Many tephra layers erupted from volcanoes in southern Kyushu, SW Japan. This paper reviews chronological studies of tephra layers during the past 30, 000 years in this area. The methods for determination of eruption ages are classified into four categories: (a) comparison with historical documents, (b) position on the calibration curve, (c) projection of ¹⁴C dates to the calibration curve, and (d) estimation from stratigraphic position. Eruption ages of two widespread tephra, the Kikai-Akahoya (K-Ah) and the Aira-Tn (AT) are estimated to be 7.3cal ka BP and 29cal ka BP based on their positions on the calibration curve. Eruptions preceding and following the large-scale ignimbrite eruptions occurred within 1, 000 to 3, 000 years. Eruption ages of the Sakurajima-Satsuma (Sz-S/P14), Ikedako (Ik) and Kirishima-Mi-ike (Kr-M) tephras, widely distributed in southern Kyushu, are approximately 13cal ka BP, 6.4cal ka BP, 4.6cal ka BP, respectively. Thus, it seems that the Ik tephra erupted after an interval of about 1, 000 years from the K-Ah eruption. In order to estimate precisely this time interval, we have to determine the position of the Ik tephra on the calibration curve.

View full abstract

Changes in depositional environments during the post-glacial period in Kagoshima Bay and seas around the northern part of the Ryukyu Islands are deduced on the basis of the analyses of ten sediment core samples which were recovered from ten boreholes on the Kagoshima City, alluvial plain and from two boreholes in Naze Bay.
The sea-level curve for post-glacials reported from the stable continental shelf of the East China Sea suggests that the sea level rose rapidly during the period ranging from the last glacial to 7, 000yrs BP. The benthic and planktonic foraminiferal data from two sediment cores indicate that deposition in Naze Bay began about 9, 300yrs BP under very shallow inner bay conditions, scarcely influenced by open sea water; with the passage of time, the bay deepened, with concomitant increasing degrees of flow of open sea water. The change in depositional environments was caused by a small supply of sediments from rivers and tectonic subsidence of the Naze Bay area. On the other hand, the benthic and planktonic foraminiferal data from the Alluvium in Kagoshima City indicate that sea-level changes synchronized with water depths during the period ranging from 10, 000 to 7, 000yrs BP. The good preservation of shallow inner-bay environments during the period may be attributable to a large supply of sediments mainly from the pyroclastic deposits distributed in this area. During the period ranging from 7, 000 to 5, 000yrs BP, the transgression reached its maximum extent to develop an alluvial plain, the rudiments of the recent alluvial plain.

View full abstract

Holocene geomorphic evolution, sea-level change, and tectonic movements around the Aira caldera are discussed based on geomorphic and depositional features, AIMS ¹⁴C ages, tephrochronology, archaeological remains, and assemblage analysis of diatom and mollusca of marine terraces and alluvium on the coastal plain along the northwestern coast of Kagoshima Bay. The results show the relative sealevel changes summarized below and uplift which partially affected geomorphic evolution in Holocene time.
Marine terraces which have emerged since the maximum stage of postglacial transgression ca. 8, 000cal BP (ca. 7, 200yrs BP) are distributed in three levels of the lowlands along the northwestern coast of Kagoshima Bay. Emergence dates of these marine terraces are 7, 300cal BP (6, 500yrs BP)-3, 500yrs BP, ca. 3, 000-2, 000yrs BP and ca. 1, 500-1, 000cal BP, respectively.
Previous sea-levels along the northwestern coast of Kagoshima Bay were +4-5m at 8, 500-8, 400cal BP (7, 700yrs BP), +6m at ca. 8, 100cal BP (ca. 7, 300yrs BP), +12m around 7, 300cal BP (6, 500yrs BP: the maximum highstand), and declined stepwise from +5-7m (3, 000-2, 000yrs BP) to +2-3m (1, 500-1, 000cal BP).
Height distributions and ages of the marine terraces and marine top sediments revealed by this study imply the existence of uplift accompanied by tilting toward the center of the Aira caldera. Observed maximum uplift rate is 12m per 7, 300cal yrs.
Coastal landform evolution has been greatly affected by the volcanic products that erupted from the maars around the Aira caldera. Reclamation of the inlets by the Yonemaru base surge (8, 100-8, 000cal BP) was particularly large-scale.

View full abstract

Volcanic damage caused by the four largescale eruptions of Sakurajima volcano in historic time is first outlined, and then, on the basis of the eruptive history, the damage in prehistoric time are also summarized. Not only large-scale eruptions but also intermittent vulcanian eruptions have been identified.
The sequence of large-scale eruption was as follows: first, precursory phenomena; second, the beginning of plinian eruption; third, generation of pyroclastic flows and surges; fourth, outpouring of lava flows; and finally, broad subsidence of the ground after the eruption. Such large-scale eruptions occurred 17 times (-19 times), and the ca. 11 ka P14 eruption was the largest among them; the erupted volume of P14 was about 11km³, causing severe damage all over the southern Kyushu. The P13, which overlies the Uenohara remains, was the second largest tephra (1.3km³). Due to the heavy ash fall, settlement in the area was estimated to be difficult for some years after the eruption.
Large quantities of fallen ash from the plinian eruption not only buried farms and houses but also became the generating factor of lahars for a few years. Many lahars occurred especially in the region mantled by ash more than 30 centimeters thick. Vulcanian eruptions did not cause any volcanic disasters around the periphery of Aira caldera, but caused many lahars in Sakurajima Island during heavy rains.
In order to reduce the volcanic hazard, it would be necessary not only to intensify hard countermeasure but also to conduct careful hazard mapping and disaster prevention education.

View full abstract

In the early history of Japan, tephra was used archaeologically for the division of relics. Tephras were used and distributed in a wide range of areas, allowing for the comparison of chronological dating of relics and traces, and the relation or extinction between upper and lower archaeological materials.
During the 1990s, the discoveries of villages which were directly covered with the tephras increased. Through excavation of the Kuroimine site, the Nakasuji site, the Hashimuregawa site, and the Old Kanbara village, we could obtain concrete information for reconstructing the disasters and the structures of villages, which were buried by volcanic eruptions during earlier ages.
These discoveries have opened new discussions, including discussion of the continuity or extinction of the archaeological materials which were in the upper and lower layers, along with specific consideration of the process of the disaster. In short, the discussion was able to solve the following problems: (1) What kinds of cultural items were damaged in the process of the accumulation of the tephra?; (2) What kind of assessment was done by victims of the rapid change of environment?; (3) What kind of strategy was chosen to deal with it. When we proceed from an a priori idea such as ‘A disaster must influence a culture’, it seems that we have to seek more concrete information of the traces for reconstruction of the ancient victims' actions.
Through comparison of two volcanic eruptions of Mt. Kaimondake in the Hashimuregawa archaeological site, we recognized that there is a difference in strategies for adapting to the disasters between the two cases. The evidence of shell mound formation begins in the 6th century, and ends in 874 A. D. When we consider the continuity of the village, we can guess that the same community continued, because the second shell-mound was formed in the same place as the first. In addition, the type changes of the potteries and of the shapes of the dwelling-pits occurred without regard to the disaster. This point of change corresponds to the period of the Ritsuryo-system extension over this area. The evidence of the last quarter of the 7th century indicates that the change of life-style was not due to the disaster, but to the expansion of the Ritsuryo-system.
In addition, as indicated by the analysis of the traces of activities of ancient people, the evidence from the last quarter of the 7th century indicates that victims chose to continue their lives, while the evidence from 874 AD suggests that the victims abandoned their village. Thus, looking at the two volcanic disasters, I think that the victims' evaluation was different. In general, the disaster influenced culture, but that alone is not enough to explain the change of culture along with natural disasters.

View full abstract

Many clastic dykes are found not only in southern Kyushu but also in Tanegashima and Yakushima islands far south of Kyushu, which are closely related to the 6.5ka BP eruption from Kikai caldeara. Clastic dykes found in southern Kyushu are mainly sand dykes which were formed by the liquefaction of fluvial deposits. Similar dykes are also found in Tanegashima and Yakushima islands; these which are composed mainly of breccia and/or pebbles probably derived from talus and/or terrace deposits. Those dykes penetrate the overlying soil layers and reach the tephra of the 6.5ka BP eruption of Kikai caldera. Tephra sequence from this eruption consists of plinian pumice, ignimbrite, and co-ignimbrite ash (Akahoya ash) in ascending order.
Field evidence suggests that breccia dykes observed in Tanegashima and Yakushima islands were formed shortly before, or at nearly the same time as, the ignimbrite eruption. However, sand dykes observed in southern Kyushu were formed in two different stages. The first group was formed at the same time as breccia dykes in Tanegashima and Yakushima islands. The second group occurred a bit later: that is, liquefied material is apparently intercalated in co-ignimbrite ash (K-Ah). This evidence indicates that two large-scale earthquakes occurred during the 6.5ka BP eruption from Kikai caldera. The first earthquake occurred around Tanegashima and Yakushima islands was quite severe, and affected a wide area of southern Kyushu. The second one, which occurred a bit north of the first one, is also considered to be a large-scale earthquake.

View full abstract

A palynological study was made on the Holocene sediments (KY core) of the Kimotsuki lowland, Osumi Peninsula, southern Kyushu, Japan. The development process of lucidophyllous forest and the influence of the Kikai-Akahoya eruption (6, 500yrs BP) on the vegetation around the Kimotsuki River were discussed. Four local pollen assemblage zones, KY-I to KY-IV, were established based on the changes of the major arboreal pollen. The stratigraphic sequence of pollen in the core sediments revealed the vegetation to be composed of a warm-temperate deciduous broad-leaved forest stage, a stage of warm-temperate deciduous and evergreen broad-leaved forests which accompanied the Celtis-Aphananthe forest, a lucidophyllous (Castanopsis) forest stage, and a lucidophyllous (Castanopsis-Quercus subgen. Cyclobalanopsis) forest stage, which corresponded to four zones (KY-I to KY-IV), respectively. Lucidophyllous elements began to expand at 9, 200yrs BP. The Castanopsis forest dominated at 8, 000yrs BP and flourished for 1, 500 years before the Kikai-Akahoya eruption. The vegetation which had been damaged by the eruption recovered as a Castanopsis-Quercus subgen. Cyclobalanopsis forest at 6, 200yrs BP and continued by 4, 000yrs BP. The investigated area is located at the northern end of the Koya pyroclastic flow deposit. As the flow deposits were not uniform in thickness and distribution around the area, the influence of the Kikai-Akahoya eruption also might not be uniform. The time required for the vegetation to recover after the eruption was relatively short, and is estimated to be 100 to 300 years, as a minimum, when observed on the whole at the terrace and the mountain area around the Kimotsuki River.

View full abstract

This study aims to clarify the impact of the explosive eruption of Kikai Caldera (7, 300cal BP) on vegetation in southern Kyushu, Japan, based on phytolith studies.
A significant change in phytolith assemblages below and above the Koya pyroclastic flow (K-Ky) elucidated that the hazardous pyroclastic flow had changed vegetation from lucidophyllous forests such as Castanopsis and Lauraceae associated with bamboo grass bush to grassland represented by Miscanthus in southern Osumi and Satsuma Peninsula. The recovery of the lucidophyllous forests in this area was considered to start about 600-900 years after the volcanic event, based on the hiatus between the erupted ages of the tephras, which was estimated as 600-900 years. However, the forest vegetation in these areas was partly in the process of recovery. The damage to lucidophylllous forests in the area northward from the mid part of Kagoshima prefecture, which the K-Ky pyroclastic flow did not reach but which was affected by the intensive Akahoya ash fall (K-Ah) alone, was considered to be relatively small.

View full abstract

Applying the age of Kikai-Akahoya eruption to the pottery chronology of the Earliest stage of the Jomon period in Kyushu indicates that the eruption occurred between the old stage of Todoroki-A type pottery and the new stage. Moreover, the making of Todoroki-A type pottery, which had wide distribution in the whole of Kyushu, was not disrupted by the Kikai-Akahoya eruption. In addition, I evaluated the recovery process of the living environment after the Kikai-Akahoya eruption on South Kyushu; most of the southern section received a direct hit of Kikai-Koya pyroclastic flow. We can confirm the condition of the sites only in the northern part of South Kyushu immediately after the Kikai-Akahoya eruption, but it is assumed to have been impossible to live in the southern part of South Kyushu, which was directly affected by Kikai-Koya pyroclastic flow. It seems that the living environment had recovered nearly completely by the middle stage of Todoroki-B type pottery (about 5, 500yrs BP) in almost all the inland part of South Kyushu mainland. But the scales of the sites are poor even in the southernmost end part of the Osumi islands or Satsuma and Osumi Peninsula which are near Kikai caldera and sustained serious volcanic damage. We can also recognize residential stability in the whole of South Kyushu after the phase of Sobata type pottery (about 5, 100yrs BP). Besides, it is likely that forest vegetation which produced nuts was considerably damaged by the Kikai-Akahoya eruption, based on the fact that the rates of polishing stones and saddlequerns in the composition of stone implements which were used for milling nuts is extremely low in deposits from immediately after the eruption in the whole of South Kyushu. It is inferred from the increase in the rate of polishing stones and saddlequerns in composition that forest vegetation recovered after the middle stage of Todoroki-B type pottery, at the latest, in the northern part of South Kyushu, and between the new stage of Todoroki-B type pottery and the phase of Sobata type pottery in the southern part of South Kyushu.

View full abstract

[in Japanese]

View full abstract