Samotherium major

Denis Geraards, 2017, Late Miocene large mammals from Mahmutgazi, Denizli province, Western Turkey, Neues Jahrbuch für Geologie und Paläontologie-Abhandlungen 284 (3), pp. 241-257 : 242-244

Samotherium major BOHLIN, 1926

Description and comparisons: This taxon is represented by a number of postcranials and several dental remains. The size range of the DP3s is perhaps greater than expected in a single species but all other remains, including the postcranials, are of homogeneous dimensions, and it is more parsimonious to assume that a single species is present.

Ma2-Gips29 ( Fig. 1B ) is the only adult maxilla. The teeth are imperfectly preserved, but only slightly worn, preserving prominent styles and pillars. The premolars are rather large but P4 remains much narrower than M1; they bear strong spurs in the central valley, as may happen in Samotherium ( BOHLIN 1926, pl. 5, fig. 10).

The best specimen is a partial mandible from Ma3 with the left i2 -c and p2-m3; unfortunately part of the diastema is missing, so that its length is unknown ( Fig. 1A ). As in other Samotherium , the two lobes of the canine are much less clearly distinct than in Giraffa , and this tooth is not much larger than the incisors. The front teeth show no evidence of the interstitial wear that is observed in some other late Miocene large giraffids of the Aegean region, such as Helladotherium from Hadjidimovo, Bulgaria ( GERAADS et al. 2005), Samotherium major from Vathylakkos, Greece ( GERAADS 1974; MERCERON et al. in press), or S. neumayri from Maragha, Iran ( RODLER & WEITHOFER 1890, p1. 4, figs. 2-3). This peculiar wear, which is certainly linked with their use as a comb when browsing, is also variably present in Giraffa camelopardalis . The premolars are strongly molarized: p2 has a long entocristid; p3 is similar to p 4 in having a long metaconid connected to the paraconid, but not to the transverse cristid, and blocking the entrance to the mesosinusid; the ectoflexid of p4 is deep, as in other Giraffidae , but the heavy wear has connected the hypoconid to the rest of the tooth. These premolars are also quite short, with a Pm/M index of 58.6 ( Table 1). These short, molarized premolars suffice to identify the Mahmutgazi large giraffid as Samotherium . In the large Helladotherium , a genus common in Aegean sites of similar age, the premolars are significantly larger, and p3 always retains a primitive morphology. Most representatives of the species S. boissieri from Samos also retain this primitive morphology ( BOHLIN 1926, figs. 141-146; KOSTOPOULOS 2009a, fig. 8), but S. neumayri from Maragha and S. major from Vathylakkos and Samos ( BOHLIN 1926; GERAADS 1974; KOSTOPOULOS 2009a) have a fully molarized p3. The shortening of the premolars also fits better S. major (Pm/M index = 53.8-66.2, mean = 59.5, N = 6) than S. boissieri (Pm/M index = 60.7-65.6, mean = 63, N = 5) (meas- urements from GERAADS 1974; KOSTOPOULOS 2009a, table 8). However, the absolute size of the teeth ( Table 1) is lower than in S. major , and closer to S. boissieri .

Upper DP3s ( Fig. 1K ) have a relatively short first lobe; lower dp3s ( Fig. 1L ) also have large metaconids completely closing the mesosinusid, and the first lobe of dp4 is fully formed, with no labial arm; these are all derived features, usually found in Samotherium .

The postcranial sample is large. There are no less than nine distal humeri , five of them from the right side, providing a Minimum Number of Individuals of five for the whole collection. It is likely that all these postcranials are indeed from a limited number of individuals, but they cannot be matched, because a single plaster jacket may have contained bones of more than one individuals ( Ma2-Gips20 has two right metacarpals).

The dimensions of the distal humeri ( Table 2) are homogeneous, but there is no complete bone ( Fig. 1E ). The available sample of measurements from nearby sites is small; these humeri are clearly larger than those of S. boissieri from Samos (dist. TD = 98-106 according to KOSTOPOULOS [2009a]), but smaller than those of its likely descendant S. major from the same area (dist.TD = 132-148). The Mahmutgazi humeri also compare well with the forms from the two levels of Kemiklitepe (KTD, lower, and KTA, upper: GERAADS 1994), and with S. neumayri from Maragha.

There are six more or less complete metapodials ( Table 3); the measurements of the metacarpals ( Fig. 1 F-H) clearly plot with S. major ( Fig. 2 ).

There are three tibiae ( Fig. 1C, D ), whose measurements ( Table 4) are clearly more similar to those of S. major than to those of S. boissieri .

There are only four astragali , two of them well preserved ( Fig. 1J ). One is similar in size to those from KTD, and the other to those from KTA ( GERAADS 1994), but they also fit S. neumayri from Maragha (pers. data). It is remarkable that, while the size distinction between the two levels of the nearby site of Kemiklitepe is clear, the size of the Mahmutgazi material overlaps both of them ( Fig. 3 ), but the difference in size between the two well-preserved specimens does not exceed what is found in a single population.

A number of other bones ( carpals, tarsals, calcanei, phalanges ) obviously belong to the same species; the first phalanges are short and stout, as in Samotherium ( GERAADS 1974); the cubo-navicular has no posterior articular facet for the metatarsal ( Fig. 1I ), as in other Samotherium , but in contrast to Helladotherium , in which it is present ( GERAADS 1974; KOSTOPOULOS 2009a).

Discussion: All postcranials can easily be assigned to Samotherium major , and the derived lower premolars fit this identification, but the lower teeth from Ma3 are smaller than in this species. If indeed Ma3 is contemporaneous with Ma1 and Ma2, these teeth may be from a small individual of S. major . In terms of biochronology, S. major clearly allies Ma1 and Ma2 with the upper part of the Samos sequence ( KOSTOPOULOS 2009a, fig. 21).

Figures

Fig. 1. Artiodactyla from Mahmutgazi. A-L – Samotherium major. A: left mandible Ma3 (A1: occlusal view; A2: lateral view of the rostral part). B, maxilla with P3-M2 Ma2-Gips29. C-D: tibiae in anterior view (C: MA2-Gips16; D: Ma2- Gips11). E: distal humerus n°54, anterior view. F-H: metacarpals in anterior view (F: Ma2-Gips17; G, H: Ma2-Gips20). I: cubonavicular Ma1-Gips10, distal view, to show the absence of plantar metatarsal facet. J: astragalus Ma1-Gips10, anterior view. K: upper deciduous tooth-row Ma1-Gips3. L: lower dp3-dp4 Ma1-Gips11. M – Tragoportax sp., right upper tooth row. N: Miotragocerus (Pikermicerus) gaudryi, left lower tooth-row. Scale bar = 5 cm for Figs. M, N, 10 cm for Figs. A, B, and I-L, 25 cm for Figs. C-H.

Fig. 2. Plot of length vs. distal width of late Miocene giraffid metacarpals (KTA = Kemiklitepe, upper level; Slq = lower Axios valley, ARAMBOURGʼS collection). Data from KOSTOPOULOS (2009a), and original measurements.

Fig. 3. Plot of medial length vs. distal width of late Miocene giraffid astragali. Note that ‘Maragha’ includes both S. neumayri and Palaeotragus coelophrys. Data from KOSTOPOULOS (2009a), and original measurements.