Metatelmatherium, Granger and Gregory 1938

Mader, Bryn J., 2008, A species level revision of Bridgerian and Uintan brontotheres (Mammalia, Perissodactyla) exclusive of Palaeosyops, Zootaxa 1837 (1), pp. 1-85 : 74-79

Genus METATELMATHERIUM Granger and Gregory 1938

Age. Uintan.

Subage. Late Uintan (one specimen assigned to this genus, CM 11380 , is from earliest Uintan deposits in the Sand Wash Basin) .

Type species. M. cristatum Granger and Gregory 1938 .

Included species. M. ultimum ( Osborn 1908) .

Diagnosis. Large-sized (length P2 to M3 is approximately 200–211 mm) brontothere in which the zygomatic arch typically exhibits a prominent flange on the ventral surface of the jugal near to where it borders on the squamosal; and in which the lateral nasal incision is shifted forward so as to lie over the upper diastema or P1.

Discussion. Metatelmatherium is one of the few brontotheres known to have existed in both North America and central Asia. Specimens of the North American species ( Fig. 26 ) were the first to be discovered and were originally described ( Osborn 1908) as a new species of Telmatherium , T. ultimum . Granger and Gregory (1938), however, correctly concluded that this taxon was not referable to the genus Telmatherium , but instead was congeneric with a skull and jaw (AMNH 26411) that they were studying from the Irdin Manha Formation of central Asia to which they assigned the new generic name Metatelmatherium .

The type species of Metatelmatherium , M. cristatum , is from Asia and, according to Granger and Gregory (1938; 1943; see also Mader 1989), is distinguished from the North American species, M. ultimum , by its slightly larger size. It should be noted, however, that although the type of M. cristatum is larger than any North American specimen of Metatelmatherium that I have measured, for most variables the size difference between M. cristatum and some North American specimens is minimal. The sample of North American Metatelmatherium is small (n = 6) and it is quite possible that North American specimens will eventually be recovered that more closely approach or even exceed the size of the type of M. cristatum . There is at least reason to question whether the population from which the type of M. cristatum was drawn was significantly larger (in a statistical sense) than the population from which the North American specimens were drawn. I do not believe, therefore, that there is sufficient evidence to separate the North American and central Asiatic forms at the species level on the basis of size.

In addition to the presumed greater size of Metatelmatherium cristatum over M. ultimum, Granger and Gregory (1943) listed the following as "specific and individual" characters of M. cristatum : P1 very large and roundly compressed; P2 large and asymmetric; P3 and P4 moderately wide with but a slight cingulum; anterior border of orbit to front of canine 190 mm (160 mm in M. ultimum ); occiput relatively broad; lower jaw differing from M. ultimum in its short, low, broad coronoid process, relatively deeper ramus, and longer more sloping symphysis; and P3 relatively wider than in Epimanteoceras (an Asian taxon not discussed in the present paper) but relatively less wide than in Desmatotitan (another Asian taxon). Because Granger and Gregory list these as specific and "individual" characteristics, they evidently did not regard all of these characters as diagnostic of the species but considered some to be merely descriptive of the type specimen. If this is the case, Granger and Gregory did not specify which of these characters they regarded as diagnostic.

After my own examination of the type of Metatelmatherium cristatum I find that the following characters (cited in part by Granger and Gregory) are potentially diagnostic of the species: compared to specimens of M. ultimum the posterior cusp on P1 is higher and wider, the P2 is proportionately larger, the cingula on P3 and P4 are thinner, the occiput is broader, the coronoid process is broader, and the mental symphysis is longer and more sloping. Because these morphological difference are based on a comparison of only a few specimens (AMNH 26411, the type of M. cristatum ; AMNH 2060, the type of M. ultimum ; and AMNH 2004, a referred specimen of M. ultimum ), however, I am skeptical that they will all prove to be of diagnostic value once the intraspecific variation of each species is better understood. Of all the characters listed above, the relatively long and more sloping mandibular symphysis of M. cristatum seems most likely to continue to be a valid diagnostic character when more specimens become available. Most recently, Mihlbachler (2005) has synonymized M. cristatum with M. ultimum , but for the present, given their wide geographic separation, I prefer to continue regarding them as distinct. If Mihlbachler is correct, however, then M. cristatum Granger and Gregory, 1938 , (the type species of Metatelmatherium ) will become invalid and a junior synonym of M. ultimum ( Osborn 1908) .

In addition to Metatelmatherium cristatum , two other Asian species of Metatelmatherium have been named: M. browni from the Pondaung Formation of Burma ( Colbert 1938), and M. parvum from the Irdin Manha Formation of central Asia ( Granger & Gregory 1943). A third Asian taxon, Eotitanotherium (?) lahrii from the Pondaung Formation of Burma, was referred to Metatelmatherium by Colbert (1938). In my opinion, M. browni , M. parvum , and M. lahrii are all probably nomina dubia (and are treated here as such), but, since these are Asian taxa, a detailed discussion of the reasons for not accepting their validity is beyond the scope of the present paper.

In North America only a few skulls of Metatelmatherium are known, most of Late Uintan age. There are four Late Uintan specimens that are clearly referable to Metatelmatherium (AMNH 2060, the type of M. ultimum ; AMNH 2004; CM 2339; and CM 2388, the type of " Manteoceras " uintensis ), all from the Myton Member of the Uinta Formation. Of these four specimens, only three are measurable (CM 2339 lacks dentition), making the sample too small to analyze statistically. A badly crushed skull (AMNH 2029) from the Myton Member of the Uinta Formation is also probably referable to Metatelmatherium , but does not significantly increase the sample size. All of the Late Uintan specimens are highly similar in morphology and there is no basis for the recognition of more than a single Late Uintan species of Metatelmatherium . A skull (CM 11380) from the Sand Wash Basin of Colorado, however, may represent a distinct Early Uintan species and Mihlbachler (2005) has gone even further and assigned this specimen to a new genus.

The Sand Wash specimen is of earliest Uintan age and lacks the characteristic flange on the zygomatic arch. In all other respects, however, it closely resembles specimens of Metatelmatherium ultimum from the Uinta Basin. Given the earlier temporal occurrence and difference in zygomatic arch morphology, the Sand Wash skull might represent a species different from M. ultimum . Although I have previously stated ( Mader 1989) that I am inclined to regard the difference in zygomatic arch morphology as intraspecific variation and probably sexual dimorphism, I am now less certain of this conclusion. I am not willing, however, to recognize a new taxon until more Late Uintan specimens of Metatelmatherium ultimum are available so that the variation of M. ultimum can be better documented. Even if the specimen should be regarded as a new species, however, I find no justification for Mihlbachler’s (2005) suggestion that it be placed in a new genus.

Osborn (1908; 1929) illustrated (see Figs. 26D and 27A ) a lateral view of the type skull of Metatelmatherium ultimum showing an arrangement of the cranial sutures in which the nasal is excluded from contact with the lacrimal by the frontal and maxilla, which are in broad contact. The same morphology occurs in metamynodontine rhinoceroses, and Wall (1989) has used this suture pattern as one of the synapomorphies defining that tribe. If Osborn was correct in his interpretation of the cranial sutures, this pattern would be an excellent derived character for M. ultimum and perhaps for the genus Metatelmatherium in general. Gregory (1920) demonstrated that the nasal-lacrimal contact is plesiomorphic for perissodactyls, and it appears that this is the typical arrangement in brontotheres. I have confirmed that the nasal is in contact with the lacrimal in Palaeosyops , Mesatirhinus , Metarhinus , Sphenocoelus (= Dolichorhinus ), and Telmatherium .

It is very difficult to discern the pattern of cranial sutures in the type skull of Metatelmatherium ultimum , however, and I have not been able to confirm Osborn's observation on this or any other specimen of M. ultimum that I have examined. I suspect that the arrangement of the nasal and lacrimal may be the same in M. ultimum as it is in other brontotheres (see Fig. 27B and 28 ). At least the exclusion of the contact between the nasal and lacrimal should not be used as a diagnostic character of M. ultimum or the genus Metatelmatherium until this arrangement can be more clearly demonstrated.

Although I have not been able to confirm that the nasals are excluded from contact with the lacrimal, I have been able to confirm that a triangular projection of the frontal overlaps the nasal as illustrated by Osborn. This morphology is rather similar to that of the Brontotheriinae and may indicate that Metatelmatherium is either a member of this subfamily or is closely allied to it (see Fig. 28 ).

Interestingly, it is possible to artificially approximate the skull of Metatelmatherium by electronically warping an image of a Telmatherium skull. Figure 29 shows a skull of Telmatherium that has been modified using the transform (warp and distort) function of Adobe Photoshop CS3 Extended. By warping the entire image and selected parts of it separately, it is possible to produce an illustration that is very similar in appearance to Metatelmatherium . Based on this photographic manipulation it seems entirely plausible that Metatelmatherium may have been directly derived from Telmatherium or something very much like it (although this is certainly not proven).

I have previously expressed the opinion, however, that Metatelmatherium may be closely allied to Sthenodectes and have suggested that the two might be placed in a separate brontothere subfamily the Metatelmatheriinae ( Mader 1989). This relationship is suggested by the relatively large size of the incisors in both genera and by the possibility that the lateral nasal incision may be shifted forward in Sthenodectes as it is in Metatelmatherium . Given the present uncertainties, I have chosen not to assign Metatelmatherium and Sthenodectes to a particular subfamily and have instead treated them as Brontotheriidae incertae sedis.

Figures

FIGURE 26. Holotype skull of Metatelmatherium ultimum (AMNH 2060) in A, anterior; B, ventral; C, dorsal; and D, lateral views. A, B, and C after Osborn, 1929; D after Osborn, 1908 and 1929.

FIGURE 27. Reconstruction of the anterior cranial suture pattern in the holotype of Metatelmatherium ultimum (AMNH 2060). A, pattern according to Osborn, 1908 and 1929; B, pattern suggested in the present paper based on the morphology of other brontothere genera. Light gray shading, nasals; dark gray shading, frontals. Both illustrations after Osborn, 1908 and 1929.

FIGURE 28. Comparison between the anterior cranial suture pattern of Metatelmatherium and the brontotheriine brontothere Telmatherium. A, partly hypothetical (see text) reconstruction of the cranial suture pattern in Metatelmatherium ultimum (AMNH 2060, type); B, anterior cranial suture pattern of Telmatherium validus (AMNH 12678). Light gray shading, nasals; dark gray shading, frontals. A after Osborn, 1908 and 1929; B after Osborn, 1929.

FIGURE 29. Electronically altered image of Telmatherium skull approximating the morphology of Metatelmatherium. A, skull of Telmatherium (AMNH 12678); B, the same image modified using the transform (warp and distort) function of Adobe Photoshop. A after Osborn, 1929.