Marine resource exploitation and maritime adaptation in the Southern Ryukyus during the Neolithic: a short report on the Shimotabaru shell mound, Okinawa, Japan

In the Southern Ryukyu Islands (Okinawa, Japan), the most important and well-known site with Neolithic pottery is the Shimotabaru shell-mound, located on the south coast of Hateruma Island, approximately 200 km east of Taiwan. The site has yielded a large assemblage of ceramic sherds that are distinct from earlier Jomon pottery, which dates to between 10,000 and 6,000 years Before Present (BP) and was possibly introduced by Jomon hunter-gatherer populations from northern Japan (e.g., Kyushu Island). The site is dated to approximately 3,700–3,200 BP and has also produced numerous artefacts and faunal remains, including fish bones. This study presents a short zooarchaeological analysis of the excavated fish remains from the Shimotabaru site, recovered during the 1980s, to reconstruct maritime adaptation and marine resource use during the Neolithic. The results indicate a strong reliance on nearshore reef-associated fish, particularly large parrotfish and sharks, with no evidence of pelagic species, suggesting that fishing efforts were concentrated in shallow coastal waters. These findings contribute to broader discussions on subsistence strategies in early Austronesian migration contexts, providing insights into how prehistoric communities in the Ryukyu Islands exploited marine resources in comparison to contemporaneous sites in Island Southeast Asia and Oceania.

Outside Wallacea, the Ryukyu Islands provide some of the earliest evidence of intentional long-distance maritime voyaging in the world. Situated between Taiwan and Kyushu Island, Japan, the archipelago was first colonized by modern humans during the Late Pleistocene, at least by 37,000 years Before Present (BP) (Fujita et al., 2016, 2024; Kaifu et al., 2020, 2025; Takamiya et al., 1975, 2019). These Late Pleistocene hunter-gatherer populations are thought to have reached the islands before the Last Glacial Maximum (LGM), using watercraft to cross the open sea and subsequently adapting to insular environments (e.g., Kaifu 2022; Kaifu et al., 2020, 2022, 2025). Comparable in age to early sites in Mindoro, Philippines (Boulanger et al., 2023, 2019; Pawlik and Piper, 2019), and the Lesser Sunda Islands of Wallacea (e.g., Hawkins et al., 2017; Kaharudin et al., 2024; Kealy et al., 2020; O’Connor et al. 2011; Shipton et al., 2024), the Ryukyu Islands provide a unique archaeological context for investigating the development of maritime skills in modern humans.

Geologically, the Ryukyu Islands stretch across a 1,200 km arc connecting Kyushu to Taiwan, with the southern region comprising over 70 islands that were formed during the Quaternary (Furukawa, 1979) (Figure 1). This unique geographical and geological context positions the Ryukyu Islands as a critical nexus for early maritime migrations, facilitating human movement and cultural exchanges connecting Japan to the north with southern Taiwan and Southeast Asia (Fujita et al., 2016, 2020, 2024; Nishimura et al., 1960; Takamiya and Nakamura, 2020). It is argued that two distinct cultural identities emerged in the archipelago: a ‘Jomon’ identity in the Northern and Central Ryukyus and an ‘Austronesian’ identity in the Southern Ryukyus (Figure 1). While these identities seem to reflect broad similarities in subsistence strategies across both zones, they were constructed through distinct cultural and social practices (e.g., Hudson, 2012).

This cultural differentiation becomes particularly evident when we examine the development and distribution of pottery in the Ryukyu Islands. Pottery production in the Northern Ryukyus now dates back to 14,000 years BP (ATBE, 2024; OPMAM, 2022), where Jomon pottery has been excavated, but the timeline for early pottery culture in Southern Ryukyus remains less clear (Yamagiwa et al., 2019). In the Sakishima Islands (Figure 1), a distinct type of local pottery appears around 4,000 years BP (Ohama, 1999; OPBE, 1986), marking the beginning of the ‘Shimotabaru Age’. This period is named after the Shimotabaru shell midden on Hateruma Island (Figure 1), where this distinctive pottery type was first discovered in the 1950s (Kokubu 1972; Kanaseki and Kokubu 1979). The Shimotabaru Age is characterized by a distinctive pottery tradition, known as “Shimotabaru pottery”, which differs markedly from the Jomon pottery found in the Northern and Central Ryukyus and elsewhere in the Japanese archipelago. The origins of this pottery remain uncertain (Yamagiwa et al., 2019). Possible Jomon pottery has recently been identified on Ishigaki Island, dating to around 10,000 BP, and Jomon genomes have been recovered from Neolithic deposits at the Nagabaka site on Miyako Island (Robbeets et al., 2021). These discoveries demonstrate a Jomon presence in parts of the southern Ryukyus before the emergence of the Shimotabaru cultural tradition and indicate that interpretations of Neolithic settlement based solely on Austronesian migrations from Taiwan should be treated with caution. Around 4,000 BP, the earlier Jomon pottery tradition disappeared and was replaced by Shimotabaru pottery throughout much of the southern Ryukyus, although no pottery dating to the Shimotabaru Age has yet been recovered from Miyako Island. Chronologically, the Shimotabaru Age overlaps with the initial Austronesian dispersals from Taiwan proposed on the basis of linguistic evidence (Asato, 1993; Summerhayes, 2016). However, the archaeological and genetic evidence suggests that the relationship between these dispersals and the emergence of the Shimotabaru cultural tradition remains unresolved.

The Shimotabaru pottery tradition, which has been identified across the Sakishima Islands, including Hateruma, Iriomote, Ishigaki, and Miyako (Figure 1), became the dominant ceramic tradition in the southern Ryukyus around 3,700 BP. Its widespread distribution reflects a major cultural development during the Neolithic, which may have resulted from local innovation, increased interaction among island communities, population movement, or a combination of these processes (Summerhayes, 2016; Takamiya and Nakamura, 2020). However, recent analyses have shown that Shimotabaru pottery differs from the decorated and red-slipped Neolithic pottery traditions of Taiwan commonly associated with early Austronesian populations, suggesting that its origins remain uncertain and casting doubt on earlier interpretations of strong cultural connections between the southern Ryukyus and surrounding regions (Aoyama et al., 2020).

Building on the existing findings, the limited understanding of migrations and the Neolithic period in the southern Ryukyus emphasizes the need to examine additional material evidence from archaeological sites. Among these, faunal remains, particularly fish, are vital for gaining deeper insights into subsistence strategies, maritime adaptations, and patterns of human movement (Colley, 1990). This study investigates prehistoric fishing practices in the Ryukyu Islands, with a primary focus on the Sakishima Islands, through a re-analysis of fish bones from the Shimotabaru shell mound site on Hateruma Island.

History, chronology and key findings of the Shimotabaru shell mound

The Shimotabaru shell mound site is located on Hateruma Island, a small island (about 13 km 2) situated approximately 200 km from Taiwan, in the far-western region of the Ryukyu Islands and the Japanese archipelago (Figure 1). The coastal environment of Hateruma Island varies considerably around the island, with accessible reef-flat, lagoon, and sandy coastal environments concentrated mainly along its northern and western coasts, whereas much of the southern and eastern coastline is rocky (Kawana and Oshiro, 1978; Ota et al., 1985; Radtke et al., 1996). Shimotabaru is located on the northwestern coast, adjacent to a shallow reef and lagoon environment. Such local variation is important because reef-fish communities in the Ryukyu Islands vary according to reef zone, depth, and distance from shore (Lecchini et al., 2003). The site has been excavated three times during which three main sections have been investigated. The first excavation in 1954 was conducted by Takeo Kanesaki (Tezukayama University) and Naoichi Kokubu(Kumamoto University) (Kanesaki et al., 1964; Kokubu, 1972), while the second excavation was carried out by Waseda University in 1958 (Nishimura et al., 1960). The third and most extensive excavation was carried out by the Okinawa Prefecture Educational Board in 1984-1985 (OPBE, 1986). The third and most extensive excavation was carried out by the Okinawa Prefecture Educational Board in 1984-1985 (OPBE, 1986).

The 1980s excavation of the site was conducted mainly in three areas of the site (Supplementary Figure S1). Among these areas, four large excavation trenches were excavated based on stratigraphy. The total excavation area was approximately 400 m 2 (16 × 25 m 2), according to the site report (OPBE, 1986). At least fourteen 4 × 4 m trenches, along with several smaller trenches, were opened in Area 1, while six 4 × 4 m grids were excavated in Area 2. Area 3 contained four 4 × 4 m squares with two small test pits (possibly 1 × 1 m). All artifacts, including the fish remains analyzed in the previous study, were collected by hand. Although the excavated soils were screened, the sieved materials were neither analyzed nor reported in the excavation report (OPBE, 1986). The stratigraphy of Areas 1 and 3 is divided into four layers, while that of Area 2 comprises five layers (Supplementary Table S1).

Radiocarbon dates obtained from previous excavations indicate that the site was formed between approximately 3,700 and 3,200 BP (Supplementary Table S2), further establishing Shimotabaru as a key archaeological site associated with the ‘Shimotabaru Age’ of the Neolithic period in the Sakishima Islands and the southern Ryukyus. Adjacent to Shimotabaru is the Oodomarihama site (Supplementary Figure S1), which dates to the late to post-Neolithic period and is thought to have functioned as a habitation site following the occupation of Shimotabaru (OPBE, 1986). The 1980s excavations revealed a substantial quantity of wild pig (Sus scrofa riukiuanus) remains that likely represent wild boar originating from Iriomote or Ishigaki Island, as wild boar are not native to Hateruma Island. In addition to Sus remains, a wide range of other faunal materials was recovered, including birds (Diomedeidae and Corvidae), marine turtles (Chelonidae), marine mammals (Delphinidae and Dugongidae), micromammals (Pteropodidae and Muridae), fish, and molluscan shellfish (OPBE, 1986). Marine shellfish were also particularly abundant across the site, with assemblages dominated by Turbinidae, Strombidae, Tridacnidae, Cyrenidae, and Trochidae; with especially high concentrations recorded in the earlier occupational layers (OPBE, 1986).

Previous analysis of the Shimotabaru fish remains

The previous analysis of fish bones excavated from Shimotabaru during the 1980s was partially conducted by Kaneko, who reported a total NISP of 842 (Kaneko 1986; OPBE, 1986). All bones were recovered by hand as in situ finds during the excavation. Among these, Kaneko specifically identified 546 specimens representing seven fish families. The majority of the identified fish remains derive from inshore coral reef environments and are primarily dominated by parrotfishes (Scarinae), along with elasmobranchs (sharks and rays) (Figure 2; Supplementary Table S3). Under current taxonomic frameworks, parrotfishes are classified within the subfamily Scarinae of the family Labridae. However, in order to maintain consistency with the analytical framework adopted by Kaneko, parrotfishes (treated as Scaridae in the original study) and Labridae are discussed separately here, following the original classification scheme.

Additional inshore coral reef fish families identified include Diodontidae (porcupinefish), Balistidae (triggerfish), Muraenidae (moray eel), Lethrinidae (emperor fish), Labridae (wrasse), and Serranidae sensu Kaneko, which corresponds to the family Epinephelinidae (groupers) under current taxonomy (Figure 2). The lethrinid species Monotaxis grandoculis (humpnose bigeye bream) was also identified. Shark remains consist of teeth, vertebrae, and worked teeth of Galeocerdonidae (tiger shark) and Lamnidae (Supplementary Table S3) (Kaneko 1986; OPBE, 1986).

Re-analysis and identification of the Shimotabaru fish remains

We recently conducted a re-analysis of the fish remains originally studied by Kaneko, as these are the only specimens with detailed records of excavation locations. Currently housed at the Okinawa Prefectural Archaeological Center, the bones are stored with associated stratigraphic information and sorted into individual plastic bags.

We identified the remains using updated methods that incorporated a wider range of skeletal elements than in previous studies, including vertebrae, fin elements, teeth, palatine, hyomandibular, pharyngeal bones, cranial parts, and caudal elements, which substantially increased the number of identified specimens (Boulanger, 2023; Ono and Clark 2012; Ono and Intoh, 2011). All analyses were conducted on-site at the Okinawa Prefectural Archaeological Center (OPAC), where the Shimotabaru assemblage is curated. Due to institutional regulations, the faunal remains could not be removed from the facility for off-site study, and all analytical procedures therefore had to be carried out in Okinawa during several research visits (six days in total). As no comprehensive modern comparative osteological reference collection was available at OPAC, identifications relied primarily on the ichthyoarchaeological expertise of the first two authors and published osteological descriptions (e.g., Bellwood, 1994 for parrotfish identification). The present analysis provides the most comprehensive taxonomic assessment currently achievable under these conditions. In addition, a limited number of comparative specimens from the National Museum of Ethnology (Minpaku), particularly Scarinae, Acanthuridae, Labridae, Lethrinidae, and Serranidae, were brought to OPAC to assist with comparative checks during the analysis. For fish bones that could not be confidently identified on-site, photographs were taken and subsequently compared with the more extensive reference collection at Minpaku. Consequently, uncertain specimens were conservatively assigned to higher taxonomic levels where appropriate. Although these logistical constraints limited access to a comprehensive comparative collection throughout the study, they do not diminish the value of documenting this unique archaeological assemblage, and the present analysis provides the most comprehensive taxonomic assessment currently achievable under these conditions.

The primary aim of this work was therefore to reassess the assemblage, evaluate its taxonomic potential, and make more detailed information available to an international readership, as previous documentation of the material has largely remained accessible only in Japanese. Particular attention was given to selected diagnostic elements, especially among Scarinae specimens, in order to strengthen taxonomic determinations. Nevertheless, given the short duration of the analysis, the limited comparative material available, and the exceptionally high marine biodiversity of the region, located adjacent to the Coral Triangle, the world's richest marine biodiversity hotspot, spanning six countries in Southeast Asia and the western Pacific many specimens could only be identified to the family level.

Quantification of the assemblage followed standard zooarchaeological procedures using the Number of Identified Specimens (NISP) as the primary analytical unit. In addition, the Number of Minimum Elements (NME) was calculated by estimating the minimum number of each skeletal element present based on element portion and taxonomic identification. The Minimum Number of Individuals (MNI) was not calculated for fish remains, as the conditions of the analysis in the present study constrained its applicability: the assemblage derives from older excavations in which material was recovered primarily by hand rather than systematic sieving, and the on-site study conditions limited the possibility of conducting exhaustive refitting or element-based sorting. Quantification therefore focused on NISP and NME, which are generally considered more appropriate for ichthyoarchaeological assemblages (Butler, 1993; Casteel, 1976; Lyman, 2008). All fish remains were examined under a stereomicroscope (Shimadzu Rika STZ-171-TLED) to identify evidence of cut marks, and digital images were captured using a Shimadzu Moticam BTW. Associated lithic materials, including quartz and chert flakes, were also examined under the same stereomicroscope to assess potential use-wear on their edges.

Taxonomic identification

By expanding the range of skeletal elements used for taxonomic identification, the present analysis substantially increased both the NISP and the taxonomic resolution of the assemblage. In total, 3,172 fish elements were identified (Figure 2; Supplementary Table 3), and the family-level NISP increased markedly from 546 in the previous study to 1,413 specimens in the current re-analysis (Supplementary Table S3).

Vertebral elements constitute the largest and most diverse category of identified remains, and their inclusion significantly increased both the number of recognized families and their associated NISP values (Supplementary Table S3). Families identified primarily through vertebral elements include Muraenidae (moray eel), Epinephelinidae (grouper), Balistidae (triggerfish), Acanthuridae (surgeonfish/unicornfish), Scarinae (parrotfish), Belonidae (needlefish), Sphyraenidae (barracuda), Carangidae (jack/trevally), and Holocentridae (squirrelfish/soldierfish) (Figure 2; Supplementary Table S3). Whereas previous analyses recognized a species richness of ten taxa, the present study identifies at least fifteen taxa, including five families not previously reported: Acanthuridae, Belonidae, Sphyraenidae, Carangidae, and Holocentridae (Figure 2; Supplementary Table S3).

Scarinae (parrotfish) represent the most abundant fish group at the Shimotabaru site, and their NISP increased substantially in the present re-analysis. They also exhibit the greatest diversity of identifiable skeletal elements (Supplementary Table S3). Diagnostic pharyngeal elements confirm the presence of at least two genera, Scarus and Chlorurus. In addition, Cetoscarus ocellatus (spotted parrotfish) is identified, representing the only species of Cetoscarus currently recorded from this region (Froese and Pauly, 2026). Some large pharyngeal specimens may also correspond to Bolbometopon muricatum (green humphead parrotfish), the only extant species in the genus. This species has been reported primarily from the Central and Southern Ryukyus (Fujiwara and Motomura, 2020; Mochida and Motomura, 2018; Motomura and Harazaki, 2017; Motomura and Uehara, 2020), with no previous records from the Northern Ryukyus or the Amami Islands until its recent discovery on Amami Island (Hata et al., 2019). Additional cranial elements, including premaxillae and dentaries, were also identified for this taxon (Supplementary Table S3). Osteological differentiation between Bolbometopon muricatum and Cetoscarus ocellatus is difficult because of similarities in their pharyngeal morphology (Bellwood, 1994). Nevertheless, several particularly large pharyngeal bones from the assemblage seem to exceed the documented size range of Cetoscarus ocellatus and could therefore tentatively be attributed to Bolbometopon muricatum. However, this observation is based on macroscopic assessment of the specimens and should therefore be considered descriptive rather than analytical. Modern populations of Bolbometopon muricatum have experienced significant declines, and their capture is now restricted in many regions due to conservation concerns (e.g., Aswani and Hamilton, 2004; Donaldson and Dulvy, 2004).

Re-identification of the shark remains yielded a total NISP of 166, including both vertebrae and teeth (Figure 3). The increase compared to Kaneko (1986) primarily reflects the inclusion of additional shark teeth that had previously been excluded from analysis. Of these remains, 97 specimens are vertebrae and 61 are teeth (Supplementary Table S3). Among the identified teeth, 40 are attributed to Galeocerdo cuvier (tiger shark) and 20 to Carcharhinidae (requiem sharks). Five Lamnidae teeth previously reported by Kaneko were not included in this assemblage, as they have been curated separately. Half of the tiger shark teeth (NISP = 20) exhibit perforations, suggesting intentional modification, possibly for use as ornaments or tools (Figure 3).

Cut marks and stone flakes

We identified at least three fish bones bearing cut marks (Figure 4). These marks suggest the use of sharp tools, with stone flakes being the most likely implements used for processing shark and fish meat. Previous reports did not report the presence of such tools or flakes at the Shimotabaru shell mound site. However, our re-examination of all excavated lithic materials confirmed the presence of 80 quartz and four chert utilized flakes (UFs) that may have been used as cutting tools (the latter clearly classified as lithic tools) (Supplementary Figure S2).

Our re-analysis of the fish remains from the Shimotabaru shell mound has expanded the known taxonomic diversity of the assemblage, increasing both richness and the total NISP, and highlighting the importance of incorporating additional skeletal elements and updated comparative references in the study of archaeological fish bones.

Although fish remains in archaeological contexts often degrade rapidly (Willis and Boehm, 2014), the Shimotabaru assemblage includes numerous specimens with preserved surfaces and identifiable diagnostic elements, including relatively large and morphologically informative bones. Such preservation is particularly valuable given the scarcity of Neolithic sites in the southern Ryukyus with comparable faunal assemblages. This has allowed for more detailed taxonomic analyses, which can provide insights on fishing practices. However, our findings are inevitably constrained by the limitations of comparative specimens and the original excavation methods, which did not include sieving. As a result, smaller fish remains, particularly from small-bodied taxa or juvenile individuals, are likely underrepresented. Fine-mesh sieving would have enhanced diversity, especially for identifying small reef fish (Nagaoka, 1994, 2005; Zohar and Belmaker, 2005; Zohar et al. 2005), which are known to be a crucial component of subsistence in many island contexts (e.g., Leach, 2006; Najima 2013). The absence of sieving also limits our ability to fully assess the diversity of fishing strategies employed at Shimotabaru, as certain capture methods, such as netting or trapping smaller fish, may be underrepresented (e.g., Bouffandeau et al., 2018; Chew et al., 2025; O’Connor et al. 2019).

Prehistoric fishing at Shimotabaru appears to have been heavily focused on nearshore environments, particularly coral reefs and shallow coastal waters, where marine resources were plentiful. This reliance on inshore fishing likely reduced the need to venture into pelagic waters. The identified fish taxa align with species commonly found in these habitats, highlighting a subsistence strategy centered around easily accessible coastal resources. This interpretation is supported by the presence of fish taxa typically associated with coral reefs, the most abundant being parrotfish.

Notably, these findings align with Kaneko's earlier study (1986), which documented prehistoric fishing around Hateruma Island. The discovery of large parrotfish, possibly including Bolbometopon muricatum, Cetoscarus ocellatus, and Chlorurus spp., suggests that these taxa were once abundant on the coral reefs surrounding Hateruma Island. Among them, Bolbometopon muricatum typically forms schools in shallow reef habitats during the day and shelters around large coral formations at night (Bellwood, 2001; Hamilton et al. 2016; Roff et al., 2017), making it a readily accessible target for Neolithic fishers.

Ethno-archaeological and ecological anthropology studies on reef fishing in Oceania, Island Southeast Asia, and the Ryukyu Islands report that parrotfish are mainly captured using netting during the day and spearing at night. Stone tidal weirs and basket traps are also effective in capturing parrotfish species. Other fish families identified primarily through vertebral elements, including Muraenidae, Balistidae, Acanthuridae, Belonidae, Sphyraenidae, Carangidae, and Holocentridae, are also associated with specific fishing strategies documented in ethnographic and archaeological studies across the tropical Pacific (Boulanger et al., 2022; Butler, 1994; Dalzell et al., 1996; Kirch and Dye, 1979; Masse, 1986; Najima, 2002; Ono, 2003, 2010). Moray eels, which inhabit reef crevices and rocky substrates, are commonly captured through spearing or by extracting individuals from reef holes using hooked implements or baited traps. Triggerfishes (Balistidae) and surgeonfishes (Acanthuridae) are typically caught using nets, traps, or spearing, particularly in shallow lagoon and reef flat environments. Needlefishes (Belonidae) and barracudas (Sphyraenidae), which occupy near-surface and pelagic reef-edge habitats, are frequently targeted using line fishing, trolling, or spearing from canoes. Similarly, Carangidae (jacks and trevallies), which are mobile predators often forming schools, are commonly captured using hook-and-line fishing, netting, or trolling techniques. Holocentridae, which are nocturnal reef dwellers, are often caught through night spearfishing or trapping methods in reef environments (Boulanger et al., 2022; Kirch and Dye, 1979; Najima, 2002; Ono, 2003, 2010). The expanded taxonomic diversity suggests that prehistoric fishing strategies at Shimotabaru were more varied than previously recognized and likely incorporated multiple capture techniques adapted to different ecological niches.

The identification of sharks in the assemblage also indicates that they played a substantial role in prehistoric fishing at Shimotabaru. Ethnographic records from the Ryukyu Islands, together with the presence of large shark remains at the site, suggest the possibility that sharks were captured using hooks and ropes, employing a nosing method (Nakama and Demura, 1981). Similar techniques have also been documented in Oceania, where sharks are lured with bait before being secured with ropes (Anell, 1955; Boulanger et al., 2025; Ono and Intoh, 2011). Although direct evidence for this practice is lacking at Shimotabaru, the ethnographic parallels from both the Ryukyu Islands and Oceania suggest that comparable fishing techniques may have been employed in prehistoric Hateruma (Kishaba, 1977; Nakama and Demura, 1981).

Together, these findings may highlight a targeted approach to nearshore fishing involving diverse techniques and tools that allowed prehistoric populations to exploit a wide range of marine habitats that are also reported in other prehistoric sites on Miyako Island and Ishigaki Island in the southern Ryukyus (Matsui et al., 2017; Najima, 2007, 2013; OPAC, 2013, 2017; Toizumi, 2007, 2010).

The presence of relatively large tiger sharks raises the possibility that the Shimotabaru people may have occasionally used small watercraft to access reef-edge or outer-reef habitats, where these sharks are most commonly found. Although this cannot be demonstrated directly from the faunal assemblage, dugout canoes are plausible candidates, as they are known from the Japanese Archipelago by at least 7,500 BP (Hattori and Ota, 2019; Kaifu et al., 2020, 2025). Furthermore, the recovery of numerous stone adzes from Shimotabaru and nearby sites on Iriomote and Ishigaki provides indirect evidence for woodworking activities that may have included the manufacture of watercraft. Similar tools have been widely reported across the western Pacific, including the Ryukyu Islands, where they are commonly associated with boat construction and maritime subsistence practices (Anderson and Summerhayes, 2008; Asato, 1991; Pearson et al., 1978; Shipton et al., 2020). Together, this evidence suggests that small watercraft may have facilitated the exploitation of reef-edge environments, although their use at Shimotabaru remains hypothetical. While Austronesian-speaking populations introduced advanced maritime technologies around 4,000 BP, including deep-sea fishing techniques in some areas (e.g., Amesbury, 2013; Campos, 2013 Ono and Intoh, 2011);, the evidence from Shimotabaru suggests that such innovations did not significantly influence local fishing practices. Instead, the community's reliance on inshore fisheries reflects a subsistence strategy closely adapted to the ecological conditions of the southern Ryukyus. Although the subtropical coastal environments of the Ryukyu Islands differ from the tropical reef ecosystems of Oceania, similar patterns of intensive inshore fishing have been documented across many Austronesian and Pacific Island archaeological sites despite the development of sophisticated maritime technologies (e.g., Boulanger, 2023; Ono, 2010; Ono and Clark, 2012; Ono et al., 2019). In both regions, this emphasis on inshore fishing is likely to reflect ecological constraints, including strong ocean currents and deep waters beyond the reef, which would have made offshore fishing more challenging. These comparisons are intended to provide a broader ecological framework rather than to imply direct cultural continuity or identical fishing practices between the Ryukyu Islands and Oceania.

Finally, the recovery of cut marks on fish bones from the Shimotabaru site provides new evidence of butchering and processing activities using lithic tools. Because the raw materials used to manufacture these tools do not occur naturally on Hateruma Island, they were likely transported intentionally for tool production and use. Although a small number of presumably flaked slate tools have previously been reported from the Neolithic Nagabaka site on Miyako Island (Hudson, 2013), the present study provides the first evidence for the use of flake tools at the Shimotabaru site and contributes new insights into technological practices and subsistence activities during the Neolithic of the southern Ryukyus. The presence of cut marks indicates that fish were not only caught but also carefully processed, using tools to remove flesh, bones, and other edible parts from the carcasses. However, despite this evidence of fish processing, it is noteworthy that no direct evidence of fishing tools, such as fishhooks, nets, or other specialized implements, has been recovered from Shimotabaru and other archaeological sites in the Southern Ryukyus.

To better understand the dynamics of fish exploitation at Shimotabaru and the ecological implications of early fishing practices, future research should prioritize further excavations and the continued expansion of the comparative osteological reference collection through the addition of new specimens. Complementary biomolecular approaches, including ZooMS and ancient DNA analyses, will also be valuable for resolving selected taxonomic questions as proteomic and, where available, genomic reference datasets continue to expand (Boulanger et al., 2025). Additional excavations could reveal greater taxonomic diversity and document changes in the composition and size of exploited fish through time. Together, these approaches will provide critical insights into broader patterns of marine resource use in the Ryukyu Islands and improve our understanding of how prehistoric populations adapted to and interacted with coastal ecosystems. Our findings suggest similarities between prehistoric fishing strategies on Hateruma and those of the Lapita culture and other Neolithic Austronesian groups, further comparative analysis is essential. Expanding this research to include other sites in the Ryukyu Islands, as well as Island Southeast Asia and Oceania, would provide a broader context for understanding early fishing practices and their regional significance. The relatively frequent presence of large individuals in the archaeological assemblage therefore provides an important historical baseline for evaluating long-term changes in reef fish populations and the cumulative impact of human exploitation on marine ecosystems.