Prehistoric Migration of Homo sapiens in the Ryukyu Islands

Abstract

Archeological sites dating back 30,000 years have been discovered throughout the Ryukyus and provided evidence for the migration of Homo sapiens across the sea during the late Pleistocene. A possible relationship to Southeast Asia is suggested, both from studies of artifacts and human remains. Subsequently, prehistoric people continued to inhabit at least several islands until the Holocene. From the end of the Pleistocene to the Holocene, the influence of Jomon culture extended to the Ryukyus, and people of the Jomon culture probably migrated from mainland Japan. One of the oldest evidence of pottery manufacture possibly related to incipient Jomon is found on Tokuno-Shima Island in central Ryukyu. This paper reviews the history of human migration to the Ryukyu Islands and discusses changes in their environment and culture.

1. Introduction

The first migration across the ocean to the islands was archived in the islands of the western Pacific by Homo sapiens during their worldwide dispersal [1, 2, 3, 4]. The Japanese Archipelago is an area where H. sapiens migrated across the ocean during the Upper Paleolithic period or the terminal Pleistocene. Since 38,000 years ago, the number of Paleolithic sites has increased in mainland Japan [5]. The similarity of Paleolithic culture with neighboring regions has prompted scholars to discuss cultural inflow via the Korean Peninsula (38,000 years ago) and Siberia/Hokkaido (25,000 years ago) to mainland Japan [5, 6]. However, the extent of migration that occurred along each route remains unclear. Although no human bones have been discovered at these sites, except for the partial H. sapiens fossils such as fragments of cranium, humerus, and pelvis from the Hamakita Negata Site in mainland Japan [7, 8], almost all Paleolithic sites in Japan are considered to be associated with H. sapiens based on the cultural point of view. The lower sea level during the Last Glacial Maximum (LGM) means that several islands were connected to each other; however, Paleo-Honshu island, which was formed by connecting Honshu, Shikoku, and Kyusyu, was always separated from the continent by the ocean. The Paleolithic people in the Japanese Archipelago, having come across the ocean, must have had nautical voyaging skills. Unfortunately, there exist no archeological remains of the used watercraft. However, raw materials of obsidian taken from Kozu Island at many Paleolithic sites in Honshu indicate that Paleolithic people were able to sail back and forth over distances of approximately 80 km in search of obsidian from Honshu to Kozu Island [9].

H. sapiens also inhabited the Ryukyu Islands, located in the southwestern part of the Japanese Archipelago, during the Upper Paleolithic period. Ryukyu comprises more than 200 relatively small, isolated islands, extending approximately 1200 km between Kyushu and Taiwan (Figure 1). Okinawa Island, which is the largest in the Ryukyus, covers an area of 1208 km². The distance between the islands is the largest between Miyako Island and Okinawa Island, at approximately 240 km or more. Kinoshita [10] examined interisland visibility and found that it was impossible to distinguish one island from the other in at least three places. However, archeological sites where lithic tools and human bones have been found in 35,000–30,000-year-old sediments are distributed across the relatively large islands of the Ryukyus (Table 1) [1, 4]. Paleolithic people in the Ryukyus must have achieved long-distance voyages across the Kuroshio Ocean Current, which is one of the world’s fastest ocean currents. Recent simulation studies have pointed out that the Kuroshio Current during the terminal Pleistocene may have been almost the same as it is today, or even stronger [27].

Figure 1.

Migration routes suggested by the human remains and/or genetic studies (blue arrows) and archeological remains (orange arrows). Note: Probability is high for solid line and low for dotted line.

Area	Island	Site	Chronological Age	Human Remains	Pleistocene Artifacts and Feartures	Dated Materials	References
Northern Ryukyu	Tanegashima	Yokomine C	35 k–30 k cal yBP	none	Pebble tools, mealing stone, pebble cluster	older than TaneIV tephra (35 ka)	[11]
	Tanegashima	Tachikiri	35 k–30 k cal yBP	none	pebble tool, edge-ground adze, burnt pebble clusters hunting pitfalls	older than TaneIV tephra (35 ka)	[12]
Central Ryukyu	Amami	Tsuchihama Yaya	30 k cal yBP	none	flakes with polished surface, shale flakes	around AT tephra (30 ka)	[13]
	Amami	Kishikawa	30 k cal yBP	none	shale flakes	around AT tephra (30 ka)	[14, 15]
	Tokunoshima	Garazo	earlier than 30 k cal yBP	none	grinding stone (granite and sandstone)	oleder than AT tephra (30 ka)	[16]
	Okinawa	Yamashita-cho Cave I	36.5 k cal yBP	infant right femur and tibia	hammer/grinding stones	charcoal	[17, 18]
	Okinawa	Sakitari Cave	layer I: 13 k–16 k cal yBP; layer II: 20 k–23 k cal yBP; layer III: 23.5 k–35 k cal yBP	a deciduous canine; and a carpal	quartz flakes (I); shell beads (I + II); shell scrapers (II + III); whet stone (II); shell fishhooks (II)	charcoal	[19, 20, 21]
	Okinawa	Minatogawa	22 k–20 k cal yBP	four adult individuals	none	charcoal	[22, 23]
	Kume	Shimojibaru Cave	18 k cal yBP	an infant individual	none	charcoal	[24]
Southern Ryukyu	Miyako	Pinza-abu	29 k cal yBP	partial adult individual(s)	none	charcoal	[25]
	Ishigaki	Shiraho-Saonetabaru Cave	layer IIIC: ~19 k cal yBP; layer IV: ~27.5 k cal yBP	partial adults at least 19 individuals (including four well-preserved skulls)	none	human and animal bone coragen	[26]

Table 1.

Major Late Pleistocene Sites in the Ryukyu Islands.

2. Late Pleistocene migration to the Ryukyus viewed from the culture and human remains

The earliest site on the Ryukyus is Yamashita-cho Cave I on Okinawa Island. Partial infant leg bones were found in 36,500-year-old sediments, accompanied by stone cobbles and possible food remains such as fish bones and marine shells, although they were not keenly studied [17, 18, 22]. About 30,000-year-old flakes and cobbles were found on several other islands throughout the Ryukyus (Table 1) [1, 4]. The cultures of these earliest settlers differed among the Northern, Central, and Southern Ryukyu regions. On Tanegashima in Northern Ryukyu, a variety of stone tools such as edge-ground axes, flake tools, trapezoids, grinding stones, and mealing stones were used during the Upper Paleolithic period, and burnt pebble clusters with charcoal and pitfalls for hunting traps have been detected [11, 12]. The microblades found at terminal Paleolithic sites in Tanegashima were similar to those found in Kyusyu, mainland Japan, and many other similarities were observed in the Paleolithic cultures of Tanegashima and Kyushu. One reason for this similarity is the land bridge between Tanegashima and Kyushu due to the sea level drop during the LGM period, but almost no obvious transport of stone materials from Kyushu has been recognized. In contrast, various stone materials were brought to Tanegashima from Kyushu after the Incipient Jomon period.

Paleolithic flakes and pebble tools have been reported on several islands in Central Ryukyu (Table 1). On Amami Oshima, shale flakes and flakes with polished surface (possibly fragments of an edge-ground adze) were discovered in the sediments and correlated with the Aira-Tanzawa (AT) tephra, which is approximately 30,000 years old, at the Tsuchihama-Yaya Site (Figures 2 and 3) [13]. At the Kishikawa site, chert flakes and pebble clusters were found in the sediments around AT tephra [14, 15]. On Tokunoshima, two grinding stones were unearthed below the AT tephra at Garazo Site [16]. Retouched chert flakes with secondary modification, which may be related to the Paleolithic trapezoidal artifacts of Kyusyu, were reported at the Amangusuku site (Figure 4), although the stratification of the Amangusuku flakes was not keenly studied [28]. Kato [29] pointed out a possible relationship between these flake industries (“amorphous flake industries” named by Kato) in Central Ryukyu and the Paleolithic culture in Southeast Asia. However, if the flake industry in Tokunoshima is linked to the trapezoid industry in mainland Japan, we must consider the cultural relationship between Central Ryukyu and mainland Japan [30].

Figure 2.

Shale stone tools obtained from the Paleolithic layer (IIIC) of Tsuchihama Yaya site (stored in Kagoshima prefectural archaeology Center).

Figure 3.

Flakes with polished surfaces (possibly fragments of an edge-ground adze) found at Paleolithic layer (IIIC) of Tsuchihama Yaya site. The length of the rightmost fragment is about 2 cm (stored in Kagoshima prefectural archaeology Center).

Figure 4.

Retouched chert flake tools of Amangusuku site (stored in Isen-cho Board of education).

In the Okinawa Islands, three sandstone cobbles were discovered with partial infant human bones at the Yamashita-cho Cave I site, which is 36,500 years old, based on ¹⁴C dating of charcoal [17, 18, 27]. These cobbles were reported to be throwing stone balls [18], but Oda [31] claims they were grinding stones or hammer stones that might share a cultural relationship with those found at Garazo in Tokunoshima and several sites in Tanegashima. Following these cobbles, a culture unique to the Ryukyus possibly developed on Okinawa Island. A variety of shell artifacts dating back 23,000 to 13,000 years ago were discovered in Sakitari Cave, Okinawa Island (Figure 5) [19, 20, 21, 32]. At this site, well-preserved strata dating back to 35,000 years ago have been excavated, and a variety of shell artifacts, including several types of shell scrapers, shell fishhooks, shell beads, and sandstone, which were probably used to grind fishhooks, were discovered in the sediments of 23,000–20,000 years ago [19, 20, 21, 32]. Pigment staining was observed in one tusk-shell bead [32]. From the 16,000 to 13,000-year-old sediments, quartz flakes, snail shell beads, a “crayon” (pigment processed into a rod shape), and a small abalone shell were unearthed [20, 21]. The assemblage mainly consists of a variety of shell tools and almost lacks stone tools; it can be considered a unique culture adapted to islands with poor lithic materials and rich shell resources. The combination of bone and shell tools with “simple” stone tools is also known in the Philippines during terminal Pleistocene to early Holocene [33, 34]. The late Pleistocene H. sapiens who expanded into the western Pacific islands may have widely shared techniques for processing bones and shells. These technological innovations enabled them to adapt and expand into the island environments with limited stone resources such as the Ryukyus.

Figure 5.

Major artifacts and human remains from Pleistocene layers of Sakitari cave. Materials from 20,000 to 23,000-year-old layer are shown above the dotted line, and the 13,000 to 16,000-year-old materials are below the dotted line. Human teeth (a, q), shell fishhook (b), shell scrapers (c-k), shell beads (l-p, u, v), and quartz flakes (r-t).

In the terminal Pleistocene, the Hoa Binh culture, which mainly consisted of cobble tools and edge-ground tools and adzes, developed in continental Southeast Asia, whereas assemblages in Islands of Southeast Asia consisted of siliceous flake tools and small amounts of shell artifacts. Both methods consist of processing tools that have been discussed in relation to plant processing [35]. In contrast, Paleolithic cultures in Northeast Asia, including mainland Japan, are characterized by blade production and a variety of hunting tools, such as points, knives, and blades. In this context, the Sakitari assemblage resembles Southeast Asia in terms of the dominance of processing tools but has a major difference in poor stone tools in Sakitari. Moreover, fishhooks made from Trochidae shells have been reported from Pleistocene to early Holocene sites in the Southeast Asian islands [3, 36, 37]. The oldest fishhook in this area was found at the Jerimalai site (Asitau Kuru) in East Timor [36], which dates back 23,000 to 16,000 years ago and is as old as the Sakitari site. Symbolic items, such as pigments and ornaments, appear to be similar across a wide range of regions, and it is difficult to clarify their origin(s), while practical items usually vary by region in their morphology, material, etc. From this perspective, fishhooks are not only subsistence tools, but also symbolic items, and the possibility of multiple origins of shell fishhooks from Southeast Asia and Okinawa has been discussed [37]. However, the fishhooks were made of similar materials and had similar shapes in these regions. We should not deny the single-origin hypothesis for shell fishhooks. The shell fishhook in Okinawa may provide evidence of cultural connections between Southeast Asia and the Ryukyu Islands. In contrast, there are no fishing hooks in Philippines, but fishing gorge have been reported dating back 30,000 years ago [34, 38]. Considering this, late Pleistocene people in the western Pacific islands may not have shared shell fishhooks themselves, but the activities such as processing bones and shells and utilizing maritime resources.

In Southern Ryukyu, late Pleistocene human fossils were discovered at the Pinza-abu Cave on Miyako Island [25] and at the Shiraho-Saonetabaru Cave on Ishigaki Island [26], but there are no associated cultural remains. The oldest cultural materials in the Southern Ryukyus were found at early Holocene sites, but their cultural context has not been determined. We will discuss this later from the viewpoint of Holocene migration.

Pleistocene human fossils in the Ryukyus also provide information about their origin(s). Based on a qualitative comparison of craniums, Minatogawa I was more similar to Wadjak in Indonesia than to the Upper Cave and Liujiang in China [39, 40, 41]. Recent quantitative studies of Minatogawa mandibles have suggested the morphological affinity to the Australo-Melanesian [42]. The Shiraho 4 skull resembles prehistoric Southeast Asia, although it also resembles mainland Jomon and Mintogawa [43]. From a genetic perspective, the mitochondrial haplogroup B4e detected in the Shiraho specimen has a distribution center in Southeast Asia [44]. These studies suggest Paleolithic migration from Southeast Asia to South and Central Ryukyu. This finding is consistent with the cultural relationship between Southeast Asia and Central Ryukyu.

Moreover, Minatogawa I in Okinawa and Shiraho 4 in Ishigaki have similarities in short facial height and short cranium length, but they have different characteristics, such as the width of the frontal region and protrusion of the temporal region [43]. They also differ in genetics; Minatogawa I was haplogroup M [45], whereas Shiraho specimens were determined as haplogroups B4e and R [44]. If these studies indicate different origins of the Paleolithic people, we need to consider multiple migrations from different areas of Southeast Asia to the Ryukyus. If some of the Paleolithic tools on Amami Island are related to those in Kyusyu, migration from the north cannot be denied, or we should consider the possibility that there was an exchange system with the north.

3. Intentionality of the Paleolithic voyage to the Ryukyus

As Paleolithic voyages have attracted the attention of researchers, their intentionality has become another important issue. Especially in the Ryukyus, the islands are relatively small and isolated, and the world’s fastest ocean current, the Kuroshio, makes it difficult for people to cross the ocean [1]. However, human remains found at Paleolithic sites of the Ryukyus contain both male and female adult and infant bones. Kaifu [4] states that Paleolithic migration(s) to the Ryukyus were not accidental drifting, which is thought to have been carried out mainly by men but were deliberate voyages involving both men and women. Ihara et al. [46] conducted a simulation study that showed that immigration should include 10 or more unrelated young males and females to maintain the population on the island. It is difficult to imagine a large group drifting simultaneously, but the multiple drifts within a relatively short period of time can have the same effect as a large number of people drifting at the same time. Another reason to deny the drift hypothesis is that the floating buoys that started in Taiwan almost never reached the Ryukyus, and the few buoys that arrived at the Ryukyus were adrift for more than 21 days [47].

In contrast, Oda and Ono [48] compiled cases of modern drifting in Oceania and mentioned that accidental drifting should not be denied because there are many cases of surviving a drifting period of more than 21 days. Although rafts are less maneuverable than dugout canoes, they can carry more people and daily supplies, allowing them to survive at sea for a longer period and increasing the likelihood of successful reproduction after reaching the island [48]. Further, the motivation for a deliberate voyage to an invisible island could not be explained [41].

Regarding the visibility of islands, some migratory birds, such as gray-faced buzzards, migrate along the Ryukyus from the mainland Japan to Taiwan, Philippines, and the Southeast Asian countries every autumn [49]. During the migration season, thousands of gray-faced buzzards migrate simultaneously under favorable weather and wind conditions. It would have been possible for Paleolithic people to recognize invisible islands from the presence of such migratory birds. Additionally, hunters sometimes track their prey over a long distance, and it is possible that Paleolithic people discovered the island after tracking their prey at sea. Considering the dispersal speed of terrestrial reptiles by drifting, it may have taken 100,000 to 1,000,000 years to disperse throughout the Ryukyus [50, 51, 52]. Compared to the ocean current dispersal of these terrestrial vertebrates, the speed of H. sapiens migration, which spread throughout the Ryukyus between 35,000 and 30,000 years ago, is remarkable. If H. sapiens adapted to the ocean in Southeast Asia around 50,000 years ago and actively sailed out for fishing, accidental drifting might have increased, and some H. sapiens could reach the Ryukyus. However, given the dynamics of the migrations that took place during the global expansion of H. sapiens, it is probable that people who were strongly motivated to explore new horizons migrated intentionally to the Ryukyus. Considering the possibility of multiple migrations from the north and south to the Ryukyus, we may need to consider both intentional voyages and accidental drifting.

4. Relation between the Paleolithic and Jomon populations in mainland Japan and in Okinawa

The Paleolithic and Jomon periods in Japan differ significantly from a cultural perspective, such as the appearance of pottery, the formation of shell middens, and primitive plant cultivation. With regard to the primitive agriculture, chestnuts (Castanea crenata), Japanese walnuts (Juglans mandshurica), Soybean (Glycine max) and azuki bean (Vigna angularis) found at the middle Jomon sites are larger than wild species, and the artificial selection or domestication of these plants may have occurred during the early to middle Jomon period [53, 54]. The earliest pottery in Japan dates back to 15,000 years. However, the formation of shell middens and primitive cultivation began in various regions of mainland Japan after the early Holocene [55]. The cultural transition from the Paleolithic to the Jomon was drastical but it is difficult to imagine that the indigenous human group has been completely replaced [55, 56]. Unfortunately, we do not have any Paleolithic human remains from mainland Japan, except at the Hamakita Negata site. Although the limb bone fragments of the Hamakita specimens showed morphological similarity with those of the Jomon people [7], the evidence is weak for discussing the relationship between the Paleolithic and Jomon people because they are too fragmentary to draw a conclusion.

According to ancient DNA studies, the divergence of Jomon from the surrounding population occurred after the appearance of Tian-Yuan, which is the approximately 40,000-year-old human remains found in China and shared the proportions of DNA variants derived from archaic humans similar to present-day people in mainland Asia [57, 58], and before the formation of the Northeast Asian, Siberian, and North American populations approximately 24,000 years ago [59, 60]. This period is consistent with the timing of Paleolithic migration to the Japanese Archipelago, considered to be from lithic culture. This is also consistent with the fact that a certain degree of cultural continuity was observed from the Paleolithic to the Jomon period, and it is difficult to recognize any major discontinuities between them. Today, the adjacent population, which is closely related to the Jomon people, is found in the coastal areas of East Asia, such as Taiwan, Korea, and Ulchi, rather than in Southeast Asia [59]. This suggests that the mainland Jomon might have a different origin from the Okinawan Paleolithic people, who have a possible link to Southeast Asia.

Previous studies have considered Okinawan Paleolithic Minatogawa as the ancestor of the Jomon people [23, 39, 40, 41]. However, recent morphological analyses have identified many differences between Minatogawa and modern Japanese and/or Jomon in the glabellar region [61], cranium shape [62], tooth root size [63], and mandibular shape [42]. From a genetic perspective, the mitochondrial DNA (mtDNA) of Minatogawa I has been identified as haplogroup M [45]. It is one of the ancestral haplogroups of M7a and N9, which are popular among the mainland Jomon people, although haplogroup M is not a direct ancestor of them [45, 64]. In contrast, morphological analysis of the skull of Shiraho 4 shows similarities to prehistoric Southeast Asia and the Jomon people [43]. The mtDNA haplogroup B4e detected in the Shiraho-Saonetabaru specimen [44] has been recognized to a small extent in modern Okinawans and Japanese people. Therefore, we need to examine the relationship between the Paleolithic and Jomon people by island or site.

In contrast, Sato et al. [65] conducted genetic research on the modern Okinawan population and found that it diverged from the continental population (Han Chinese) 15,000 years ago. During this period, Jomon pottery began to appear in mainland Japan and spread to the Ryukyus, as explained in the following section. However, if the divergence 15,000 years ago indicates the divergence of the mainland Jomon people from the continental population, this contradicts the hypothesis that the Jomon people were descendants of Paleolithic people. This may be related to the assumption of single divergence from the continent, as proposed by Sato et al. [65].

5. Influences of the Jomon culture on northern and central Ryukyu

Soon after Jomon pottery appeared in mainland Japan, pottery culture moved south to Tanegashima Island in Northern Ryukyu, and the culture of Tanegashima has become almost the same as that of Kyushu. For example, the number of relatively large-scale settlements with many artifacts has increased, and stone tools and raw materials imported from Kyushu have become common.

Recently, several potsherds discovered in Shitabaru Cave on Tokunoshima Island in Central Ryukyu have attracted attention because they may be a Ryukisen-mon pottery (linear-applique pottery) dating back 16,000–15,000 years ago [66], suggesting that Jomon culture may have reached Central Ryukyu in the incipient Jomon period. Another possible evidence of the connection between Central Ryukyu and the incipient Jomon is the polished stone adze collected at Kayauchibanta, which is located at the northern end of Okinawa Island. It has a similar form to the cylindrical round-edged chisel-shaped adzes named “Kakoinohara-type” that was unique to the incipient Jomon culture of southern Kyushu [67, 68], but it is difficult to evaluate this Okinawan adze because it lacks stratigraphic information. At present, the culture of the incipient Jomon possibly reached Central Ryukyu; however, more information is needed to confirm this.

The clear influence of the Jomon culture appeared in Central Ryukyu approximately 10,000 years ago. Striation pattern pottery, similar to the shell-marked cylindrical pottery of initial Jomon pottery in southern Kyushu, has been discovered at seven archeological sites, including Teera-gama Cave on Okinawa Island and Yabuchi Cave on Yabuchi Island, which is nearby Okinawa Island (Figure 6) [68]. With the introduction of pottery, the combined use of terrestrial and marine resources, such as wild boar, blood cockles, oysters, and mangrove clams, becomes common, as seen in the Yabuchi Cave. The shellfish in Yabuchi consisted of bivalves living in the sandy mud bottom of the inner bay, similar to the shellfish consumed by the Jomon people in mainland Japan. As coral reefs did not develop in the Ryukyu Islands until 9500–8500 years ago [69], the use of shellfish in the inner bay’s sandy muddy bottoms can be explained by these environmental factors. We cannot deny that the food culture favoring these shellfish was introduced along with pottery culture. The pottery of Shitabaru Cave on Tokunoshima has a wavy striated design and resembles the pottery of Teera Gama Cave and Yabuchi Cave. It is accompanied by wild boars, endemic rabbits, marine shells, and lots of freshwater snails [66]. At the Hango site on Amami Oshima, thin pottery with a nail-tip impression, which is different from these potteries and resembles the following pottery with a shouldered linear impression pattern, was discovered together with the nuts of Castanopsis from the 11,000-years-old sediments [70]. Although there are similarities in the pottery on these islands, there are different aspects of food culture.

Figure 6.

Early Holocene potteries of central Ryukyu (Okinawa and Amami Islands). 1: Striation pattern pottery (replica, Minatogawa site, Okinawa Island), 2: Wavy pattern pottery (Umbuki underwater cave, Tokunoshima Island, found and collected by underwater explorer Toshiaki Hirobe), 3: Shouldered linear impression pottery (Kikai-cho Sogo ground site, Kikai Island), 4: Nail-tip impression pottery (Ireibaru site, Okinawa Island) 5: Sobata-type pottery (Isso-Matsuyama site, Yaku Island), and 6: Sobata-type pottery (Keji I site, Amami Oshima Island) potteries are stored in Okinawa prefectural museum and art Museum (1), Amagi-cho Board of education (2), Kikai-cho Board of education (3), Chatan-cho Board of education (4), and Kagoshima prefectural archaeology Center (5, 6).

After this pottery, earthenware with a shouldered linear impression pattern appeared approximately 9000 years ago in six archeological sites on Amami, Kikai, Tokunoshima, Okinoerabu, and Okinawa Islands of Central Ryukyu [71]. This pottery is considered the prototype of nail-tip impression pottery, which was unique to Central Ryukyu 7000 years ago. The unique culture of Central Ryukyu developed not only in pottery design but also in food culture; wild boar, shellfish living in coral reefs, sea turtles, and dugongs became common. Therefore, the interaction between Kyushu and the Ryukyus decreased 9000 years ago, and a culture adaptive to the subtropical islands developed in Central Ryukyu.

Between 6500 and 5500 years ago, Sobata-type pottery, which was popular in the early Jomon of Kyushu and Northern Ryukyu, came south of Central Ryukyu [71]. In Sobata culture, many cultural elements were introduced to Central Ryukyu: fully polished stone adzes, flake industry of siliceous chert, the use of red ocher, basketry of plant material, and breeding dogs. After the Sobata period, the pottery culture of Central Ryukyu come to take on remarkable regional characteristics against Kyushu and Northern Ryukyu.

Thus, the influence of mainland culture on Central Ryukyu was not continuous but was limited to the Pleistocene, early Holocene (10,000–9000 years ago), and early Jomon period (6500–5500 years ago). We do not know why cultural exchange across the ocean became active during this period. No obvious development of maritime technology and/or maritime adaptation was observed during these periods, at least in material culture. However, the number of archeological sites increased during the incipient and initial Jomon periods in southern Kyusyu and Tanegashima; therefore, population growth was one of a possible factor in migration to the Ryukyus. The number of archeological sites decreased in southern Kyusyu after the catastrophic eruption of the Kikai Caldera approximately 7300 years ago, and the cultural influence of Kyusyu on the Ryukyus decreased. Archeological sites increased again during the Sobata Culture (6500–5500 years ago), and the cultural influence on the Ryukyus became noticeable. It is not clear how large the population influx was during these periods; however, given the genetic similarities between the Jomon populations of the Ryukyus and Kyushu, some degree of migration is assumed.

6. Holocene cultures in southern Ryukyu

No prehistoric human occupation before 3000 years ago has been reported from Miyako Island. There is a long hiatus in human history after the 30,000-year-old partial human bones from the Pinza-abu Cave. Recently, stone tools, isolated human teeth, and animal remains, including fish bones, were found in 10,000-year-old sediments at Tsuzupisuki-abu Cave [72]. However, the cultural context in which this site should be located remains unclear. At Tsuzupisuki-abu Cave in Miyako, chert flakes and pebbles, possible pigments (limonite), shark teeth, fish bones, wild boar remains, and isolated human teeth were found in 10,000-years-old layer. At Shiraho-Saonetabaru Cave in Ishigaki, quartz flakes, cobble tools, and pottery fragments were found in 10,000-year-old sediments associated with wild boar bones, some of which had clear cut marks.

On Ishigaki Island, possible human occupation between 27,000 and 13,000 years ago has been suggested based on the radiocarbon dates of human bones, animal bones, and charcoal obtained from the Shiraho-Saonetabaru Cave, but no cultural remains have been discovered during this period. The earliest cultural remains in Ishigaki were stone cobbles, quartz flakes, and tiny fragments of pottery unearthed from the 10,000-year-old layer of Shiraho-Saonetabaru Cave. It is also unclear in what cultural context these artifacts should be placed. Their clay and mixed materials seem different from those of early Holocene potteries of Central Ryukyu [68]. The chemical composition of this earthenware clay differs from that of following Shimotabaru pottery, although it is similar to the part of the oldest Shimotabaru pottery [73].

After this early Holocene, there was a period of interruption in Ishigaki’s human history. The second oldest culture appeared in the Shimotabaru period about 4000 years ago. People of the Shimotabaru culture were hunter-gatherers who built shell middens, kept dogs, and hunted wild boars. The pottery of the Shimotabaru culture is relatively small, thick-walled, and with a simple design made of coarse clay, and is quite different from the pottery in Central Ryukyu [74]. Therefore, some researchers consider that Shimotabaru culture is different from mainland Jomon and possibly related to the cultures of southern areas, such as Taiwan [75]. However, there are many differences between Shimotabaru and the Neolithic culture of Taiwan, which was influenced by mainland China; they have an abundance of sophisticated earthenware, various implements made from jade (nephrite), domestic dogs and pigs, and agriculture also began [75]. The cultural background of Shimotabaru culture is still shrouded in mystery.

On Miyako Island and Yaeyama Islands, the (aceramic) Mudoki period is recognized following the Shimotabaru period. Mudoki culture is characterized by earth ovens (burnt pebble clusters) and shell adzes made of Tridacna clam shell; the main subsistence is based on the foraging economy such as shellfish gathering within coral reefs and wild boar hunting [10, 75]. The cultural relationship between the Mudoki and Shimotabaru is unknown [76]. They are thought to have almost no relation to the cultures in the north of Central Ryukyu, which are associated with Jomon and Yayoi cultures of mainland Japan, although parts of the shell items, wild boar tusk artifacts, and perforated shark teeth show similarities between Shimotabaru, Mudoki, and Okinawan Jomon.

In contrast, the mitochondrial DNA haplotype of the bones found in the Shimotabaru and Mudoki periods at the Shiraho-Saonetabaru Sites on Ishigaki was M7a [44], which is widely recognized among the Jomon people of Okinawa and mainland Japan [64]. Human remains from the Nagabaka Site on Miyako—which were not accompanied by any cultural materials but ¹⁴C- dated to the Shimotabaru period—were genetically similar to those in mainland Jomon [77]. These genetic studies suggest a possible link between the Shimotabaru, Mudoki, and Jomon populations. Therefore, it would not be surprising if there was a cultural relationship as a result of human migration. However, only a small number of human bones have been genetically analyzed. In addition, a major problem in Southern Ryukyu is the long absence of archeological sites in the middle Holocene.

7. Discussion

During the late Pleistocene, H. sapiens arrived in the Ryukyu Islands approximately 37,000 years ago, and since then, have repeatedly interacted with the surrounding areas across the sea. During the first stage of migration, relatively large islands (Tanegashima, Amami Oshima, Tokunoshima, Okinawa Island, Miyako Island, Ishigaki Island) were occupied. The culture of Tanegashima in Northern Ryukyu has been correlated with that of Kyushu since the Paleolithic period, while Central and Southern Ryukyu are probably related to Southeast Asia from the perspectives of fossil human morphology, ancient genetics, and artifacts. However, there is also an opinion that the Paleolithic culture of Amami Island was partly influenced by that of Kyushu; therefore, influence from the north cannot be ruled out at present despite insufficient evidence.

Human migration and maritime adaptation can be traced back to 50,000–40,000 years ago on the Southeast Asian islands. It is plausible that some of these people became more active in coastal areas and migrated to the Ryukyus. However, mainland Japan was itself separated from the continent by the ocean, and the Paleolithic people of mainland Japan must have possessed an advanced maritime technology, as seen in the transport of obsidian from the remote island [9]. Some of them may have moved south to Northern Ryukyu and possibly to Central Ryukyu. However, as mentioned above, the Central and Southern Ryukyus had a deep relationship with Southeast Asia during the late Pleistocene, and the earliest migrations were probably accomplished by deliberate voyages by the population who developed the maritime technology and cultural flexibility to adapt to the island environments in the islands of Southeast Asia.

In the Jomon period, the influence of mainland Japan on Northern and Central Ryukyu became noticeable, while the influence of Southeast Asia became unrecognizable, except for the possible connection between Southern Ryukyu and Taiwan. In the Japanese Archipelago, fishing and shell gathering became noticeable, and shell middens began to be left during the Jomon period. In the early Jomon period, pelagic fishes were consumed, complex fishing gear appeared, and migration beyond the Kuroshio Current to Hachijo Island occurred [78]. The activation of maritime behavior and technology in Jomon might be the cause of its increasing influence on the Ryukyus from Jomon. However, maritime activity must have increased in Southeast Asia at the same time. Therefore, we must consider causes other than the development of marine activities and voyaging technologies, such as population growth and/or movement, to explore new horizons.

People have inhabited Okinawa Island, Central Ryukyu, continuously or intermittently from around 35,000 to 13,000 years ago. A unique culture, in which a variety of shell artifacts were made, developed in Sakitari Cave as an adaptation to island conditions. However, the genetic influence of these Paleolithic people has not been recognized in modern Okinawan populations, and it appears they were replaced by the Jomon and later immigrants from mainland Japan. Repeated influences of mainland Jomon were observed in archeological remains. As this culture spread, there must have been a considerable number of immigrants based on the genetic studies. In addition, the population growth rate of Holocene immigrants may have been higher than that of the Paleolithic population.

In the Northern and Central Ryukyus, cultural changes became basically consistent with the Jomon culture, while each island developed its own individuality. In Southern Ryukyu, there is a long hiatus between the Paleolithic occupation and the early Holocene settlement. The cultural positions of the Shimotabaru and Mudoki periods have not yet been determined, and ancient DNA research has only recently begun. There are many unanswered questions regarding the prehistoric migration history of Southern Ryukyu.

There are also many issues to be resolved regarding the migration history and culture of the Paleolithic period in Central Ryukyu. This is due to the lack of information and the paucity of Paleolithic and early Holocene archeological sites. We hope that these problems will be solved by recently increasing cave excavations throughout the Ryukyus and by analyses of archeological and human remains using modern techniques, such as chemical analyses and ancient DNA studies.

Acknowledgments

This work was supported by JSPS KAKENHI Grant Numbers JP 22H00027 and 21H04368. We would like to thank Editage (www.editage.com) for English language editing.