Protitanotherium koreanicum Takai, 1939

Mihlbachler, Matthew C., 2008, Species Taxonomy, Phylogeny, and Biogeography of the Brontotheriidae (Mammalia: Perissodactyla), Bulletin of the American Museum of Natural History 311 (1), pp. 1-475 : 341-344

Protitanotherium koreanicum Takai, 1939

HOLOTYPE: UMUT CV6034; a partial right maxilla with P2–P3 originally deposited at the Palaeontological Laboratory, Waseda University, Tokyo. Bombing during World War II destroyed the holotype, although a photograph of it can be seen in Takai (1939: pl. 1, fig. 6 and is reproduced here as fig. 173n).

AGE: Middle or late Eocene ( Russell and Zhai, 1987).

TYPE LOCALITY: Hôsan Coal Field in the central part of Kôdai-dô, Northwest Tyôsen, ( Korea).

REFERRED SPECIMENS: (From the same locality as the holotype; see Takai [1939] for a complete list) Takai (1939) referred numerous specimens to Protitanotherium koreanicum . Takai (1939) provided no specimen numbers, although 20 of the specimens were published as photographs by Takai (1939) have been given numbers in the UMUT collection. Table 12 matches the UMUT numbers with the figures in Takai (1939).

DETERMINATION: Nomen dubium. The holotype and referred material are too fragmentary to adequately compare to other species. The referred material might represent more than one species.

DESCRIPTION

The material referred to Protitanotherium koreanicum consists of a maxillary fragment and isolated teeth. The more diagnostically useful of these specimens are pictured in figure 173; additional specimens are figured in Takai (1939). UMUT CV29046 (fig. 173a, b) is an upper incisor, probably an I2. It is rather small with a subconical crown and a thick lingual cingulum, very similar to that of Protitanotherium emarginatum or Diplacodon elatus . There are two distinctly sized canines; the larger is seen in fig. 173q. Takai (1939) interpreted the canine size variation as sexual dimorphism. The P2 ectoloph of the holotype (UMUT CV6034) has a lingually angled parastyle, but with a straight metastyle (fig. 173n). The labial rib of the paracone is more prominent than the labial metacone rib. There are no distinct lingual cusps, but rather a single crest formed by a prominent preprotocrista and lingual crest. A small swelling in the position of the protocone can be seen on another P2 (UMUT CV6035) (fig. 173p). The P3 of the holotype has a straighter parastyle and a prominent labial paracone rib, but it lacks a distinct labial metacone rib; its lingual surface is damaged. The P4 included with UMUT CV6028 (fig. 173e) is heavily worn with a damaged metacone. However, UMUT CV6038 (not shown) is a complete and unworn P4 that offers additional details. The P4 parastyle is strongly angled labially. The labial paracone rib is weak and the labial metacone rib is absent. A P4 mesostyle is lacking. There are two distinct lingual cusps, a protocone and a slightly smaller hypocone that is close to the protocone but does not appear to have been connected to it by a lingual crest. The P4 of UMUT CV6028 lacks any sign of a preprotocrista, although a very rudimentary preprotocrista can be seen on other P4s (UMUT CV6037 and CV6038). All available premolar specimens have a continuous lingual cingulum, but the labial cingula of these premolars are discontinuous, especially at the base of the paracone of the P4s.

UMUT CV6028 includes a complete upper molar row (fig. 173e). The specimen has been distorted in such a way that the M2 has been rotated, with the anterior side smashed against the posterior edge of the M1. The M1 is very heavily worn, although typical brontotheriine apomorphies are evident in the more moderately worn M2 and M3, including strongly lingually angled ectolophs, a very weak labial rib on the paracone (visible on M1 only), no labial ribs on the metacones, thin lingual ectoloph enamel, and wedge-shaped lingual margins of the paracone and metacone. A shallow central molar fossa is evident on M2. A small triangulate anterolingual cingular cusp bearing a distinct wear facet can be seen on M3. There are no traces of paraconules or metalophs. The M3

TABLE 12 Specimens referred to Protitanotherium koreanicum , along with present UMUT catalog numbers and matching figure references in Takai (1939)

lacks a hypocone, although the distolingual corner of the crown forms a small peak that bears a distinct wear facet. Lingual molar cingula are discontinuous and faint to absent. Labial molar cingula are thin, beaded, and discontinuous around the mesostyle and parastyle.

The lower dental elements include an i3 (fig. 173c, d), numerous premolars (fig. 173f– m), and a few isolated molars and molar fragments (fig. 173o). The i3 (UMUT CV29047) is small with a mesiodistally elongate crown; it closely resembles those of Protitanotherium emarginatum and Diplacodon elatus . Typically in brontotheres the lower premolars are progressively more molarized posteriorly, so that p1 is the least molarized and p4 most molarized. However, the p1 (UMUT CV6040) is unusual in its degree of molarization and out of step with this typical sequence. (These teeth may represent more than one species; see below). The trigonid consists of a short, somewhat lingually angled paralophid that creates a small lingual trigonid notch. The protolophid is angled about 45 ° posterolingually. The talonid is broad with a prominent cristid obliqua and long hypolophid, forming a broad talonid basin. The p2 (UMUT CV6041) is significantly less molarized than the p1; the trigonid is elongate and lacks a significant lingual notch, and neither the paralophid nor the protolophid are angled lingually. The talonid is the same width as the trigonid with a more poorly developed cristid obliqua and hypolophid. The lingual side of the talonid forms a slightly concave sloped surface, but a basin has not developed. The p2 lacks a metaconid. There is progressively more molarization in p3 (UMUT CV6042) and p4 (UMUT CV6041) than in p2, which represents the typical pattern. The trigonid of p3 is longer than the talonid, while the p4 trigonid is shorter than the talonid. The paralophid and protolophid of p3 are angled slightly lingually, creating a shallow lingual notch. The paralophid and protolophid of p4 curve strongly lingually, creating a molariform trigonid basin. The p3 lacks a metaconid, while p4 has a large, lingually positioned metaconid. The talonids of p3 and p4 are broader than their trigonids. The p3 talonid has a shallow basin; that of p4 is nearly molariform. Lingual premolar cingulids are absent, while labial premolar cingulids are very weak. The molars are typical: elongate with thin lingual enamel, weak lingual ribs, and shallow talonid basins.

REMARKS

Takai (1939) reported specimens of Protitanotherium koreanicum from four collections: (1) the Palaeontological laboratory of Waseda University, Tokyo, (2) the museum of the Geological Survey of the Government- General of Tyôsen ( Korea), (3) the museum of the Geological Institute, Kyoto Imperial University, and (4) from the private collection of Tosisaburô Hodosima. Takai (1939) provided photographs of 20 specimens from the Waseda University, Kyoto University, and T. Hodosima collections. Takai (1939) gave no numbers; however, all 20 are now assigned numbers in the UMUT collection (table 12). These numbers include specimens originally from the Waseda University and Kyoto University collections that are now lost, as well as material from the T. Hodosima collection that has mostly survived. Most unfortunately, the specimens in the Waseda collection, which included the holotype specimen (UMUT CV6034) and what was described by Takai (1939) as ‘‘innumerable teeth, both complete and incomplete’’ were destroyed during World War II (Hiromichi Hirano of Waseda University, personal commun., 2005). Fortunately, Takai (1939) provided quality photographs of these specimens. The surviving material in the UMUT collection includes some of the best specimens, such as the maxilla fragment with P4–M3 (UMUT CV6028). Additionally, I encountered three more unidentified specimens in the UMUT collection from the same locality that were not mentioned or figured by Takai (1939).

None of the available material is sufficient to fully diagnose a distinct species; therefore, Protitanotherium koreanicum is a nomen dubium. And while the fossil material referred to Protitanotherium koreanicum may represent multiple species, the taxonomic identity of at least some of the material can be narrowed down considerably. Takai’s (1939) referral of Protitanotherium koreanicum to the genus Protitanotherium follows Osborn’s (1923, 1925, 1929a) practice of lumping several large Asian horned brontotheres into Protitanotherium . These species were eventually split into separate genera, such as Protitan and Rhinotitan , by Granger and Gregory (1943), although they did not consider P. koreanicum in their revision. At the species level, Takai (1939) considered Protitanotherium koreanicum to be most similar to Protitan grangeri , although examination of the material leads me to disagree. The relatively advanced upper premolars rule out all brontotheres that lack premolar hypocones or only occasionally have poorly developed premolar hypocones. Among those ruled out are Protitanotherium emarginatum and all species of Protitan and Rhinotitan . The upper premolars and molars are most consistent with a number of brontotheres. However, among those species with two lingual premolar cusps, only Diplacodon elatus has incisors that are like those associated with Protitanotherium koreanicum , although a comparison of incisors cannot be made with Pachytitan ajax , another brontothere very similar to Diplacodon .

The upper dentition seems to represent a species similar to Diplacodon elatus or Pachytitan ajax . The lowers of P. koreanicum strongly contrast with Diplacodon . Most notably, p3 lacks a distinct metaconid. More perplexingly, the p1 referred to Protitanotherium koreanicum is disproportionately molarized with respect to the other premolars. Only specimens of Megacerops coloradensis from the late Eocene White River deposits of North America are known to have a similarly molarized p1. Unfortunately, the lower dentition of Pachytitan is unknown, so a comparison cannot be made with that species.

To summarize, the upper dental material assigned to Protitanotherium koreanicum is consistent with Diplacodon elatus and Pachytitan ajax . The lower dental material, excluding p1, is inconsistent with Diplacodon elatus , but a comparison with the lower dentition of Pachytitan ajax is not yet possible. Because of the possible synonymy of Protitanotherium koreanicum and Pachytitan ajax , P. koreanicum must be considered a nomen dubium. Nonetheless, the material is of great biogeographic significance; most of the specimens originally attributed to P. koreanicum seem to document a Diplacodon - or Pachytitan -like brontothere in Korea. A middle Eocene age for the Hôsan Coal Field is also supported by this conclusion. The anomalously molarized p1 (UMUT CV6040) is distinctly out of character with the other lower premolars attributed to this species, and could represent a second species, perhaps similar to Dianotitan or Megacerops .