<<

The World of - International Congress, Rome 2001

Proboscidean from the Japanese Archipelago and Islands and their relationship with the Chinese

K. Takahashi1, C.H. Chang2, Y.N. Cheng2 1 Lake Biwa Museum, Shiga, - [email protected] 2 National Museum of Natural Science, Taichung, Taiwan [email protected], [email protected]

SUMMARY: The colonization, evolutionary processes and the relationship between the faunas of the Japanese Islands and Taiwan Island have been summarised, based on a comparison of their pro- boscidean taxa with those of the Chinese mainland. The comparison indicates that Mammuthus protomam- monteus and aurorae evolved in the Japanese Islands and Taiwan respectively, and that S. shin- shuensis, S. orientalis and Palaeoloxodon naumanni migrated to Japan from the mainland of China. Although P. huaihoensis migrated to Taiwan, P. naumanni did not migrate there.

1. INTRODUCTION There are two main factors influencing the faunas in these islands. Firstly, the depth of The Japanese Archipelago and Taiwan are the strait between the islands and the Asian located along the east margin of the Asian mainland. The maximum depth of the Strait of , with Japan in the north and Taiwan Korea between the continent and the Japanese in the south, ranging from 45° ~ till about 24° islands is about 130 m. Most of the depth of the morthern latitude. Between the Japanese island Strait of Taiwan between the continent and of Kyushu in the north and Taiwan are situated Taiwan is less than 100 m depth. These differ- the , a series of small islands ences suggest that Taiwan has been connected belonging to Japan. longer or more frequently to the continent than For convenience , , the Japanese Islands, and such a difference has and Kyushu, the main Japanese greatly affected the faunal composition of these islands, are herein referred to as the ‘Japanese two island groups. Islands’. A second factor is latitude. The Japanese The faunas of all these islands were strongly Islands are located further north than Taiwan. influenced by local tectonic movements and Nowadays, the Japanese Islands are situated in climate changes, resulting in changes in the sea the temeprate zone, while Taiwan is situated in level in the Plio-. During periods of the subtropical zone. These difference in lati- low sea level, the Strait of Korea immediately tude must have resulted also in a difference in northwest of the southernmost Japanese climate during the Plio-Pleistocene, which islands, the Strait of Taiwan and the East resulted in different faunas. It would have China Sea, opposite the Ryukyu islands, were turned up as the movement of the boundary line exposed as dry land, and these areas then between the Palaearctic and the Oriental that is played a central role in the migration of presently located on the Ryukyu Islands. . When sea levels rose, islands were The colonisation, evolutionary processes and again formed, and endemic sometimes the relationship between the faunas of these evolved on those islands. islands are summarized, and a comparison is

148 made of the fossil proboscidean taxa of these boundary of NN19/NN20 is approximately 0,9 islands with those of the Chinese mainland to 0,4 Ma. (Takahashi and Namatsu 2000). We began our study in 1999 with a reexami- nation of 232 proboscidean molar fossils from 2. JAPANESE PROBOSCIDEAN FOSSILS the sea bottom around the Penghu Island and Tsiliao area near Tainan, including much new In recent years Japanese Plio-Pleistocene material. Our comprehensive study confirmed strata have been well studied by teprochronolo- the presence of two species of Stegodon,two gy, palaeomagnetic stratigraphy and other dat- species of Mammuthus from the Early to ing methods, and recent studies of widespread Middle Pleistocene, and one Palaeoloxodon volcanic ash deposits have allowed for the cor- from the . There are no fossils relation of event horizons at geographically before the Pleistocene. distant areas (Oda 1977; Machida et al. 1980; Yoshikawa et al. 1996; Satoguchi et al. 1999, 4. COMPARISON OF JAPANESE AND TAIWANESE etc.). Therefore, the geological horizon and the PROBOSCIDEAN FOSSILS age of proboscidean fossils are known accu- rately. A comparison of Japanese and Taiwanese The earliest proboscidean fossils from the proboscideans by Shikama et al. (1975) shows Japanese Islands are found in Miocene deposits two mammalian faunas in Taiwan. One was of Gifu prefecture, Honshu (Matsumoto 1926; named the Cho-chen fauna, being represented Kamei et al. 1977). After the the num- by Mammuthus armeniacus taiwanicus, S. ber of specimens increases and proboscidean aurorae and S. sinensis. fossil taxa appear in the following chrono- The other was the Penghu-Tainan fauna logical sequence: Shinomastodon sendaicus, represented by Palaeoloxodon naumanni. Stegodon shinshuensis, S. aurorae, Shikama et al. (1975) placed the first fauna in Mammuthus protomammonteus, S. orientalis, the Middle Villafranchian and the last in the Palaeoloxodon naumanni and M. primigenius. Late Pleistocene. These fossil sites number are over 350 (Kamei In the Japanese Islands, Stegodon and 1991). Mammuthus occur in the Early to Middle However, only two proboscidean molars Pleistocene, the same age as on Taiwan. S. auro- have been found from the Ryukyu Islands rae is known from the Late Pliocene to the Early (Tokunaga 1940; Nohara and Hasegawa 1973). Pleistocene (2.5 Ma – 1.0 Ma), M. protomam- Otsuka (1997) identified one of these speci- monteus from the Early to the Middle mens as be similar to Mammuthus paramam- Pleistocene (1.0 Ma – 0.7 Ma), and S. orientalis monteus shigensis, and estimated the age as from the Middle Pleistocene (0.5 Ma – 0.4 Ma). the Late Pleistocene. There are too few pro- Using data from tephrostratigraphy and oxygen boscidean fossils recovered from the Ryukyu isotope stratigraphy, Konishi and Yoshikawa Islands to include this area in the present dis- (1999) estimated the immigration date of M. cussion. protomammonteus into the Japanese Islands as 1.15 Ma, and that of S. orientalis as 0.62 Ma. 3. TAIWANESE PROBOSCIDEAN FOSSILS S. sinensis, S. (Parastegodon) akashiensis, S. (Parastegodon) aurorae, hysudricus The horizon and the age of proboscidean fos- and M. armeniacus taiwanicus were recorded sils of Taiwan are not yet known accurately. among the Cho-chen fauna described by Recently, the layers of the area in which Shikama et al. (1975). S. (Parastegodon) Stegodon and Mammuthus occurred were stud- akashiensis has subsequently been synonymised ied. The nannofossil biostratigraphy (Shieh and with S. aurorae (Taruno 1991), and our studies Shieh, personal communication) suggests that indicate that the specimens identified as Elephas the area involved must be placed in NN19. The hysudricus by Shikama et al. (1975) are identi- 149 The World of Elephants - International Congress, Rome 2001

cal to M. armeniacus taiwanicus. Furthermore, parison with the Japanese sequence, we posit a S. sinensis and S. aurorae of Shikama et al. sequence of S. aurorae, M. a. taiwanicus and S. (1975) are morphologically indistinguishable orientalis, with the latter the most recent taxon. from S. orientalis and S. aurorae of the Japanese Shikama et al. (1975) described Pa- islands. Although M. armeniacus taiwanicus laeoloxodon from Taiwan as Palaeoloxodon described by Shikama et al. (1975), is slightly naumanni. However, our detailed investiga- different from M. protomammonteus of the tions indicate that Palaeoloxodon of Taiwan Japanese Islands in the mumber of lamella and differs markedly from P. naumanni of the molar size, those general characters are very Japanese Islands in the size and wearing pattern similar. Therefore, the proboscidean taxa from on the occulusal surface of the molars. the Cho-chen fauna can be placed into S. orien- We think that three species of Palaeoloxodon talis, S. aurorae and M. a. taiwanicus. (P. namadicus, P. naumanni and P. huaihoen- The fossils from the Cho-chen fauna of sis) are represented in China, and that Taiwan are considered as resedimentated fos- Palaeoloxodon of Taiwan is the same as P. sils, and as such the chronological sequence of huaihoensis, as Qi (1999) describet. P. nau- the proboscidean fossils cannot be decided manni has been recorded from Japan, but P. accurately. However, based upon direct com- huaihoensis has not.

Fig.1 - Distribution maps of proboscidean fossils in eastern Asia during the Plio-Pleistocene. 150 Proboscidean fossils from the Japanese Archipelago and Taiwan Islands...

The migration of P. naumanni into the Nohara, T. & Hasegwa, Y. 1973. An elephan- Japanese Islands has been estimated as at 0.43 toid tooth from Kyan, Okinawa-jima Ma by Konishi and Yoshikawa (1999). After (Studies of the paleovertebrate fauna of this age, there exists no evidence that the Ryukyu Islands, Japan. Part III). Mem. Nat. Japanese Islands were connected with the Sci. Mus. 6: 59-63. (in Japanese) Asian Continent. Because Palaeoloxodon of Oda, M. 1977. Plantonic foraminifera biostratig- Taiwan is considered to be of the Late raphy of the Late Cenozoic sedimentary Pleistocene (Hu and Tao 1993), it is under- sequence, central Honshu, Japan. Tohoku standable that P. huaihoensis has not been Univ. Sci Rep., 2nd ser. (Geol.), 48, (1): 1-72. found in Japan. Otsuka, H. 1997. On the molar from Comparison of the research results in the Miyako island of Nansei Islands as reported Japanese Islands and Taiwan with data from by Tokunaga (1940). Abstract paper of the mainland China indicates that M. protomam- 1997 Annual Meeting Paleont. Soc. Japan monteus and S. aurorae evolved in the Japanese 66. (in Japanese) Islands and Taiwan, and that S. orientalis, P. Qi, G. 1999. On some problemes of naumanni and P. huaihoensis migrated there Palaeoloxodon in China. In Wang, Y. And from the mainland. Deng, T. (eds.), Proc. VII Annual. Meeting Chinese Soc. Vert. Paleont.: 201-210. China 5. REFERENCES Ocean Press. Satoguchi, Y., Nagahashi, Y., Kurokawa, K. & Hu, C.H. & Tao, H.J. 1993. The fossil fauna of Yosikawa, S. 1999. Tephrostratigraphy of Penghu Islands, Taiwan. Penghu District the Pliocene to Lower Pleistocene forma- Cultural Center Publications: Taiwan. (In tions in central Honshu, Japan. Earth Chinese) Science, 53: 275-290. (In Japanese with Kamei, T. (ed.). 1991. Japanese proboscidean English abstract) fossils. Tsukiji Shokan: . (In Shikama, T., Otsuka, H. & Tomida, Y. 1975. Japanese) Fossil from Taiwan. Sci. Rep. Kamei, T., Okazaki, Y., Nonogaki, I. & Yokohama National University, 2nd. sec. 22: Paleontology Club, Aichi Gakuin 13-62. University. 1977. On some new materials of Takahashi, K. & Namatsu, K. 2000. Origin of Gomphoterium annectens (Matsumoto) the Japanese proboscidea in the Plio- from the Mizunami Group, central Japan. Pleistocene. Earth Science, 54: 257-267. Bull. Mizunami Fossil Museum, 4:1-8. Taruno, H. 1991. Stegodon fossils from Japan. Konishi, H. & Yoshikawa, S. 1999. In Kamei, T. (ed.), Japanese Proboscidean Immigration times of the two probocidean Fossils: 82-99. Tokyo: Tsukiji-Shokan. (In species, Stegodon orientalis and Japanese) Palaeoloxodon naumanni, into the Japanese Tokunaga, S. 1940. A fossil elephant tooth dis- Islands and the formation of the land covered in Miyakozima, an island of the bridge. Earth Science 53: 125-134. (In Ryukyu Archipelago, Japan. Proc. Imp. Japanese with English abstract) Acad. Japan, 16: 122-124. Machida, H., Arai, F. & Sugihara, S. 1980. Yoshikawa, S., Satoguchi, Y. & Nagahashi, Y. Teprochronological study on the middle 1996. A widespread volcanic ash bed in the Pleistocene deposits in the Kanto and Kinki horizon close to the Pliocene-Pleistocene districts, Japan. Quat. Res. 19: 233-261. (In boundary: Fukuda-Tsujimatagawa-Kd38 Japanese with English abstract) volcanic ash bed occurring in central Japan. Matsumoto, H. 1926. On two new mastodonts J. Geol. Soc. Japan, 102, 258-270. (In and an archetypal stegodont of Japan. Sci. Japanese with English abstract). Rep. Tohoku Imp. Univ., 2nd ser.(Geol.) 10: (1): 1-11. 151