This is a summary of the presidential address to the public at the 19th annual meeting of the Japanese Association for Quaternary Research, held at Tottori August 18-20, 1989. The address had two parts: one was an introduction to the concept of “Quaternary” and the significance of Quaternary research, and the other was an elucidation of the discussion at the symposium on “Paleogeography and Paleoenvironments around the coastal areas of the Japan Sea”, exemplified by elephant fossils from the sea-bottom of the southern Japan Sea. The pioneers of Quaternary research in Japan, the German geologists E. NAUMANN and D. BRAUNS, were very interested in the elephant fossils of Japan and published papers on them in 1881 and 1883, respectively. Since that time, the study of elephant fossils and Quaternary research in Japan have been closely related.
Since about 20 years ago, several elephant tusks and molars have been dredged by dragnet fisheries off the San'in district and also off the Noto peninsula in the southern Japan Sea. They were obtained from depths of 120m to 400m, on either the continental shelf or on the drowned bank of the sea-bottom. Formerly, those materials were considered to verify the presence of a landbridge in the past around the area of the Tsushima strait between Korea and Kyushu, but now this idea has come to be rejected. The results of analysis for the boring cores drilled at several places on the bottom of Japan Sea afford much information about paleoenvironmental changes during the Late Pleistocene. For example, an inflow of the Tsushima warm current to the Japan Sea, which is one of the remarkable tributaries of the Kuroshio, was reduced or arrested at that time by mixing with fresh water from the Hwang Ho River running through North China. This created a stagnant condition in the bottom water of the Japan Sea, and much influenced the biotic community there. Again, the inflow of the Tsushima warm current to the Japan Sea was regenerated beginning about 6, 500 y. B. P.
The elephant fossils on the sea-bottom consist of tusks and molars of Naumann's elephant Palaeoloxodon naumanni and a molar tooth of woolly mammoth Mammuthus primigenius. The radiocarbon dating carried out for these and others found on land showed that the former was older than 30, 000 y. B. P. and the latter was younger, around 20, 000 y. B. P. It is known that during the maximal cold phase in the Last Glacial, woolly mammoths came down from Siberia southward to Hokkaido, but did not cross over the Tsugaru strait to Honshu. Therefore, it is reasonable to assume that the remains of woolly mammoths would be transported by drifting from the Hwang Ho area to that off San'in and then sink to the bottom of the Japan Sea. However, the idea is still a matter of imagination, more investigation for the environmental changes during the Quaternary in Japan needs to be done.

View full abstract

Changes in the palaeoenvironment along the Japan Sea since the early Pleistocene are reviewed with reference to the level of sea water and to the connection between islands and the continent and the nature of ocean currents. In addition, the sea level during the Murodo and Tachikawa stages is discussed. The nine palaeoenvironment stages in ascending order are as follows; 1. Omma, 2. Utatsuyama, 3. Riss II, 4. Hiradoko, 5. Murodo, 6. Tachikawa, 7. Nyuzen, 8. Jomon transgression, and 9. Yayoi regression.
Early Pleistocene
1. Omma stage; older than ca 0.8Ma B.P.
Environmental analysis of the Omma-Manganji fossil molluscs shows that the main cold currents flow from the north and the occasional warm currents from the south.
Middle Pleistocene
2. Utatsuyama stage; ca 0.5-0.3Ma B.P.
This stage is represented by the Utatsuyama, Hanyu and Kurehayama formations in the Hokuriku region and the upper part of the Uonuma group in the Niigata area; these formations consist mainly of terrestrial materials. An embayment molluscan fauna such as Batillaria or Anadara is sometimes found in these formations, which shows a marine enviroment near the depositional basin of these formations.
3. Riss II stadial; ca 0.16Ma B.P.
Although there is no direct evidence of this stage, the climate of the stadial is proved by the distribution of morains and other evidence which suggests a climate colder than those of the Würm I, II, and III stadials. At the Riss II stadial the sea level would have been 140m lower than the present, if sea leval change corresponds to the air temperature. Therefore it is reasonable to think that the Japan Sea at that time was enclosed and that land bridges connected the islands to the Asian continent.
Late Pleistocene
4. Hiradoko stage; ca 120ka B.P.
There are several marine terraces which contain marine shells along the Japan Sea: Hiradoko in Noto, Katanishi in Oga and Kurobe in the Tango area. Ecological analysis of molluscs in these marine terrace deposits shows that the sea water temperature at this stage was the same as, or a little warmer than, that at present. The sea level may have been 5m higher than at present.
5. Murodo stage (stadial); ca 60-40ka B.P.
These are two glacial stadials in the Würm age in Japan. The earlier one is the Murodo, and the later one is the Tateyama. To build a glacial cirque requires a lot of snow, which suggests warm currents flowing into the Japan Sea. The sea level at this stage must have been about 100m deeper than at present.
6. Tachikawa stage (Tateyama stadial); ca 20ka B.P.
The coldness of the Tateyama stadial was less severe than that of the Murodo stage. So the sea level was not deeper than that of the Murodo stage, presumably 80-100m deep. But warm currents still flowed into Japan Sea before the beginning of this stage, and these warm currents opened the Tsushima and Korea straits.
Holocene
7. Nyuzen stage; ca. 10ka B.P.
A Nyuzen submerged forest is found 40m under the sea on the continental shelf off the Kurobegawa alluvial fan, Hokuriku region. The climate of this stage was 2-3°C colder than at present.
8. Jomon transgression stage; ca. 6ka B.P.
Warm currents flowed into Japan Sea. The sea level was 3-6m higher than at present. This transgression invaded most of the alluvial plains around the Japanese Islands.
9. Yayoi regression stage: 2ka B.P.
An Uozu submerged forest has been found 1-2m under the sea; such a phenomena have been found in several coasts along the Japanese islands.

View full abstract

Hokkaido constituted the northern landbridge between Japan and the Asian continent during most of the last glacial, since the Soya Strait, between Hokkaido and Sakhalin, continuously emerged after about 60ka B.P. This paper reconstructs the permafrost environment and vegetation of this northern landbridge from the distribution of fossil periglacial phenomena and pollen data. The edge of the continuous permafrost zone shifted southwards, down to about 45°N in northeast China, and to about 43°N at the southern foot of the Shikhote-Alin' Range in easternmost USSR. The southern boundary of the discontinuous permafrost zone reached about 40°N in Hebei, 44°N in northeast China, and 40°N at the southern foot of Changbai Shan, North Korea. The limit of each zone runs southward from the Pacific side inland, just as the present limit of the permafrost zone does. The lack of typical ice-wedge casts and pingo scars on the one hand, and the richness of soil wedges and palsa scar-like features on the other, in northern and eastern Hokkaido strongly suggest that most of Hokkaido was located in a discontinuous zone of permafrost during the last glacial. The continuous zone seems to have covered only the northernmost part of Hokkaido. The fossil pollen assemblage from the coldest phase (25-20ka B.P.) in Hokkaido is characterized by Larix gmelini, Pinus pumila and Picea (probably P. ezoensis and P. glehnii). Although NAP pollens dominate in several localities, especially in eastern Hokkaido where volcanic activity continued through the last glacial, the dominance of light and dark conifers in the pollen assemblage indicates the spread of boreal forest in the main part of Hokkaido. This suggests that, during the last glacial, Hokkaido was part of the refugia for the taiga zone which consisted of Larix dahurica, very similar to (or synomymous with) species of Larix gmelini. On the assumption of a lowering of the vertical zonation by 1, 500m, the alpine tundra spread down to about 750m a.s.l., thus taiga forest covered the main part of Hokkaido. The vegetation of northern and eastern Hokkaido, which some researchers regard as tundra or forest tundra, seems to have been a complex one made up of patches of coniferous forest, grassland, mire, and alpine tundra.

View full abstract

The problem of the maximum depth to which the sea level dropped during the last glaciation is even more difficult and obscure than that of high sea levels. Some evaluations have been proposed, based on the estimations of the depth of the Holocene sediment base in the coastal plain or on the consideration of submerged sediments or geomorphological features now found on continental shelf. Many Japanese geologists estimate -100 to -140m for the last glacial low stand sea level, but some doubt is cast on their data and opinions. Our estimate of -80±5m was a conservative interpretation based on dates for shells and peat obtained from the shelf and coastal plains of the Japanese Islands.
1. Charts show that wherever sizable inland seas are separated from the ocean by narrow straits, current erosion holes (sea caldrons) exist either in the narrow straits or directly adjacent to them. The Tsushima strait is a part of the shelf, although there is a relatively deep hole (to -160m) along the east side of Tsushima Island. The depth of this deep hole attains 50m below the surrounding relict wave-cut terrace (-110m to -115m). In this case, the depth of the relict wave-cut terrace may reflect the sea level (about -100m) at the time of the Tsushima strait formation.
2. As a result of continuous seismic profiling surveys in Ishikari Bay, the buried valleys on the continental shelf are known to be deeper than 90m. The buried valleys were formed at the stage of the lowest sea level, during the last Glacial Age. However, we observed that Ishikari river may be 10m to 15m deep near its mouth. The available data indicate a relative rise of sea level since the valley was cut, amounting to about 80m.
3. The breadth to depth ratio of the straits around the Japanese Islands seems to reflect each still standing sea-level stage, such as -100±10m, -80±5m and -45±5m, after the formation of these straits.
4. For most of the Pleistocene, the Japanese Islands were connected with the Korean Peninsula, and the present major islands themselves were tied to each other. It is sure that large mammals such as elephants migrated into the Japanese Islands through land bridges. In the early Shimosueyoshi transgression (about 100, 000y.B.P.), when the sea level was about -100m, the Japan Sea was connected with the Pacific through narrow passages located in the Korean and Tsugaru straits. At the time of the last Glacial Age, when the sea level fell to about -80m, land bridges between Honshu and other lands were never formed. We now believe that it might have been 12, 000 years ago when the sea level rose to about -45m. This was the final stage of the land bridge in the Soya strait between Sakhalin and Hokkaido.

View full abstract

A long-range, tropospheric eolian dust transported from the Saharan desert and the Asian continent has been deposited on the terrestrial and aquatic environments in the northern hemisphere. Soil loss due to wind erosion in the arid and semiarid source areas is more significant than previously assumed. Global emission of desert dust and mineral aerosol material is estimated to amount to more than 1.0×10⁶ton yr^-1. Long-range eolian dust is an important factor in soil formation and nutrient input in many deposition areas. Physical, chemical, and mineralogical characteristics of long-range eolian dust derived from the Takla Makan and Gobi deserts and the Loess Plateau in China and their influence and significance to the soil and paleosol formations in Japan and Korea are reviewed in this paper.
The long-range eolian dust in East-Asia was characterized by a predominance of soil particles 3 to 30μm in diameter. Their dominant minerals were 2:1 layer silicates, kaolinite, quartz, and feldspar. Nonallophanic andosols, red-yellow soils developed on limestones, basalts, and other diverse parent materials, and paleosols buried in paleodunes in the area along the coast of Japan Sea, were strongly influenced by the long-range eolian dust derived from China. Oxygen isotope abundance of the fine-grained quartz (1 to 10μm) isolated from soils revealed that fine quartz and 2:1 layer silicates in diverse soils and paleosols in Japan and Korea and pelagic sediments in the Japan Sea were of eolian dust origin. The eolian dust flux from the atmosphere to terrestrial environments in Japan is significant in the heavy snowfall area along the coast of Japan Sea and was more prominent in the last Glacial age than in the Holocene. Dust flux from East China Sea, Yellow Sea, and Japan Sea pelagic sediments dried during the marine regression period in the last Glacial age to soils and paleosols was also significant in Japan. Thus the desert dust phenomenon is of relevance to geophysical science in general, e. g. geography, geochemistry, climatology, soil science, ocean sedimentology, and Quaternary studies. Desert dust emission and long-range transport are useful indicators for dynamic change in the tonal circulation system, influencing the discussion of future climatic change.

View full abstract

In the eastern and middle San-in districts, on the southern coast of the Western Japan Sea, the following floral change is recognized since the middle Pleistocene.
There were three types of flora during the middle Pleistocene. First was the interglacial flora composed of cool temperate forest elements such as Fagus crenata, Quercus sp., Cryptomeria japonica, etc., and a pollen assemblage which is predominated by Fagus or represented by Fagus and Cyptomeria. Second is the glacial flora composed of subarctic elements as Picea jezoensis, Tsuga diversifolia, Menyanthes trifoliata, etc., and pollen assemblage represented by Pinus Subgen Haploxylon and Picea. Third is the interstadial flora composed of pollen assemblages represented by Picea and Cyptomeria.
During the late Pleistocene, a fourth type of flora appeared, composed of warm temperate forest elements with pollen assemblages containing Cyclobalanopsis. Moreover, some species occur in each type of flora: Pinus koraiensis occurs in glacial flora, Lagerstroemia pollen in interglacial and Buxus microphylla var. japonica and Pseudotsuga japonica in interstadial. By means of pollen analysis, the latest Pleistocene and Holocene pollen zones are clearly identified as follows.
Quercus and Alnus PZ: before about 10, 000y.B.P.
Quercus and Aphananthe-Celtic PZ: about 10, 000 to 9, 000y.B.P.
Fagus and Tsuga PZ: about 9, 000 to 7, 000y.B.P.
Pinus and Abies PZ: about 7, 000 to 5, 500y.B.P.
Cyclobalanopsis and Castanopsis PZ: about 5, 500 to 2, 400y.B.P.
Gramineae PZ: about 2, 400y.B.P. to present
Cryptomeria Subzone: about 2, 400y.B.P. to about A. D. 700
Typical Subzone: about A. D. 700 to 1, 500
Pinus Subzone: about A. D. 1, 500 to later than 1, 900
Pinus and Cyptomeria Subzone: later than A. D. 1, 900 to present.

View full abstract

This report is a summary of 17 Proboscidean fossils discovered in the Japan Sea. Although most specimens were discovered off the coast of the San'in district (N35°10′-36°10′, E131°30′-132°), molars were also discovered off the coast of the Noto Peninsula and northeast of the Yamato ridge. These specimens were identified as Palaeoloxodon naumanni (MAKIYAMA), with one specimen being Mammnuthus primigenius (BLUMENBACH). ¹⁴C datings indicates their ages to be about 23, 000y.B.P.-38, 000y.B.P.
It is impossible to consider that M. primigenius migrated from Hokkaido to Honshu over a land bridge in the last glacial period as the manifestation of mammal fauna for Quaternary geohistory. It is desirable to take into consideration that the remain's most likely drifted from China or that the animal crossed the channel by swimming.

View full abstract

The purpose of this paper is to present the result of recent investigations into the palaeolithic sites of the Chugoku Mountains, and to hypothesize some routes of cultural contact in southwestern Japan, with particular reference to the Japan Sea side area.
1. The Palaeolithic cultural chronology in the Chugoku mountainous region can be recognized as follows: The knife-shaped tool tradition is divided into phases, I, II and III. Phase I is characterized by Moro-type or Tatenogahara-type knife-shaped tools or trapezoids. Axe-shaped tools are associated with most assemblages. Phase II is marked by more elaborated Moro-type knife-shaped tools and the Sunagawa-type blade technique. Phase III is distinguished by the predominance of knife-shaped tools on wide flakes. The micro-blade tradition includes two kinds of assemblages: one is characterized by subconical micro-cores, and the other contains wedge-shaped micro-cores made by the Yubetsu technique.
2. The distribution of eight palaeolithic sites in the Chugoku Mountains seems to suggest the appearance of a traffic route along the ridge of the mountains during phases I-II of the knife-shaped tool tradition; it also suggests the development of other routes crossing the mountains during phase III. It can be thought that there was a main route along the coast of the Japan Sea, which branched off the route along the ridge of the Chugoku Mountains in the San'in region. Since wedge-shaped microcores were recently unearthed at Onbara site, Okayama Prefecture, it should be possible to pursue the route connecting the micro-blade industries in the Korean Penisula and on the Japan Sea side of Honshu, Japan.

View full abstract

Paleomagnetism was measured with a cryogenic magnetometer and a spinner magnetometer for 458 semi-consolidated mud samples from eight sections of the Plio-Pleistocene Tsunozu Formation at its type locality, the Tsunozu area, San'in district. The results, revealing a magnetostratigraphic sequence with repeated magnetically normal and reversed intervals, are discussed with regard to the chronology of the formation. Based on previous data on magnetostratigraphy and the fission track age of the Oe-Takayama Volcano Group northeast of the Tsunozu area, a correlation with the standard paleomagnetic polarity time scale leads to the conclusion that the entire Tsunozu Formation ranges from just before the Cochiti normal event (about 4.1Ma) to just before the Jaramillo normal event (about 1Ma), and that the Plio-Pleistocene boundary (1.64Ma) is situated at the horizon just above the M3 marine clay bed in the Middle Member. An approximate chronology of four marine clay interbeds within the Middle Member is: M1 bed, 3.8-4.0 Ma; M2 bed, 2.5-2.7Ma; M3 bed, 1.7-1.8Ma; M4 bed, 1.1-1.5Ma, in ascending order. The apparent sedimentary rate of the Middle Member is about 2cm/10³y. Our estimation of the stratigraphic horizon of the Plio-Pleistocene boundary is quite consistent with the previous view based on pollen stratigraphy.

View full abstract

The existence of a land bridge in Tsugaru Strait, Northeast Japan, is hypothesized with reference to caldron-saddle topography and submarine terraces in the western part of the strait and to late Pleistocene marine terrace deformation on land. It is possible that the Tsugaru Land Bridge was formed after the lowest submarine terrace (Terrace VII) and, before the paired caldrons. If the land bridge existed in the last glacial maximum stage, 18 to 20ka ago, the sea level is estimated to have fallen below the depth of 130 meters. However, Terrace VII has not been age-determined yet. Tectonic movement, deduced from marine terrace deformation on land is not detected from height changes in continental shelves, caldrons, and saddle features. Differential movement is also not found in the height distribution of submarine terraces, at least in the Tappi Saddle. Much data on subsurface geology and structure beneath continental shelves are necessary in order to know the age of the Tsugaru Land Bridge.

View full abstract

Characteristic distributions of shallow crustal earthquakes and active faults can indicate tectonic activity, because both are thought to arise chiefly from tectonic instability.
Central Kyushu is the most seismically active region of the island. This active seismogenic region can also be defined as a NE-SW-trending active normal fault zone, which would lead to progressive spreading. On the other hand, northwestern Kyushu, including the Tsushima Strait, is postulated to have a low concentration of strain energy, and a comparatively stable condition is inferred in this region in terms of geomorphic indices and landforms. Thus, the Tsushima Strait region cannot be regarded as a potential tectonically active zone. However, the tectonics of the region are still not understood in detail. Sufficient information on Holocene marine terrace surfaces and on submarine tectonic structures in and around the Tsushima Strait remains to be obtained.

View full abstract