The Ryukyu Islands, about 1, 200km. in length, are arc islands between Kyushu and Taiwan (Fig. 1). The islands have been considered unique in their Quaternary landforms and geology compared with the main islands of Japan, because of the distribution of living coral reefs and the Quaternary Ryukyu Limestone. In addition to the geomorphological and geological researches, anthropology, paleontology and archeology of the islands have been intensively studied. Especially the ages, routes and methods of arrival of the paleolithic men and mammals to the islands in the Pleistocene period have been discussed. In the past several years, discoveries of some artifacts indicating the older limit of the neolithic age were presented. The theme, “Problems of the late Pleistocene-Holocene periods in the Ryukyu Islands, ” is selected for this symposium so that it will provide a common ground for discussion from the various fields. The following topics may be discussed in the symposium. (1) Sedimentary environments and ages of the Ryukyu Limestone. (2) Stratigraphy of the Ryukyu Limestone and the Kunigami Gravel bed. (3) Modes of the crustal movement. (4) Sea level changes (ex. sea level of the last glacial maximum). (5) Routes and methods of arrival of the paleolithic men and mammals to the islands. (6) Relation between the neolithic cultures in the Ryukyu Islands and the Jomon-Yayoi cultures in the main islands of Japan. (7) Relation between the neolithic cultures of the Sakishima Islands and those of the Okinawa Islands and northward.
The Ryukyu island arc is divided into three regions by Tokara channel and Miyako depression (Figure 6), between which a region is called the Central Ryukyus. The region south of the Miyako depression is named the South Ryukyus. Throughout the two regions the Pleistocene coral reef sediments are distributed and united into the Ryukyu Group (MacNeil, 1960), in which the Upper and Lower Formations are classified (Table 7). The Upper Formation is thought to have been deposited in the last interglacial stage and the following sub-interstadials, according to the dating by 230Th and 231Pa growth method (Konishi et al., 1974) and the nannofossil biostratigraphy. Emiliania huxleyi, a nannofossil species, commonly occurs in the lowermost part of the Upper Formation in Ishigaki-jima (Figure 2) and such occurrence has been considered to show the δ18O stages from the 6th on by GARTNER (1972) and others. The age of the Lower Formation may be the Middle Pleistocene, probably δ18O stages 15 to 13 (Table 7), because its lowermost part contains rare Pseudoemiliania lacunosa and abundant Gephyrocapsa oceanica (NISHIDA, 1977; MINOURA, 1979). The Upper Formation forms the Upper Middle, Lower Middle, Low and Lowest terraces, while the Lower Formation sometimes makes the High terrace. Depending upon the characters of the basement beneath the Ryukyu Group, the Ryukyu Islands are also classified into four areas. As the backbone of the Ryukyu island arc was made up in some period of the Late Miocene (KONISHI and SUDO, 1972), the distribution of the Ryukyu Group and of the accompanied marine terraces must be related to the tectonic movement after the Late Miocene. The above-mentioned four areas are the Late Miocene to Early Pleistocene Shimajiri Group area, pre-Late Miocene areas of the Central and South Ryukyus, and Late Miocene to Holocene volcanic inner arc area (Table 7). The Lower Formation of the Ryukyu Group is not distributed in the pre-Late Miocene area of the South Ryukyus. In the Shimajiri Group area, however, the two formations of the Ryukyu Group are distributed although the High terrace is not recognizable. The Lower Middle terrace keeps its reef barrier topography in many islands. The Low terrace consists of “Awaishi limestone”, foraminiferous calcarenite with uniform texture, sometimes. The distribution of Pleistocene marine terraces is outlined in Figure 3. While the Late Pleisotocene terraces are ubiquitously distributed, that of the Middle Pleistocene ones is limited into the Central Ryukyus. This fact may mean that the period both the Central and South Ryukyus commenced to be upheaved is the latest Middle Pleistocene. Although KONISHI and SUDO (1972) pointed out the tilting toward the Asiatic continent of the whole Ryukyu arc, it is not recognized concerning the South Ryukyus. From the facts that the Central Ryukyus tilts toward the continent, that Hateruma-jima in the South Ryukyus does westwards, and that the height of the last interglacial terrace in the Shimajiri Group area decreases southwestwards (Figure 5), a hypothesis that the Philippine Sea plate is being subducted under the Asian continental plate from the southeast or southeast-east can be supported. The distribution of the Late or Middle Pleistocene maximum sea-level height (Figures 4 and 5) shows that the uplifting rate of the pre-Late Miocene area in the Central Ryukyus is faster in the middle part and slower near the Tokara channel and Miyako depression. In the volcanic inner arc area behind the above-mentioned area the rate is rather faster near the two tectonic depressions. These phenomena may be related to the movement of such a left strike-slip fault as predicted for this case by KONISHI (1965).
Most recently (UJIIÉ, OKI and HIGASHIKAWA, 1979), a remarkable island shelf ca. 50 to 70km wide and ca. 360km long has been recognized at the apprroximately 500m water depth on the southern Ryukyu Island Arc by an examination of 1/200, 000 bathymetric charts prepared by the Hydrographic Office of the Maritime Safety Agency for publication (Figure 1). Meanwhile the southern two thirds of the shelf was figured out in details by a number of seismic profiling surveys around the area carried out by the Hydrographic Office (HAMAMOTO, SAKURAI and NAGANO, 1979). In this paper I suspect the shelf as to be covered with a submerged reefal limestone according to the reasons mentioned below. 1) A petroleum exploration well located on the shelf (“WELL” in Figure 1) revealed the presence of 64m thick reefal limestone at the top. 2) Smith-McIntyre-type snapper and Phleger-type gravity corer failed to acquire enough amount of sediment samples from the shelf in the October 1978 cruise aboard on the R/V Kagoshimamaru. Besides, great damage upon the core edges at every time of operation suggests that the samplers always hit on the hard rock. 3) Grain-size composition of the sediments from the shelf rather resembles to that of some sand patches observed on the littoral reef off Okinawa-jima (Figure 3). 4) Benthonic foraminiferal assemblages from the same samples show generic compositions characteristic to the inner neritic environment, differently from those expected from such a depth as 500m. The discrepancy may be ascribed to the present environment similar to coral reef, although Recent littoral reef is characterized by common occurrence of Soritidae, Amphisteginidae, Calcarinidae, and so on, which are seldom on the island shelf. 5) Geomorphology of the island shelf provided with fairly flat surface and sharp marginal edge suggests that the surface is inevitably capped with such hard rocks as reefal limestone resistant against the erosion by water turbulence, particularly by the Kuroshio Current flowed over. 6) Depending upon the same reason, the “Southern Boundary Fault of the Central Ryukyu Island Arc” could well keep its sharp scarpment in the shelf area, whereas it is unable to trace the scarpment on the shelf slope which is composed of the semiconsolidated muddy facies of the Shimajiri Group (Figure 4). If the limestone-capped island shelf around 500m water depth is accepted as a new hypothesis, some significant events will be supposed for the Quaternary geological history of the Ryukyu Island Arc system. Several blocks shallower than 200m, including all island masses and banks, abruptly stand up on the island shelf bounded by many faults running in perpendicular to the Arc trend (Figure 2). Almost all the blocks are also covered with 60 to 70m thick reefal limstone, called the Ryukyu Group, in a narrow sense. Therefore, it is able to regard that these blocks represent relatively upheaved parts of the island shelf. The hereby deduced block-movement corresponds to the “Uruma Crustal Movement” of KIZAKI and TAKAYASU (1976) who proposed this by geological survey on land. By the movement, moreover, the “South Boundary Fault of the Central Ryukyu Island Arc” named in this paper may have been formed in accompanying about 1500m or more down-throw of the northern side, in another word, of the southern end of the Central Ryukyu Island Arc. The Fault must replace the Miyako Depression, which has currently been utilized since KONISHI (1965) for the discussion in structural geology and biogeography; the Depression has ever implied no precise definition in geography. Beneath the reefal limstone spread over the whole of the island shelf region, gently folded or inclined strata of the Shimajiri Group are distributed throughout.
Pattern and rate of late Quaternary tectonic movement of the Ryukyu Islands, belonging to an active island arc, are discussed on the basis of the height distribution of the coral reef terraces and characteristics of active faults dislocating them. The main terrace (S terrace) and the Holocene terrace (H terrace) which are considered to have been formed at the high sea level of the Last Interglacial and Post-glacial ages, respectively, are used as reference surfaces for the estimation of vertical displacement. The Holocene terrace is rather easily identified as the lowest raised coral reefs fringing the present shoreline with abundant 14C dates from the reef complex. Recognition of the S terrace, however, is difficult, because of the absence of the radiometric dating except for Kikai-jima and other few islands. In this paper, therefore, the most extensive and well-preserved depositional surface in each island is tentatively assumed as the S terrace formed at 12-13×104y.B.P. Active faults are investigated mainly by the air-photo interpretation. Results of the investigation of each island are listed in Table 1 and summarized in Figs. 1, 15 and 16. As a whole, the Ryukyu Islands have uplifted since the Last Interglacial age at the latest, judging from the age-height diagram (Fig. 2). The height of the S terrace varies from about 200m to 30m, indicating a notable differential vertical displacement and the H terrace also ranges from 22m to few meters in height as shown in Table 1, Figs. 1 and 15. The Ryukyu Islands can be classified into three zones in terms of pattern of vertical deformation, as represented in a schematic diagram of Fig. 16. The Zone 1 is characterized by a notable westward tilting from Kikai-jima, situated close to the axis of the Ryukyu Trench, toward the eastern part of Amami-Oshima. Such a tilting characteristic to “trench margin” may be an accumulated result of co-seismic deformation associated with the subduction of the Phillipine Sea Plate. The maximum rate of uplift reaches to about 2m/1, 000 years for 12-13×104 years. In Zone II, which consists of main axis of the island arc and may correspond to the “outer arc rise”, an upwarping-like deformation can be recognized. A hinge line of different deformation areas is presumed at the western margin of the Kasari Peninsula of Amami-Oshima. In Zone III, composed of Tertiary and Quaternary volcanic rocks west of the volcanic front, the terrace height of each island shows rather irregular distribution, implying that each island has uplifted independently. Such a zonal arrangement of the deformation pattern is more obvious in the northern area than in the southern area. It may suggest the difference of activity and history of an active island arc between both the areas. Active faults of “certainty I and II” are mapped in Figs. 3 to 14. Generally speaking, they run in NW or WNW (Fig. 17), approximately perpendicular to the trench axis as well as the island arc. Dip-slip faults, probably high-angle normal faults, are predominated, as far as ever known. Consequently, PHS max in the late Quaternary is estimated to be perpendicular to the strike of trench axis. Active faults not only result in the deformation of the coral reef terraces but also control the outline of island in the southern area (Fig. 10). They are more densely distributed on the coral reef terraces than on the ordinal marine terraces and on the hilly or mountainous lands. It is inferred that an existence of active faults may partly correspond to an existence of hard and resistant coral limestone. Activity of the faults usually belongs to “Class B” (10-1m/1, 000 years) or “C” (10-2m/1, 000 years).
Variation at three islands of Nansei-Shoto (-Islands) in the presently occupying elevations of the Last Interglacial (Riss/Würm) emerged reefs, which have been identified stratigraphically with geochemical control of both uranium-series radiometry of hermatypic corals and oxygen isotopic measurment of associated molluscs, are best explained in terms of the neotectonic interplay at plate convergence of the active Ryukyu island are system, of which frontal arc is divided morphotectonically into three major tectonic blocks, Northeast, Central and Southwest Ryukyus. The islands closest to the trench axis (e.g. Kikai) in the Central Ryukyu block are being uplifted rapidly and tilted towards the Asian continent (“reverse dipping”) through compression between the relatively resistive continental lithosphere and gently subsiding (25°-35°in dipping angle of the Wadati-Benioff zone) West Philippine Sea lithoshere in contact (“Chilean-type” of plate convergence by Uyeda and Kanamori). Atolls of at least 50-60m. y. old (e.g. Minami-daito) on the two remnant arcs (Daito and Okidaito) resting on the West Philippine Sea lithosphere to the south of the trench began to uplift to be bulge since Early Pleistocene as the result of lithospheric flexure, when the lithosphere traveled northward and lately approached close to the trench. In Southwest Ryukyu block south of Miyako Depression, the most trenchward island of the frontal arc (i.e. Hateruma) indicates far slow rate of uplift, if ever, in the magnitude of practically almost none, if compared with Kikai, as deduced from the altitude of, besides the referred Last Interglacial, the preceding “Interglacial” (correlative to Deep Sea Core Oxygen Isotope Stage 7: ca. 220k. a. B. P.), which represents the first radiometric confirmation of the Middle Pleistocene in the raised coral reef stratigraphy in the Northwest Pacific. This morphotectonic block sits next to a steeply dipping (55-65° in average and more at the lower tip) Wadati-Benioff zone and tesile field behind the frontal arc where Okinawa Trough (backarc basin) is actively spreading, like the case of the “Mariana-type” of plate convergence. The present study verifies that neotectonic rate of uplift within a frontal arc of the same island arc system varies systematically not only in transverse as “reverse dipping” from trench toward backarc basin or continent, but along its axis (longitudinally), depending upon its morphotectonic blocks which may differ in mode of subduction of convergence underneath.
In this paper, some problems of the late Quaternary sea-level change are discussed, based upon the following topographical data of five straits around the Okinawa Islands. The breadth to depth ratio of the straits (Figure 2) seems to reflect still-standing sea-levels; -110±10m, -80±5m and -50±5m. According to the radiocarbon ages of shallow-water shells and peat obtained from the continental shelf and the coastal plains in the Japanese Islands (Figure 4), it is supposed that the sea-level at the maximum Würm (17, 000 to 20, 000y.B.P.) was about 80m below the present level. Therefore, the result agrees with the evidence from the strait topography around the Okinawa Islands. Even though the sea-level was down to about -80m at the maximum Würm, any land bridge between the Okinawa Islands and other lands can not be expected. Thereby it is inevitable to consider that the immigration of Minatogawa-jin (a paleolithic man) into the Okinawa Islands about 30, 000y.B.P. was made by means of primitive boat or raft.
Fossil bones of the Pleistocene man have been found at the Yamashitacho cave in Naha city and the Minatogawa quarry near the southern coast of the main island of Okinawa. Charcoal fragments from these two sites yielded the radiocarbon date 32, 100±1, 000y.B.P. and 18, 250± 650y.B.P., respectively.
The mammalian fauna of Okinawa in the period from Late Pleistoceneto Recent is discussed. The Ryukyu Islands are known for abundant occurrence of deer fossils of Miocene type. The Yamashita-cho site of fossil man, 32, 100±1, 000y.B.P. (TK-78) in age, yields deer fossils and no other mammals. On the other hand, at the site of the Minatogawa man, 18, 250±650y.B.P. (TK-99) to 16, 600±300y.B.P. (TK-142) in age, deer fossils become few while boar fossils increase remarkably. This indicates that the Ryukyu boar gained ground taking the place of deer. Wild boars are living in such islands as Amami Oshima, Okinawa, Ishigaki and Iriomote, and many shell-mounds reveal the existence of boar. In all probability, migration of Sus must have taken place since the time of the Minatogawa man, that is, during the Würm maximum. Geologically the Ryukyu Islands are supposed to have been already separated from the continent at that time, but the occurrence of Sus suggests that the islands were partly connected with the continent even in the Würm maximum.
Environments of the sites in the Ryukyu Islands differ in many aspects from those in Kyushu and other parts of Japan. They largely reflect unique and local topography and geology. In the Amami, the Okinawa and the Sakishima Islands, in the Ryukyus a great effort has been paid on archaeological chronology by using excavated potteries. The present study involves stratigraphy and absolute ages of some pottery-bearing formations distributed in those areas. As shown in Table 1, the present study makes it possible to correlate these pottery-bearing formations one another, and the Holocene chronicle of the Ryukyus was affiliated with that of Kyushu for the first time.
Recent archeological interests in Yaeyama district are as follows: 1. Characters and dating of the neolithic aceramic culture which has been chronicled as to be of the First Period. 2. Problem about the “Red Ware” which was newly discovered. 3. Re-examination of previous chronological framework about the prehistoric sites. 1. Recently the Funakuya, Kanda and Nagura-B shell mounds, which have been known as the sites of the neolithic aceramic culture, are investigated in purpose to find out any occurrence of pottery. But no pottery fragments were discovered from the includes of all the sites. In the chronological framework, the site of the neolithic aceramic culture has been regarded as the oldest and as fundamental of the Yaeyama prehistorical culture. Referring to the stratigraphic relationship between the Otabaru and the Kanda sites, however, it became clear that the previous idea should be re-examined. Many data recovered hitherto and comparative study of the cultural remains reveals that this culture had developed in not so old age but probably during from 5 to 13 A.D. The culture was characteristically developed near the river mouth, on the sea coast or sand dune accompanying with no pottery but stone artifacts and shell artifacts. Usage of the earth-oven cooking together with fire stones is supposed. The stone artifacts are composed of stone axes in abundance, beat-stones and whet-stones. The stone axes are large-sized in comparison to those of the other periods and, in relatively many cases, polished or semi-polished. Characteristic technique making the stone artifacts may be pinpointed on the polished-blade, which seems to have been pursuited by the following Yaeyama prehistorical cultures. Considering by the location of the sites and the combination of cultural remains, the life of this period may have been supported through fishing and esculent culture, namely, Imo-Culture. 2. Recently the Fune No. 1 site and the Otabaru site in the Ishigaki Island yielded new form pottery which differs from the prehistoric pottery reported previously from Yaeyama district and called as Yaeyama style pottery (provided with external lugs). This pottery was named Red Ware because of reddish tint. The ware is hard baked with generally thick wall and designed by nail impression. Red Ware-type includes the associated potteries without design and with ox-horn handle. This type of potteries is widely distributed in the Yaeyama Islands and is thought to be the oldest among the pottery cultures without imported chinaware. Besides, it became clear that it precedes the aceramic culture by stratigraphic survey of the Otabaru and Kanda sites. The Red Ware culture accompanies stone axes which are generally small sized, and beaten or semi-polished. A 3, 850±75y.B.P. 14C age is measured for the Otabaru site including the Red Ware. This age is not incompatible with the stratigraphic sequence. Standing upon the 14C age, a gap exists between the Red Ware and the following aceramic culture. Beside of this problem, searching the culture anteceded to the Red Ware is a subject to be progressed in the comming future. 3. Judging from the situation mentioned above, it become necessary to re-examine the previous chronicles, Development of the Red Ware culture may have been antecedent to the “aceramic culture of the first period” in the chronicle proposed by the Waseda university. After this aceramic culture, culture yielding Yaeyama style pottery accompanied with imported chinaware appears. The pottery and the Red Ware are belonged to the same lineage, although they are separated by the aceramic culture from each other. Therefore, it can be presumed that peoples who ceased to use pottery once in the aceramic culture began to use pottery again. Social background of the repeated usage of pottery beyond a gap is also the subject to be solved in the future.