A Review of Coral Studies of the Ryukyu Island Arc to Reconstruct Its Long-Term Landscape History
Coral studies of the Ryukyu Island Arc have resulted in the accumulation of a vast knowledge quite relevant to a future reconstruction or interpretation of its landscape history, i.e., the process of viewing a present landscape as a historical artifact contingently and cumulatively changed through interactions between two types of agents, natural and human. For the geohistory of the late Miocene to the late Holocene, in particular, I review three topics: insularization of the Ryukyu Arc, coral growth during the last glacial maximum, and the formation of a coral reef environment in the middle Holocene. The latter two topics are considered in connection with archaeological and anthropological evidence of late Pleistocene and Holocene human settlement and subsistence. The latest phase after the middle Holocene, however, appears to still not be fully examined in geohistorical coral studies; thus, I briefly describe my geoarchaeological surveys of the shallow sea and alluvial land in the Nagura region of Ishigaki Island, the future examination of which will make us aware of the difficulty of simply determining anthropogenic changes as either environmental degradation or landscape enhancement.
KeywordsCoral studies Landscape history Geoarchaeology Ryukyu Island Arc
A present landscape can be viewed as a historical artifact of long-term interactions between two types of agents, natural and human. The holistic elucidation of such a landscape should be facilitated by transdisciplinary approaches that bridge the natural sciences and the humanities. For instance, archaeology in the Pacific through the past 20 years has contributed substantially to understanding this issue from the viewpoint of historical ecology, which incorporates data collected from a variety of natural sciences, including geology, geomorphology, paleontology, and paleobotany, as noted by Kirch and Hunt (1997) in their landmark volume Historical Ecology in the Pacific Islands. It is noteworthy that many studies have used the term “landscape” to scrutinize long-term processes of diachronic interactions between natural processes and anthropogenic actions, probably because positioning the notion of landscape between the two enables the smooth combination of the natural sciences and humanities. Following these studies (see also Yamaguchi et al. 2009), I apply the term “landscape history” to a transdisciplinary approach for examining the diachronic processes that have formed specific present landscapes, and I review coral studies of the Ryukyu Island Arc, one of the well-studied fields in Japan, through a variety of natural sciences, making the work relevant to human settlement history. I also provide a short description of the preliminary outcomes of a geoarchaeological survey of Ishigaki Island in the Yaeyama Islands.
4.2 Long-Term Environmental History of Insularization of the Ryukyu Arc
Massive silty and sandy sediments of the Shimajiri Group, which appear to have been deposited since the late Miocene, extensively lie on various basement rocks. As there is substantially less earlier marine sediment dating back to the middle Miocene, the former Ryukyu Arc probably formed a part of the eastern margin of the Eurasian Continent (Kizaki 1980; Machida et al. 2001:301–2). New Quaternary nannofossil biostratigraphic data date the deposition of the marine siltstone to 1.552 ± 0.154 Ma, so the arc would have since been separated from the continent (Osozawa et al. 2012). Around 1 Ma, in the early Pleistocene, the succeeding marine environment surrounding the Ryukyu Arc produced a remarkable change in the sediments, from silty and sandy deposits to the carbonate rocks of the Ryukyu Group formed in the Ryukyu Coral Sea.
Two hypotheses have thus far been proposed to explain the main factors promoting the coral growth at this high latitude; one is the appearance of the Okinawa Trough and its trapping of muddy fluvial sediments from East China and Taiwan just before the Ryukyu Arc (Nakagawa 1983), and the other is the entrance of the Kuroshio (Black Current) to the Okinawa Trough and its transport of a tropical warm water mass to the Ryukyu Arc (Koba 1992). The Kuroshio could have entered the Okinawa Trough through the deepening and widening of the Yonaguni Depression that would have occurred as the result of the collision of the Luzon Volcanic Arc with Taiwan before 0.7–0.6 Ma (Koba 1992). This deepening and widening of the depression also separated the China–Okinawa land bridge and promoted insularity of the Ryukyu Arc.
Located at a plate boundary, the arc appears to have been subjected to highly complex vertical and horizontal movements that occurred along with glacioeustatic changes in sea level. Therefore, various paleogeographical hypotheses on land configuration changes in the Pleistocene have been proposed, and paleobiological evidence of fossilized terrestrial vertebrates and their associated, but less reliable, radiometric ages have also confused the issue. However, the recent advance of phylogeographical and molecular phylogenetic studies of terrestrial flightless vertebrates has provided a different set of convincing evidence on the historical process of insularization, or geographic isolation, of the islands (Hikida and Ota 1997; Ota 1998, 2003; Matsui et al. 2005). Although the details of this hypothesis are still somewhat inconsistent with newly examined Quaternary nannofossil biostratigraphic data from Shimajiri Group siltstone and overlying sediments (Osozawa et al. 2012), the Ryukyu Arc likely became an insular environment from the formation of the Ryukyu Group, beginning at latest by the early Pleistocene.
4.3 An Insular Environment with Coral Growth Even During the Last Glacial Maximum
Remarkable progress in calcareous nannofossil and sedimentological studies of the Ryukyu Group carbonates by the COral-REef Front (COREF) Project (Matsuda et al. 2012) has revealed that the group includes rhodolith and large foraminifera limestones in addition to reefal sediments deposited intermittently during the early and middle Pleistocene. The aggradational part indicates two long-term trends of Pleistocene sea-level change: the rhodolith and large foraminifera limestones would have formed on island shelf in deeper sea during periods of sea-level highstand, and the latter reefal sediments would have been deposited in shallow water during periods of sea-level lowstand under the condition of continuous tectonic subsidence (Ikeda et al. 1991; Obata and Tsuji 1992; Sagawa et al. 2001).
Obata and Tsuji (1992) analyzed high-resolution seismic reflection profiles and depositional facies distributed from the insular shelf to the shelf slope of the western offshore regions of Miyako and Irabu Islands and summarized the geohistory of the southern Ryukyu. They suggested that tectonic movement culminated with the upheaval of the Ryukyu Arc and that the long-term tectonic trend shifted at 0.4 Ma from subsidence to uplift. This shift also caused a part of the Ryukyu Limestone deposited between ca. 1.2 Ma and 0.4 Ma to become emergent on land.
Recent surveys of the insular shelf have revealed mound structures, characterized by strong reflections and chaotic internal structure in high-resolution seismic reflection profiles, in both central and southern Ryukyu. One of these, at the Amami Spur, is delineated by 80–120-m isobaths northeast of Amami-Oshima Island in the northern part of the central Ryukyu and had a measured thickness of 15 m and width of 400 m. Some irregularly shaped topographic highs were also detected at the shelf edge southwest of Kikai-shima at the northern edge of the central Ryukyu (Matsuda et al. 2011). These structures are considered biogenic reefs or banks consisting of coarse-grained bioclasts deposited during a stage of sea-level lowstand. High-resolution seismic reflection profiles of the insular shelf off of western Irabu Island also show the existence of mound structures located at present in water depth of ca. 126 m that were interpreted as reefal facies (Obata and Tsuji 1992). One of these was drilled, and its carbonate sediments were analyzed in detail; some coral fragments were dated by the 230Th/234U and 14C methods (Sasaki et al. 2006). This lithological examination and identification of coral fauna confirmed that the upper part of the mound structure is a small-scale coral reef, which was likely deposited until ca. 22 ka during the last glacial period.
4.4 Natural Environment of the Late Pleistocene Human Subsistence
Most intriguing from the perspective of landscape history is the possibility of coral reef formation during even the last glacial maximum (LGM) because the Ryukyu Island Arc had already been colonized by groups of Pleistocene humans in this period. Fossilized human remains have been excavated from calcareous fissure and cave sediments of limestone areas on Okinawa and Kume Islands in the central Ryukyu and also on Miyako and Ishigaki Islands in the southern Ryukyu. Many of the discovered Pleistocene human remains from Japan are concentrated within this arc (Matsu’ura 1996). So far, the earliest radiometric age for these Pleistocene human remains is ca. 32 ka from Yamashita-cho Cave 1 of Okinawa Island, where a juvenile femur and tibia were excavated in association with abundant extinct deer (Cervus astylodon). The layer including these remains was covered with a deposit containing dated charcoal flecks, probably from an earth oven; thus, the layer coincides with or is older than the conventional age of 32,100 ± 1000 BP (Takamiya et al. 1975).
Human skeletal remains of more than five individuals were excavated at a fissure site of the Minatogawa limestone quarry on Okinawa Island (Suzuki and Hanihara 1982). The lower deposits of this fissure also yielded fossils of extinct deer and less abundant wild boar (Sus leucomystax taivanus), whereas wild boar fossils were found without deer in the upper deposits. To confirm the vertical, or chronological, provenance of the human remains, a comparative analysis of human and terrestrial vertebrates was applied to their fluorine contents taken from post-burial surrounding sediments (Matsu’ura 1996). The result indicated that the human remains are close in relative age to the vertebrate bones from the lower deposits. Two conventional ages of 18,250 ± 650 and 16,600 ± 300 BP, respectively, were also obtained from charcoal flecks from the lower deposits (Suzuki and Hanihara 1982).
Most of the ages are less reliable ones from charcoal flecks collected from associated deposits of the fissure and cave sites, and the possibility remains that these could be contaminated by secondary deposits (Matsu’ura 1996). However, some exceptional human remains excavated with terrestrial vertebrates (including wild boar) at Shiraho-Saonetabaru Cave on Ishigaki Island, the southern Ryukyu, have recently provided more reliable conventional ages from well-preserved collagen samples extracted directly from the bones: 20,416 ± 113 and 18,752 ± 100 BP (Nakagawa et al. 2010). These ages are just within the LGM, when the lowstand sea level should have been at least 120 m lower than the present sea level. The islands at that time were larger and higher than the present ones, and some of them, such as Ishigaki and Iriomote, were likely connected into a “super island” (Ota 1998:189). A paleobotanical analysis of a boring core sample from Izena Island in the central Ryukyu indicated that the pollen assemblages around ca. 22 ka are dominated by coniferous trees, such as Pinus and Podocarpus with substantially less Abies and Tsuga, and deciduous broad-leaved trees of Alnus and Quercus subgen. Lepidobalanus. Thus, it has been assumed that the last glacial climatic condition in the subtropical Ryukyu Islands was not much cooler but was more arid than the climate of the present day (Kuroda and Ozawa 1996). This assemblage also included considerable pollen of evergreen broad-leaved trees, such as Quercus subgen. Cyclobalanopsis, Castanopsis-Pasania, and Myrica. This nuciferous vegetation would have supported terrestrial vertebrates such as deer and wild boar.
The late Pleistocene human groups in the Ryukyu Island Arc were likely dependent on nuts and the terrestrial vertebrates living within the evergreen broad-leaved forests. This assumption appears to be supported by the fossilized bones of deer and/or wild boar excavated along with human remains from fissure and cave sites located at the far interior of super islands, which would be within the habitats of these vertebrates. Nevertheless, recent geological investigations of some island shelves suggest that a shallow-water environment with scattered small coral reefs was also available to the Pleistocene human groups, even in the period of the LGM. This new possibility deserves to be examined archaeologically and anthropologically.
4.5 Coral Studies Linking the Late Pleistocene to the Holocene
During the post-LGM, the relative sea-level change has been dominated by a transition of the earth system from a glacial to an interglacial stage, and the large-scale ice-to-ocean mass redistribution has produced a dramatic increase in ocean water volume as well as a large and ongoing isostatic response of the solid Earth (Fleming et al. 1998; Milne et al. 2005). According to the sea-level history obtained by synthesizing three deglaciation curves based on long drill cores from Tahiti, Barbados, and New Guinea in the far-field region, where the glacial isostatic adjustment should be smaller in magnitude and the eustatic signal therefore dominant, there are two brief periods of accelerated ice melting, at ~13.8 and 11.3 ka, respectively, and these so-called meltwater pulse events are thought to have induced rapid sea-level rises that drowned the late Pleistocene coral reefs, which could not keep up with the rising water level (Camoin et al. 2005).
The present coral reefs surrounding the islands of the southern and central Ryukyu Arc started to grow some tens of meters below the present sea level since ca. 8.5–7.9 ka, but this seems to have occurred more than 1500 years later in the northern islands (Kan and Kawana 2006). The global melting rate of ice sheets has been significantly decreased from around 7 cal. kyear BP to the late Holocene (Milne et al. 2005); thus, the smoother sea-level rise should have been a suitable environment for coral reef growth. Additionally, the main axis of the Kuroshio, a warm current, is considered to have reentered the Okinawa Trough around 7.3 cal. kyear BP, when the sea surface temperature and depth of the thermocline abruptly increased to higher than modern values in both southern and northern troughs (Jian et al. 2000). The warm water mass transported from the tropics would also have increased the vertical growth rate of coral reefs in the Ryukyu Island Arc (Kan 2010:24).
Studies have illustrated the drastic changes in the insular environment from the late Pleistocene to the mid-Holocene. In particular, portable submersible drilling devices, newly developed in the 1970s (e.g., Macintyre 1975), enabled marine geologists to directly observe the interior coral reefs; description of their developmental processes was also facilitated by the prevalence of the radiocarbon dating method (Kan 2002; Hongo 2010). Consequently, it has been demonstrated that the Holocene coral reefs caught up with sea level during the mid-Holocene, which should be crucial to archaeological investigation of the Neolithic hunting-and-gathering subsistence economy and habitation in the insular environment of the Ryukyu Arc. For instance, as the coral reefs developed and their wave-dissipating effect increased, the Holocene high-energy window directly eroding island coasts would be closed, and the shallow lagoon environment could then form (Kan 2002). The prehistoric subsistence strategy should have been influenced by such a change in the coastal environment, resulting in a much greater dependence on marine resources available from the shallow water of the coral lagoon (e.g., Kinoshita 2012:17–18; Toizumi 2010:16).
However, considerable differences exist among the “catch-up” chronologies of the coral reefs, which range from 6.5 to 3.5 ka (Kan 2010). Even at the same island, the windward reef generally developed earlier than the leeward reef (Hongo and Kayanne 2008). This chronological diversity can be attributed to local effects of tectonic movement, wind direction, wave strength, and sea surface temperature. The spatial distribution of archaeological sites should thus be reexamined in detail with the local chronologies of Holocene coral reef development. Although Neolithic subsistence in the mid-Holocene in the Ryukyu Island Arc is generally viewed to have been largely dependent on the coral reef ecosystem, a recent case study of Ishigaki Island, in the southern Ryukyu, showed that the locations of habitation sites are not exclusively restricted within the spatial range of developed coral reef landforms in either the Shimotabaru (ca. 4850–3640 cal. BP) or Mudoki (ca. 1880–930 cal. BP) periods (Kobayashi et al. 2013)1.
The recent coral studies can provide a diachronic perspective on environmental history, linking the late Pleistocene to the Holocene. These have revealed drastic changes in the insular environment, which are crucial to and must be considered in archaeological investigations and interpretations, as mentioned previously. Similarly, the consequent relative sea-level fall in the late Holocene should have changed the littoral environment. The sea-level fall and seaward accretion of the reef should have resulted in reduced wave force at the shore. Such wave attenuation would bring about changes in the sedimentary environment and also promote sand spits or bars on emergent reefs. In the case of Tonaki Island in the central Ryukyu, a prehistoric habitation site dated at ca. 3500 BP and containing earthenware and shell middens (Toma 1981) is located on a tombolo, a sand bar connecting two islands (Kan et al. 1997).
The above archaeological site indicates that the geomorphic accretion of the littoral environment enlarged the spatial range of human habitation. However, human activities, even in prehistoric time, have another aspect: they act as an agent affecting the environment. Thus, we need to consider the anthropogenic changes in the landscape as well as the adaptive aspect of prehistoric subsistence. This requires the articulation of two directions of investigation in the diachronic perspective, the geomorphology of shallow water, where corals have been distributed, and the geoarchaeology of land, where islanders have lived. Our research of the Nagura region, Ishigaki Island, is still in progress, but a preliminary outline is presented below.
4.6 The Possibility of Anthropogenic Landscape Changes in the Late Holocene
Ishigaki Island, one of the southern Ryukyu Islands, lies about 440 km southwest of Okinawa Island and 240 km east of Taiwan. It has an area of approximately 223 km2, a long axis of 32 km extending from northeast to southwest, and a maximum width of 19 km at its southwestern part, where the Nagura region is located, facing the shallow Nagura Bay (Fig. 4.1b). This region is a large alluvial fan of the Nagura River, which is located between two chains of mountain peaks: the southern side of Banna-dake and the northern side of Omoto-dake, which is the highest peak at 526 m. The lower terrace of the nearly level or rolling alluvial fan consists of light-red to yellowish-red loamy matrix with lenses and beds of clay, sand, cobbles, and boulders (Foster 1965: Geologic Map of Ishigaki-Shima, Ryukyu-Retto). Ranging from 5 to 15 m above sea level, it slopes toward the bay; it is bounded by the alluvial lowland and is dissected by the main stream and several small valleys. In front of the alluvial lowland is an extensive mangrove forest of tideland, which is separated from the tidal flat of Nagura Bay by some spits covered with coastal vegetation.
We also observed that several colonies of dead Porites, which had diameters of several meters, were scattered at the sandy bottom of Nagura Bay at a depth of approximately 2 m. Two dated samples from these suggested that their growth was restrained at about 800–1200 cal. BP. Their decline could not be attributed to sea-level change but may be related to other environmental factors. Although it is necessary to examine such factors by analyses of trace elements in coral skeletal cores2, geoarchaeological information from the alluvial lowland will also be helpful for reconstructing the landscape history, including aspects of both land and shallow sea.
We collected boring cores on the Nagura coastal lowland and within the bottom of a small valley bisecting the lower terrace. The latter core, particularly, provided a stratigraphic sequence from shallow sea, through mangrove tideland to alluvial lowland, which subsequently has been altered into irrigated pond fields for rice cultivation. Floodplain deposits of sandy silt between sediments of tideland and pond fields accumulated to a thickness of more than 50 cm in the valley bottom around 2000–1000 cal. BP and contain a large number of Pundanus and Barringtonia pollen, which increase rapidly upward. Because its silica (SiO2) content is increased compared with lower sediments, the deposit can be viewed as having been derived from inland terrace. This phase corresponds to a period of prehistoric cultural, and some archaeological sites have been discovered within the Nagura region: the Kanda habitation site at the skirt of the lower terrace and shell midden spots of Nagura Site on an early sand spit. In the boring cores, microscopic charcoal flecks are more abundant within the sandy silt than in the lower sediments, which suggests the possibility that anthropogenic ignition of broad-leaved forest occurred synchronously with deposition of the floodplain deposit. Although detailed analyses are still in progress, it seems possible that colluvial soil washed from the inland not only contributed to the stabilization of an alluvial environment suitable for irrigation but also likely degraded the coastal shallow-sea environment where Porites and Favia colonies grow, as suggested by recent damage to coral growth by outflows of inland laterite soil.
4.7 Concluding Remarks and Introduction to the Next Two Chapters
Coral studies of the Ryukyu Island Arc have resulted in the accumulation of a vast knowledge quite relevant to a future reconstruction of the long-term landscape history, which can be articulated with anthropological and archaeological evidence on late Pleistocene and Holocene human activities in the insular environment. The preliminary articulation of several colonies of dead Porites with data from boring cores from a small valley suggests two aspects of anthropogenic work: environmental degradation and landscape enhancement. The first refers to the fact that inland activities degraded the shallow-water environment of coral colonies, even in prehistoric time, and the second focuses on the notion that these activities also promoted the development and stabilization of the alluvial environment in small valleys that were consequently altered into irrigated pond fields for rice cultivation. The detailed analysis of landscape history will reveal the difficulty of distinguishing between degradation and enhancement; thus, the essential aim should be to illustrate interactions between two agents, natural and human, through a cumulative and contingent history of the landscape (Kirch 2005:432) from the distant past to the present and toward the future.
This perspective of discerning landscapes as historical products is certainly not unique, as it was shared among the natural sciences and humanities as early as A.D. 1955. In that year, the Wenner-Gren Foundation for Anthropological Research held an international symposium on “Man’s Role in Changing the Face of the Earth” at Princeton University, which was a milestone event at which researchers from several academic domains such as geography, ecology, zoology, botany, history, anthropology, and planning came together to discuss human-induced environmental changes (Thomas 1956). Carl Sauer, a noted geographer, led the organization of this symposium to consider a tripartite division: retrospect, process, and prospect. This approach considers that retrospect and prospect are different ends of the same sequence and that today is therefore but a point on a line (Williams 1987:225). The approach of landscape history should also be open to discussions of future landscapes produced by on-going interactions between the nature and humans.
A recent work in planning has also appealed to the importance of orienting the present and future landscape through a diachronic perspective (Marcucci 2000). It allows us to summarize and complement the effectiveness of this perspective as follows: (1) understanding the long-term changes in landscape moderates the narrow essentialism of admiring specific cultural traditions, pristine landscapes, or harmonized ecosystems, which appear to stand on the edge of a precipice; (2) concurrence in the dynamic interactions between humans and nature is useful for overcoming the dichotomy between fundamental environmentalism and extreme development; (3) the diachronic historical perspective aids our realization that present planning becomes part of the historical process of a landscape once it is implemented; and (4) initiating historical inquiry at the very beginning provides a means for engaging the local inhabitants in landscape planning and also allows them to be experts on the history of their landscape.
The fourth point, particularly, would be a unique conception for planning, and it suggests that archaeologists can also contribute by weaving academic outcomes and the narratives of personal histories into a local landscape history, which would promote dialogue among inhabitants, including those who are not necessarily concerned with local environmental issues, in various situations. For such a contribution, we need to familiarize ourselves with historical–anthropological studies of the recent past, in which the present inhabitants have lived, and consider ways of making use of landscape histories in their present and future lives. The next two papers will include many suggestions on these topics. One of these papers is on the disappearance of historical ecological knowledge associated with destructive coral mining for traditional plaster in Ishigaki (Fukayama), and the other is on the practical articulation of local or personal knowledge for exploiting living resources and academic outcomes of landscape history in a modern enterprise such as ecotourism (Shimoda). The first suggests that traditions do not necessarily fit in with modern environmental conservationism but that they should rather be viewed as ecological and local knowledge about living resources. The second proposes a way of thinking, not about protecting corals and coral reefs, but about inventing value in them, both economic and cultural.
Conventionally, the prehistory of Ishigaki Island in the Holocene has been divided into two cultural phases in accordance with archaeological assemblages, along with a cultural blank between these phases: the Shimotabaru (the earlier phase) and the Mudoki (the later phase).
A coral skeletal core was bored in a colony of dead Porites and dated at ca. 1200 cal. BP, which has a time window of 10 years. The latest analysis of its Ba/Ca (a proxy of sedimentation and nutrients) shows no cyclical peaks suggesting seasonal activities, such as agricultural activity (Sowa et al. 2014).
- Foster HL (1965) Geology of Ishigaki-shima, Ryukyu-retto, Geological Survey professional paper 388-A. US Government Printing Office, Washington, DCGoogle Scholar
- Hikida T, Ota H (1997) Biogeography of reptiles in the subtropical East Asian islands. In: Proceeding of the symposium on the phylogeny, biogeography and conservation of Fauna and Flora of East Asian Region, National Science Council, R.O.C, pp 11–28Google Scholar
- Ikeda S, Kasuya M, Ikeya M (1991) ESR ages of middle Pleistocene corals from the Ryukyu islands. Quat Sci 36:61–71Google Scholar
- Thomas WL Jr (ed) (1956) Man’s role in changing the face of the earth: an international symposium under the co-chairmanship of Carl O. Sauer, Marston Bates, Lewis Mumford. Univ. of Chicago Press, ChicagoGoogle Scholar
- Kan H (2002) Development of coral reefs and the environmental change in coastal zone – sea level rise and revisiting of Holocene high energy window. Bull Soc Sea Water Sci Jpn (Nippon Kaisuigakkaishi) 56(2):123–127 (In Japanese)Google Scholar
- Kan H (2010) Holocene history coral reefs in the Ryukyu islands. Archaeol J (Koukogaku J) 597:24–26 (In Japanese)Google Scholar
- Kan H, Kawana T (2006) Catch-up of a high-latitude barrier reef by back-reef growth during post-glacial sea-level rise, Southern Ryukyus, Japan. In: Proceedings of the 10th international coral reef symposium, Japanese Coral Reef Society, Tokyo, 494–503Google Scholar
- Kawana T (1989) The Quaternary changes of earth’s crust in the Ryukyu islands. Chikyu Mon (Gekkan Chikyu) 11(10):618–630 (In Japanese)Google Scholar
- Kinoshita N (2012) Formation of culture and movement of people in prehistoric Ryukyu-with focus on the mutual human- geographical relation between islands. Bungakubu-ronso (Univ of Kumamoto) 103:13–27 (In Japanese)Google Scholar
- Kirch PV, Hunt TL (eds) (1997) Historical ecology in the Pacific islands. Yale University Press, New Haven and LondonGoogle Scholar
- Kizaki K (1980) Ryukyu-ko no Chishitsu-shi. Okinawa Times Press, Naha (In Japanese)Google Scholar
- Kobayashi R, Yamaguchi T, Yamano H (2013) Remote sensing evaluation of spatio-temporal variation in coral-reef formation in Ishigaki, Yaeyama islands, toward a study of prehistoric insular resource utilization. Q Archaeol Stud 60(2):55–72 (In Japanese)Google Scholar
- Kuroda T, Ozawa T (1996) Studies of land bridges and the migration of men and other animals along them – paleoclimatic and vegetational changes during the Pleistocene and Holocene in the Ryukyu islands inferred from pollen assemblages. J Geogr (Chigaku Zasshi) 105(3):328–342 (In Japanese)CrossRefGoogle Scholar
- Machida H et al (eds) (2001) Nippon no Chikei 7: Kyushu and Nansei-shoto. University of Tokyo Press, Tokyo (In Japanese)Google Scholar
- Macintyre IG (1975) A diver-operated hydraulic drill for coring submerged substrates. Atoll Res Bull 185:21–26Google Scholar
- Matsuura S (1996) A chronological review of Pleistocene human remains from the Japanese Archipelago. In Omoto K (ed) Interdisciplinary perspectives on the origins of the Japanese, International Research Center for Japanese Studies, Kyoto, 181–197Google Scholar
- Matsuda H et al (2012) Paleoceanography reconstructed from shallow-water carbonates during glacial periods in the northern part of the Ryukyus. Chikyu Mon (Gekkan Chikyu) 34(6):363–372 (In Japanese)Google Scholar
- Nakagawa H (1983) Outline of Cenozoic history of the Ryukyu islands. Geol Soc Jpn (Chishitsugaku-ronsyu) 22:67–79 (In Japanese)Google Scholar
- Nakagawa R, Doi N, Nishioka Y, Nunami S, Yamaguchi H, Fujita M, Yamazaki S, Yamamoto M, Katagiri C, Mukai H, Matsuzaki H, Gakuhari T, Takigami M, Yoneda M (2010) Pleistocene human remains from Shiraho-Saonetabaru Cave on Ishigaki island, Okinawa, Japan, and their radiocarbon dating. Anthropol Sci J Anthropol Soc Nippon 118(3):173–183Google Scholar
- Ota H (2003) Toward a synthesis of paleontological and neontological information on the terrestrial vertebrates of the Ryukyu Archipelago, I: systematic and biogeographic review. J Fossil Res (Kaseki Kenkyu-kai Kaishi) 36(2):43–59Google Scholar
- Suzuki H, Hanihara K (eds) (1982) The Minatogawa man: the upper Pleistocene man from the island of Okinawa, Bulletin, no. 19. University Museum, University of Tokyo, Tokyo, JapanGoogle Scholar
- Takamiya H, Kin M, Suzuki M (1975) Excavation report of the Yamashita-cho Cave Site, Naha-shi, Okinawa. J Anthropol Soc Jpn (Jinruigaku Zasshi) 83(2):125–130 (In Japanese)Google Scholar
- Toizumi T (2010) Vertebrate resource use in prehistoric Amami and Okinawa islands. Archaeol J (Koukogaku J) 597:15–17 (In Japanese)Google Scholar
- Toma S (1981) Tonaki-jima no Iseki. Reports of Okinawa Prefectural Museum II (Okinawa Kenritsu Hakubutsukan Sogo Chosa Hokokusho II), Tonaki Island, pp 41–49 (In Japanese)Google Scholar