P₄, are not actually present in the type specimen. The type jaw is broken towards the back, and Scott and Osborn misinterpreted P₂–M₃ as P₁–M₂. This means that P₁ is not actually retained, and that the supposed molariform premolar is just a molar.^[10]
• Troxell's features to distinguish A. erectus from A. advenus were assessed as "quite minor", and the A. erectus fossils as "well within the range" of A. advenus.^[10]
• The dental features used to separate A. intermedius from A. advenus were found to be dubious. P₁, used in the diagnosis of the species, is not preserved in the type specimen and is present only as a plaster reconstruction. Osborn's claim that the canines of A. intermedius are unusually large and procumbent is also unsupported, since the canines of the type specimen are broken and their original position is indeterminable. Variations in canine size is also explainable via sexual dimorphism. A. intermedius has also been separated from A. advenus by its larger size and its stratigraphic position—A. intermedius is from the "Uinta C" level whereas A. advenus is from "Uinta A" and "B"—but several fossils from Uinta C are in the smaller size range of A. advenus.^[10]
  • Stock's suggestion that A. intermedius could be a species of Amynodontopsis was rejected by Wall, who pointed out that the supposed greater extent and depth of the fossae was due to the A. intermedius type skull being crushed. This is easily observed in that the fossae only extend further than normal on one side of the specimen.^[10]
• A. reedi was recognized as distinct due to its earlier age and its apparently more primitive cheek teeth. There are no intermediate fossils between A. reedi and A. advenus, which further supports A. reedi's validity.^[10]

In 2018, Jérémy Tissier et al. recovered A. reedi as only distantly related to A. advenus in a phylogenetic analysis. A. reedi was recovered in a much more derived position as the sister taxon of Zaisanamynodon. Tissier et al. suggested that A. reedi is more similar to Zaisanamynodon than to A. advenus, for instance in the lack of a lingual cingulum on the upper molars, and that it differs from Zaisanamynodon only in size and one dental feature, that the hypocone on P₃ and P₄ is anterior to the metacone.^[28] The authors proposed that A. reedi was not a species of Amynodon and should be referred to as "Amynodon" reedi until its fossils are reexamined and assessed.^[28]

Purported discoveries outside North America

Asia

Several species of Amynodon have been named based on fossils found in East Asia.^[29] In his revision of Amynodon, Wall restricted the use of the genus to fossils from North America. As Amynodon is currently understood, none of the Asian species thus belong to the genus.^[10] Wall recommended that the Asian material should be treated as Amynodontidae incertae sedis^[10] and detailed reexaminations of most of the Asian fossils are still necessary.^[29]

The best known Asian species is Amynodon mongoliensis, described by Osborn in a posthumous publication in 1936. A. mongoliensis was based on fossils from the Shara Murun Formation, at Ula Usu in Inner Mongolia.^[30] Osborn compared the fossils to Amynodon, especially A. intermedius.^[30] A. mongoliensis was reassigned to its own genus, Sharamynodon, by Miklós Kretzoi in 1942.^[31] Kretzoi's Sharamynodon was not recognized as a distinct genus by Vera Gromova in 1954, nor by paleontologists in China.^[32] The generic distinction of Sharamynodon was reaffirmed by Wall in 1989, and has been largely accepted since.^[32] In both 1982 and 1989, Wall suggested that some of the other Asian Amynodon species could belong to Sharamynodon instead.^[10][33]

A. watanabei is the earliest described Asian species, described by Shigeyasu Tokunaga in 1926 as Aceratherium? watanabei, based on a partial lower jaw from Ube in Nagato, Japan. Of the teeth, only P₃, P₄, and M₁ are preserved, complicating classification.^[9] Various other generic identifications were proposed for the fossils over the following decades, such as Rhinoceros and Chilotherium.^[34] In 1945, Fuyuji Takai suggested that A. watanabei was an amynodont, perhaps Cadurcotherium. Takai reclassified A. watanabei as a species of Amynodon in 1948, having compared its proportions to A. erectus.^[34] Tokanuga believed that A. watanabei was "not older than the Miocene"^[9] but the fossils are probably Middle or Late Eocene.^[32]

In 1930, Otto Zdansky described Amynodon sinensis,^[8] based on Late Eocene fossils from Henan, China.^[32] The fragmentary^[32] fossils include parts of lower and upper jaws, and teeth.^[8] Additional fragmentary fossils were referred to A. sinensis by Yang Zhongjian in 1937,^[35] and by Zhou Mingzhen and Xu Yu-Xuan in 1965.^[36] Zhou and Xu reclassified A. sinensis to the genus Sianodon^[36] but Sianodon is generally not considered a valid genus, and A. sinensis falls outside both Sianodon (as traditionally understood) and Amynodon in phylogenetic analyses.^[32]

In 1958, Gromova described Amynodon giganteus,^[37] from the Ergilin Dzo Formation in Mongolia.^[38] A. giganteus has since been identified as a specimen of Gigantamynodon cessator.^[38][39]

Xu and Chiu Chan-siang described Amynodon altidens in 1962, based on a fragmentary right lower jaw from Lunan, Yunnan. The only teeth preserved are M₂ and M₃. Intermediate in size between the other Asian "Amynodon" A. sinensis and A. (Sharamynodon) mongoliensis, Xu and Chiu distinguished A. altidens from both by its higher tooth crowns, lengthened M₃, distinct longitudinal grooves on the external surfaces of the molars, and more slanted transverse ridges.^[3] In 1964, Zhou, Xu, and Zhen Shuo-nan described Amynodon lunanensis, based on a skull and associated jaws from Hsiaogotze-Baiyaoshan, five kilometers north of Lunan. A. lunanensis was destermined to be distinct from A. sinensis due to its larger size, the relatively shorter upper premolars, and its lower molars having subtle longitudinal grooves on their outer walls and like A. altidens, more slanting transverse ridges. A. lunanensis was smaller than Sharamynodon mongoliensis, from which it was also distinguished by relatively narrower third molars but broader second molars, larger canines and shorter premolars, and the more slanting transverse ridges on the lower molars.^[6] In 1975, Qi Tao described A. alxaensis, based on a partial skull from Alxa Zuoqi, Ningxia. A. alxaensis was noted to be similar to Cadurcodon ardynensis and "Sianodon" bahoensis (=Cadurcodon), but most similar to Amynodon in its dental formula. Referral to Amynodon was also supported by similarities to Sharamynodon mongoliensis in dental features, the short face, and body size.^[4]

In 1963, M.D. Biryukov described Amynodon tuskabakensis based only on an isolated M₃ found near Lake Zaysan in Kazakhstan. In 1971, Elizaveta Belyaeva [ru] referred A. tuskabakensis to the genus Hypsamynodon. In 1996, Spencer G. Lucas and Robert J. Emry suggested that the species was a synonym of Cadurcodon ardynensis, found close to the A. tuskabakensis type locality.^[40] In 2017, Alexander Averianov and colleagues identified A. tuskabakensis as a species of Cadurcotherium through shared derived characteristics.^[32]

Skull

Skull size varies between Amynodon specimens. PU 10057 is 44 centimeters (17 in) long and YPM 11453 is 35 centimeters (14 in) long.^[13] 41576-19, one of the skulls described by Wilson and Schiebout in 1981, is 51 centimeters (20 in) long.^[13] The skull of Amynodon was relatively dolichocephalic (long relative to its width); the skull was about as wide as 49–54% of its length.^[10] This sets Amynodon apart from more derived amynodonts. In metamynodontines, the skull is brachycephalic (wide relative to its length) due to enlarged zygomatic arches, with the width ranging between 55–70% of the skull length. In cadurcodontines, the skull is even more dolichocephalic, only about 45–50% as wide as it is long.^[10]

Cadurcodontine amynodonts have a noticeably higher skull, and metamynodontines have elevated eye orbits; neither of these traits are present in Amynodon.^[10] The zygomatic arches did not have any noteworthy adaptations, contrary to those of caducodontines (reduced in height and thickness) and metamynodontines (increased in size).^[10] The premaxillae are not thickened or fused with each other, as in some more derived amynodonts, but the premaxillae extend relatively far back, below the nasal bones. Extensive contact between the premaxillae and nasals such as this is a primitive trait, seen in more primitive rhinocerotoids but not in more derived amynodonts.^[10] Amynodon had large preorbital fossae (grooves in front of the eye orbits).^[10]

A major trend observed in amynodont evolution is the shortening of the snout compared to more primitive rhinocerotoids; this is observed in Amynodon, which had a shorter snout than the more basal amynodont Rostriamynodon, but a longer snout than all more derived amynodonts.^[10] The nasal bones of Amynodon were relatively long, and the nasal incision extends back above the diastema (gap) between the canines and cheek teeth.^[10] The frontal sinuses of Amynodon were not as large as they were in the cadurcodontines.^[10]

The occiput of the skull was relatively narrow, and there is a distinct post-orbital constriction (constriction of the skull behind the eye orbits), due to Amynodon's small braincase.^[10] The skull had a well-developed sagittal crest.^[10]

Dentition

Side view of PU 10309

Top view of the lower jaw associated with YPM 11453

Amynodon had the dental formula 3.1.3.33.1.3.3,^[10] i.e. both the lower and upper jaws had three pairs of incisors, one pair of canines, three pairs of premolars, and three pairs of molars.^[10] The dentition of Amynodon is similar to primitive rhinocerotoids such as Heptodon and Hyrachyus, and is primitive when compared to later amynodonts.^[10] The teeth nevertheless demonstrate the early beginnings of evolutionary trends later observed among the amynodonts, including that the canines were enlarged, that the premolars had decreased in importance, and that there was an increased emphasis on shearing by the ectolophs of the molars and a corresponding simplification of the crosslophs of the molars (the ectoloph and crossloph being two prominent ridges on the surface molar).^[10]

The incisors and canines were relatively primitive in Amynodon, and the retention of three pairs of upper and lower incisors is a primitive trait.^[10] The incisors of Amynodon are more conical than in Heptodon and Hyrachyus, and vary in relative size between individuals; in some Amynodon, both the upper and lower I₂ is larger than the other incisors.^[10] The canines of Amynodon were moderately-sized semi-erect tusks, clearly larger than in more primitive rhinocerotoids but smaller than in many more derived amynodonts.^[10] The lower canines were positioned in front of the upper canines.^[10] In most Amynodon skulls, there is a short diastema (gap) in the upper jaw between I₃ and the canine, to accommodate the lower canine. This diastema is a derived trait in amynodonts, not present in more primitive rhinocerotoids. In some specimens, this diastema is not present, and the lower canine preserves a wear pattern caused by contact with I₃.^[10] The canines–premolar diastema is longer in Amynodon than in more derived amynodonts, but relatively shorter than in Hyrachyus and Heptodon due to the reduced length of the snout.^[10]

Amynodon had low-crowned cheek teeth.^[10] The upper premolar series of Amynodon was relatively short, only about half the length of the upper molar series, but still relatively longer than in more derived amynodonts. The upper P₂ was relatively small, with its long axis oriented posteriolingually (backwards and towards the tongue). The protoloph (an internal ridge joining the cusps together) of the upper P₂ is angled sharply to the back, and there is usually a small metaloph (another ridge). The hypocone (a cusp in the corner of the tooth) is either missing or small in the upper P₂.^[10] The upper P₃ is larger and more complex than the P₂, and its long axis is less posteriolingually oriented than in P₂. The P₃ had a large parastyle, an anterior rib, and a small posterior rib along its ectoloph. In most known P₃ there was also a lingual cingulum.^[10] The upper P₄ was relatively large, taking up about 42% of the length of the premolar series. In most other amynodonts, this tooth is relatively smaller.^[10] The P₄ had anterior and posterior ribs more developed than those of the other cheek teeth, as well as a developed parastyle. The metaloph of P₄ was distinct from the proloph, and relatively larger than the metaloph of P₃.^[10] The molars do not have distinct crochets (a crest on the metaloph) or crichets (a crest on the ectoloph), and only sometimes an antecrochet (a crest on the protoloph). The upper M₂ is larger than the M₁, though these teeth are otherwise similar.^[10] The parastyles of the upper molars were relatively small, but larger than in more derived amynodonts. Developments of the cusps and ridges of the upper M₃ gave the tooth a somewhat quadratic outline; this is feature unique to amynodonts, which separates them from all other rhinocerotoids.^[10]

The lower premolar series was shortened, similar to the upper one, about 45–50% the length of the lower molar series. The lower P₂ is relatively large and double-rooted. In other amynodonts, P2 is either much smaller (Sharamynodon), only sometimes there (Amynodontopsis), or completely lost (Cadurcodon).^[10] Most of the lower P₂ was taken up by the protoconid (the most prominent cusp) but the P₂ also had a small paralophid, small protolophid, and small talonid (the other cusps).^[10] The lower P₃ was much larger than the P₂. In the P₃, the other cusps were enlarged but the protoconid was still dominant.^[10] The lower P₄ had an even smaller protoconid, though it was still the largest cusp. P₄ also had a well-developed paraconid, and distinct trigonid and talonid basins (depressions, for the cusps of the upper teeth).^[10] Both the lower P₃ and lower P₄ were submolariform, i.e. somewhat similar to molars.^[10] All of the lower molars were relatively broad, though M₃ was relatively narrower than the others, and they increased in size from front to back.^[10]

Postcranial skeleton

The postcranial skeleton of Amynodon has not been described at the same level of detail as the skull or dentition.^[10][13][20] The centra of the vertebrae of Amynodon were relatively small. The anterior (forwards) dorsal vertebrae had long and slender spines.^[20] The humerus of Amynodon was similar in length to the femur, and very broad. The ulna and radius were long and slender.^[20] The hip bones were elongated and slender, the pubis was short, and the ischium was long.^[20] Like other amynodonts, Amynodon had four toes on their manus (front feet) and three on their pes (hind feet),^[45] a primitive condition among perissodactyls.^[46] The metapodials (long bones) of the manus were longer than the metapodials of the pes, and all were relatively slender.^[20] The third metacarpal was the longest of the metacarpal bones, but not by much.^[20]