宁夏同心中新世铲齿象化石

本文记述了宁夏同心地区中中新世哺乳动物群中铲齿象化石——同心铲齿象 (Platybelodon tongxinensis) (原定为同心嵌齿象 (Gomphotherium tongxinensis). 通过与国内外已现的有关铲齿象对比,确认其与国内已发现的葛氏种关系最近,可能属同一枝系.同心种与葛氏种相比,是性质相对原始分布层位较低的我国第二个铲齿象种.文章初步讨论了其 M_3 的某些变异性状.并结合现有的其它标本初步阐述了铲齿象类下门齿的两种结构类型,原始型的同心层状结构和衍生型的齿质柱状结构.建议根据这两种结构仍将铲齿象类划分为 Platybelodontinae 和 Amebelodontinae.

PLATYBELODON (PROBOSCIDEA, MAMMALIA) FROM THE MIDDLE MIOCENE OF TONGXIN, NINGXIA

Abundant mammal fossils were collected from near Gujiazhuang village, Tongxin County, Ningxia Hui Autonomous Region (fig. 1, showing fossil locality) in 1982, supplemented in 1983 The fossil-bearing deposits are Miocene fluviolacustne sediments, which consist of mudstone, siltstone and sandstone. The fossils were mainly preserved in sandstone metrix. The collection is very plentiful, including Artiodactyla, Perissodactyla, Lagomorpha, Carnivora and Proboscidea. The present paper deals specially with the species Platybelodon tongxinensis out of this colleetion. Although most samples are isolated teeth, yet these samnples are of speeial interest in understanding the Platybelodon group. Our specimens are very similar to Gomphotherium tongxinensis (described by Chen, 1978) collected from the same locality. It is, therefore, considered that the type specimen of Gomphotherium tongxinensis and our specimens belong to the same species. DP_4, M_1, M_2 of the new specimens tend to be tetralophodont, the lower incisors are platelike, the posterior serrated crest of the pretrite cusps of M_3 are developed. They should be best assigned to genus Platybelodon. Order Proboscidea Family Amebelodontidae Barbour, 1972 Platybelodon tongxinensis (Chen), 1978(Pls. I, II) Horizon and Locality Middle Miocene, near Gujiazhuang village, To ngxin County; Ningxia. Emended diagnosis Lower incisor flattened plate-like, similar to that of Pl. grangeri. The crown of deciduous premolars and molars higher than that in Pl. danovi but lower than in Pl. grangeri. In general the molars are narrower and longer than in Pl. danovi but relatively wider and shorter than in Pl. grangeri. M~1 and M~2 with anterior and posterior serrated crests of pretrite cusps, anterior buttress of pretrite cusps biger than posterior ones in M~3, M_3 hyposodont and angustieoronate, with 4-5 ridges, developed posterior serrated crest of pretrite Crisps and plentiful cement as well. Remarks Pl. tongxinensis is very different from Pl. (T.) loomis, Pl. kisumuens and Pl. danovi in the morphology of lower incisors. The former possesses flattened plate-like lower incisors, while incisors of the latter three species are narrower, thicker and medially higher. Furthermore, compared with Pl. danovi, the crowns of deciduous premolars and molars are higher, the contour of molars is narrower and longer; M_3 is larger and the posterior serrated crest of its pretrite cusps is more developed and cement is thicker. The juvenile incisor fragment (Pltybelodon sp.) from Loperto area, northwestern Kenya (Maglio, 1969) is similar to that of, Pl. tongxinensis, but the dentinal rods are relatively larger and more loosely packed than those of Pl. tongxinensis. Pl. tongxinensis is most similar to Pl. grangeri, however with differences from Pl. grangeri: (1) M~1, M~2 with posterior Serrated crest of pretrite cusps. (2) M~3 without anterior serrated crest of pretrite cusps. (3) The molar crown is lower and not so narrow and long as that of Pl. grangeri. (4) The tendency to tetralophodonty of the intermediate molars is not so distinct as in Pl. grangeri. (5) The fewer ridges (4-5), the relatively wider valleys and the less distinct alternation between the pretrite cusps and posttrite cusps than in Pl. granger (6) The less cement than in Pl. grangeri. The above mentioned differences indicate that Pl. grangeri is a species with more specialized characters and Pl. tongxinensis represents a relatively more primitive species. The associated fauna shows that Pl. grangeri is younger than Pl. tongxinensis in geological age. Selenolophodon spectabilis (Hopwood) from Lintong, Shaanxi (Chang & Zhai, 1978), characterized by hypsodonty, angusticoronate, plentiful cement, 4 ridges in M_2, ridges in M_3, and the developed posterior serrated crest of pretrite cusps in M_3. These characterS clearly show that Lintong specimen belongs to Platybelodon, and is quite similar to Pl. tongxinensis. Because of the absence of upper molars and lower incisors, however, its Specific position is uncertain. Perhaps it may represent a form of Pl. tongxinensis with progressive chaxacters: higher tooth crown and thicker cement than in Pl. tongxinensis. In studying these material two interesting points were noticed. 1. The samples kept in I. V. P. P. confirmed that there are two kinds of internal structures of "shovel trsk", namely, dentinal rods and dentinal laminae. In dentinal rods structure incisor, the rods are packed in the core of "shovel tusk" and surrounded by dentinal laminae (see fig. 2-a). Although they have not been weathered out as discrete and isolated rods, the diffferent kinds of fracture in external laminae and internal rods are revealed. In the incisor with dentinal laminae structure the overall plate of incisor is made up of laminae (see fig. 2-b). V2407, a fragment of shovel tusk from Qinan, Gansu, is of dentinal laminae structure. About 14 laminae can be distinguished with naked eye. The lo- wer incisors of Phiomia from late Eocene of Fayum are also of dentinal laminae structure, so this kind of structure must be of a primitive character and the dentinal rods structure may be of derivative character. Generally speaking, the internal strtucre of animal organs are usually more stable than their morpholgy, therefore the structure of shovel tusk should be taken into account in the classification of Amebelodontidae. After checking up on the plates and illustration so far published, we found that there seems to be two kinds of molars relative to two kinds of structures of "shovel tusk" respectively. Because the type specimen of Amebelodon (Barbour, 1927) is of dentinal laminae structure, this kind of structure therefore should be taken as a generic character of Aniebelodon. Am. frieki from Kansas (Barbour, 1941) possesses lower incisors of dentinal rods structure, in addition, molars are of extreme hyposodonty and narrow contours, M~2 4 ridges and M~3 6 ridges, it belongs therefore to platybelodonts instead of Amebelodon. According to the fact that the lower incisors of Pl. (T) barnumbrowni and Pl. (T.) cf. barnumbrowni are of dentinal laminae structure (Shotwell, 1963) it is possible that both forms should belong to amebelodonts, although they possess the broad, paatelike incisors which have usually been considered as a generic characteristies of Platybelotion. It is suggested that Am. fricki and "Pl. (T.) barnumbrowni" represent two different adaptive patterns of amebelodonts, the former with relative narrow lower incisors and the latter with broad, plate-like incisors. 2. There are considerable individual variation in M_3 Of Pl. tongxinensis. The third lower molars can be divided into 4 ridges type and 5 ridges type. The conulete and talonid are variable in every tooth. Even the size and form of the last ridge, and talomids on the opposite teeth of the same individual are different (see V 8029, V 8027). The size ranges from 175×63.5 mm to 252×84.2 ram. It is therefore apt for these M_3 to be taken as different species. But table 5 straws no obvious gaps in size. The phenomena of transitional change from 4 to 5 ridges is possibly due to the variation of talonid. All M_8 share the following characters: hypsodont and angusticoronate crown, the accessory structures restricted to the pretrite side, developed posterior serrated crest of pretrite cusps, plentiful cement. The associated fossils and their distribution in stratum show that they lived in the same short period of geological time. It would surely be unreasonable to separate these M_3 of Platybelodon into different species. We believe that they belong to a single species of Platybelodon: Besides the molars, considerable individual variations, espeeially age variations also exhibit in the morphology of lower incisor symphysis and horizontal branch of mandible. The details will be discussed in another paper.