Portal de Periódicos da CAPES

Discovery of Iron Grapnel Anchors in Early Modern Ryukyu and Management of Underwater Cultural Heritage in Okinawa, Japan

2016; Taylor & Francis; Volume: 45; Issue: 1 Linguagem: Inglês

10.1111/1095-9270.12145

1095-9270

Rintaro Ono, Chiaki Katagiri, Hironobu Kan, Masayuki Nagao, Yumiko Nakanishi, Y. Yamamoto, Fumiaki Takemura, Norimitsu Sakagami,

Maritime and Coastal Archaeology

The Yarabuoki underwater site contains seven iron grapnel anchors and Early Modern Okinawan ceramic jars and is dated to the 16th–19th centuries. The site lies at a depth of 12–32 m off the western coast of Ishigaki Island in Okinawa, Japan. Based on underwater archaeological and broadband multibeam surveys, as well as historical research of the artefacts and Early Modern Ryukyuan shipping, we discuss the possible anchor and vessel types in Ryukyu and Eastern Asia in Early Modern times. We also discuss the efficacy of low-cost ROV for assisting surveys in shallow-water environments and the value of educational programmes for promoting the management and conservation of underwater cultural heritage. 要約 屋良部沖海底遺跡は、16ཞ19世紀頃と推測される沖縄近世陶器群と7点の四爪鉄錨からなる。遺跡は沖縄県の石垣島西岸の水深12ཞ32mの海底に位置している。本論文では、この遺跡で発見された遺物群を対象としたこれまでの水中考古学的調査とマルチビーム探査の成果から、近世琉球王国における海運で利用されていた船や、それらの船が搭載していた錨のタイプについて論じる。このほかに本論文では、この遺跡調査で新たに開発した低コスト型のROVによる沿岸海域での効用、水中文化遺産の保全と活用を目的とした教育プログラム開発における新たな可能性についても論じる。 Academic and public interest in underwater cultural heritage (UCH) is increasing in Japan, partly because of the first designation of a permanently submerged site by the state government in 2011 at Takashima-Kōzaki. This site, in Nagasaki Prefecture in Kyushu, where 13th-century ships of the Yuan Dynasty China were discovered, was designated a National Historic Site after an extensive underwater investigation. Over the past two decades, the number of registered underwater archaeological sites subject to the Japanese Law for the Protection of Cultural Properties increased from only about 300 in 1998 to 512 sites in 2012 (Japanese Agency for Cultural Affairs, 2012: 32). Of these 512 listed sites, 34 are located in Okinawa Prefecture, one of the leading areas for underwater archaeological investigations. Okinawa Prefecture, or the Ryukyu Islands, consists of 199 islands in a chain that stretches some 1200 km between Kyushu Island and Taiwan (Fig. 1). Underwater archaeological surveys and research projects have been conducted in Okinawa since 1998. A total of 211 sites had been recognized by the Okinawa Prefectural Archaeological Center (the Center, hereafter) by 2014, many in a relatively good state of preservation (Katagiri, 2010; Okinawa Prefectural Museum, 2014). Research has been conducted by Kagoshima University and Nansei Islands Underwater Cultural Heritage Study Group (2010; 2011), the Center, Okinawa Prefectural Museum and Art Museum (for example Katagiri, 2007; 2009; 2010; Miyagi et al., 2005), and the Asian Research Institute of Underwater Archaeology (2012). The Yarabuoki underwater site near Ishigaki Island is one of those discovered by these recent surveys. Ishigaki is the most populated of the Yaeyama Islands, a subgroup of the Ryukyu Islands. The other major islands in the Yaeyama group are Iriomote, Taketomi, Kuroshima, Hateruma and Yonaguni. A total of 69 underwater archaeological sites are currently known off the coasts of these islands (Fig. 2). Our team is currently undertaking an interdisciplinary research project to evaluate the historical and archaeological significance and potential of the site as a cultural resource. This article presents up-to-date results of our current research and discusses the significance of the site in the context of reconstructing overseas trade routes in pre-Modern East Asia. We also report on cultural resource management practices for UCH in the Yaeyama Islands. The site was discovered in 2009 by a local professional diver, Mr Seiji Fujii, who has been running a diving service for more than 30 years from Ishigaki Island. He reported the approximate location of the site to the Center. In 2010, Katagiri (2010) conducted a site inspection survey as the chief investigator at the Center with the assistance of the Nansei Islands Underwater Cultural Heritage Study Group (2011). The survey confirmed the location of seven iron grapnel anchors of various sizes and a cluster of Early Modern Okinawan ceramic jars (tsuboya-yaki), which were originally produced in Okinawa Island from the 16th to the 19th centuries, during the Edo Period in Japan and the Early Modern Ryukyu Kingdom Period in Okinawa. They are the first grapnel anchors found in Okinawa and grapnel anchors have not previously been found in the region. Thus, these remains provide the opportunity to investigate both the past anchor types that equipped Ryukyuan and other vessels that navigated around the Ryukyu Islands and maritime trade during the Early Modern Ryukyu Kingdom Period. Based on this preliminary survey, our aim was to locate the site with more accuracy, to find evidence for a precise date, and to attempt to characterize the site to determine, for example, if it was a harbour, anchorage, or shipwreck site. Accurate and detailed recording is the essential first step to understand a site. Moreover, geographical and topographical data can greatly contribute to reconstructing site formation. Therefore, multiple methodologies were employed including GPS positioning (Table 1), a broadband multibeam sonar survey, the creation of a high-definition video record using a small-sized, lightweight underwater robot, and a record and analysis of each artefact. A bathymetric map was provided by the Geospatial Information Authority of Japan (GIS) and is shown here with the location of artefacts plotted (Fig. 3). However, this map proved to be insufficiently detailed and did not accurately represent seafloor topography. To solve this problem, we conducted a multibeam echo sounding (MBES) survey around the site to visualize the bathymetric results at a lateral grid resolution of one metre. The MBES survey was conducted in August 2011, using Sonic 2022 (R2 Sonic, LLC) and its accessory system which had been introduced to the laboratory at Okayama University in 2011, where H. Kan was affiliated. The Sonic 2022 has a variable ultrasonic frequency of 200–400 kHz with 256 ultrasonic beams and provides selectable swath coverage of 10–160°. The typical ultrasonic beam width, parallel and orthogonal to the direction of travel, is within one degree of each other when an ultrasonic frequency of 400 kHz is selected. We used a VS111 GPS compass system with A20 and A30 antennas (Hemisphere Inc.) combined with a dynamic motion sensor (DMS-10, Teledyne TSS Ltd), a sea surface sound velocity sensor (miniSVS, Valeport Ltd), and a sound velocity profiler (MicroSVP, AML Oceanographic Ltd). The sounding method is similar to that used in the Nagura Bay survey (Kan et al., 2015) but the ultrasonic frequency of 400 kHz was selected for the entire area of the Yarabuoki underwater site. The depth of the surveyed area (Fig. 3B) ranges from 0.5 to 51.0 m. The HYPACK2010 software was used for both hydrographic survey and data processing. IVS3D Fledermaus was used for three-dimensional visualization with a grid size of one metre for the entire area. The MBES bathymetric map enabled us to create a high-resolution site map on which to plot the artefacts (Fig. 3C). An experimental part of the underwater survey at Yarabuoki, conducted by N. Sakagami and F. Takemura, used a purpose-built low-cost, portable remote operating vehicle ROV (Fig. 4) with a high-resolution digital video camera. Its use had the great advantage of saving time; as the survey area ranges in depth from 15 to 30 m, SCUBA divers could work for only limited periods. Anomalies could thus be captured and viewed on board via the video screen without entering the water. The ROV is 0.7 m long, 0.55 m wide and 0.3 m high, and weighs about 15 kg. It can reach a maximum depth of approximately 30 m. The robot is equipped with four thrusters to achieve surge, heave, and yaw motion: each generates a force of 2.2 kgf under water. The cylindrical waterproof housing contains equipment including a high-resolution camera, a small computer, depth sensor, magnetic compass, LED lights and batteries. We have also developed software so that an untrained user can manoeuvre the ROV easily using a video-game controller. The aim of this system was to invite public involvement and the promotion of the site to non-divers through the experience of operating the ROV and viewing the site. Two separate areas of artefact concentrations have been recorded at Yarabuoki (Fig. 3). The grapnel anchors can be divided into an eastern and a western zone, some 400 m distant from each other. Grapnel anchors Nos 1–4 are located in the eastern zone at a depth of 13–21 m, along with the ceramic jars. Anchor Nos 5–7 lie in the western zone at a depth of 25–32 m (Tables 1 and 2, Fig. 5). The shank of anchor No. 1 measures 1.97 m. The maximum width of the shank is 0.20 m and thickness 0.15 m. All four arms survive in good condition, with two of them buried in the sandy seabed (Table 1, Fig. 6a). Each arm is square in section with a length of about 1.80 m. The flukes are flattened and rectangular in shape. A loop formed at the top of the shank is 0.08 m in diameter. No additional ring is attached. Anchor No. 2 is somewhat smaller than No. 1: its shank measures 1.80 m in length, with a maximum width of 0.18 m and thickness of 0.12 m. Of the four arms, two survive in good condition, while two of them are buried in the sandy seabed (Fig. 6b). The crown, connecting the shank and arms, is embedded in coral. The arms are estimated to be square in section with a possible arm length of 1.20 m. The shank loop is oval and measures 0.13 m by 0.06 m. An additional ring, 0.27 m in diameter, is attached to the loop. Anchor No. 3 is much shorter than No. 1 and 2, with a shank length of only 1.50 m. Of the four arms, two of them are buried in the sandy seabed, while only the flukes of the other two are exposed. Both of the exposed flukes have been damaged and are incomplete (Fig. 6c). The shank of anchor No. 4 is one of the shortest, being 1.20 m in length. Of the four arms, two are buried in the sandy seabed, while only the points of the other two are exposed, and these are broken and incomplete (Fig. 6d). The crown is embedded in coral, but the arm section can be seen to be square in shape. The oval shank loop is 0.12 m by 0.09 m. An additional ring of 0.18 m diameter is attached to the loop. The ring has a circular 0.03 m-diameter section. The shank of anchor No. 5 is 1.60 m in length. All four arms survive in good condition, while one of them is buried in the hard coral seabed (Fig. 6e). The oval shank loop measures 0.22 × 0.16 m. An additional ring with a diameter of 0.30 m is attached to the loop. The ring has a circular, 0.08 m-diameter section. The shank of anchor No. 6 is 1.65 m in length. Its maximum width is about 0.15 m. All the four arms survive in good condition. One of them is buried in the hard coral seabed (Fig. 6f, g). Both of the exposed fluke points are incomplete. The anchor crown is embedded in coral, but the cross section of the arms is confirmed as rectangular. An additional ring is attached to the loop of 0.30 m diameter. The ring has a circular, 0.05 m-diameter section. The shank of anchor No. 7 is 2.05 m in length, with a maximum width near the crown of 0.20 m and 0.15 m thickness. Of the four arms, three are exposed in good condition and are complete (Fig. 6h), while one is buried in the coral rocky seabed. The anchor crown is embedded in corals, but the cross section of the arms can be confirmed as square. The flukes are flattened and rectangular in shape. The oval shank loop is 0.10 × 0.05 m. An additional ring is attached to the loop of 0.34 m diameter. The distribution of the anchors indicates that at least two or three ships lost their anchors at the site, as it is unlikely that all of these spatially dispersed anchors belong to a single ship (see Fig. 3C). There may be a relationship between the anchor sizes and their distribution. The largest two anchors (Nos 7 and 6) are located close together; as are the next two largest anchors (Nos 1 and 2). Anchor No. 4, the smallest found, in contrast, was located at some distance. The ceramic cluster is close to anchors Nos 1, 2 and 3, and thus all these finds may belong to a single ship. The ceramic jars are located close to grapnel anchors Nos 1, 2 and 3 on the seabed at the depth of about 20 m. In total, 12 jars have been recognized ranging in size from about 0.40 m to 1 m in height. All the jars are similar in form and material, which can be identified as the Early Modern Okinawan pottery known as tsuboya-yaki, originally produced on Okinawa Island (Fig. 7). The jars were recorded and left in situ, except for a small jar that was less embedded in the coral and easily recovered. The jar is housed in the Okinawa Prefectural Museum awaiting further study to identify the production site and precise date. These are the first iron grapnel anchors discovered in Okinawa and the Ryukyu Islands. The following sections will discuss the history of shipbuilding and marine transportation of the region and possible origins of these artefacts. In terms of dating, we estimate that these artefacts belong to the Early Modern period between the 17th and the 19th centuries AD (mainly the Edo Period in Japan and the Early Modern Ryukyu Kingdom Period in Okinawa) based on our historical knowledge and archaeological chronology of Okinawan pottery and ceramics, though it is hard to determine a more precise date at this stage. The Ryukyu Kingdom was an independent kingdom ruling most of the Ryukyu Islands from AD 1429 to 1874. The Kingdom was unified by Sho-Hashi, who was the last king of Chuzan Kingdom from three separated kingdoms called san-zan (san = three, zan = Ryukyuan dialect for 'country') including Hokuzan, Nanzan, and Chuzan in 1429. In the Sanzan Period from the early 14th century (c. 1322 AD) to 1429 AD, each of these kingdoms made inroads into maritime activities, and the Chuzan Kingdom settled a tributary relationship with the Ming Dynasty of China in 1372 AD (Matsuda, 2001). During the 15th and 16th centuries, the Ryukyu Kingdom thrived and was prominent in maritime trade with Southeast and East Asia for nearly 200 years (Okamoto, 2008). The Ryukyu Kingdom conquered the Yaeyama Islands at the southern end of the chain c. AD 1500, expanding its maritime network. In the late 16th century, however, the kingdom's commercial prosperity fell into decline, coinciding with the increased threat of wakō, or Japanese pirates, among other factors. The Ming gradually lost their power and interest in the region. The Tokugawa Shogunate that had ruled Japan since AD 1601 authorized the Shimazu Clan, the feudal lord of Satsuma Domain, to send an expeditionary force to conquer the Ryukyu Kingdom in 1609. The Shimazu Clan succeeded in occupying the Ryukyu Kingdom in the same year. In 1644, when the Qing Dynasty took over the former Ming Dynasty in China, the Ryukyu Kingdom remained a tributary state. The Tokugawa Shogunate formally approved this arrangement in 1655. Here, we define the Early Modern Ryukyu Kingdom as having started with the invasion of the Shimazu Clan of Satsuma in 1609 and ended with the formal annexation by the Japanese State in 1879. There were three major types of trading ship during the Early Modern Ryukyu Kingdom. The largest ships were shinko-sen (sen = ship) (進貢船) and Ryukyuan kai-sen (楷船), and were a type of junk. Shinko-sen were used for international trade, mainly with China (Ming and Qing Dynasty), while the Ryukyuan kai-sen (Fig. 8a) were often second-hand shinko-sen, repurposed for local trade between Satsuma and Ryukyu once the armaments had been taken off. Another ship type, maran-sen (馬艦船) were middle to small-sized wooden junks (Fig. 8b) made by Ryukyuan merchants for domestic trade, mostly within the Ryukyu Kingdom (for example Tomiyama, 2012). A shinko-sen could also be called To-sen, or 'Chinese ship', as they were predominantly used for international maritime trade between Naha, the capital port of the Ryukyu Kingdom, and Fujian (Fuzhou) in China. Between AD 1383 and around 1450, shinko-sen-type vessels were granted to the Kingdom by the Ming when Ryukyu became a tributary state. Possibly more than 30 vessels were granted to the Ryukyu Kingdom over a period of 70 years (Okamoto, 2008). These granted Chinese-built ships seemed to be separated in two types by size: 1) large-sized vessels with a loading capacity exceeding about 600 kL/m3 or a capacity of over 200 persons, and 2) middle-sized vessels with a loading capacity exceeding about 300 kL/m3 or a capacity of about 100 persons, based on Ryukyuan historical documents (for example Yamagata, 1996; Okamoto, 2008). However, when such grants from China ceased in the mid 15th century, the Ryukyu Kingdom appears to have started building such vessels at certain Fujian dockyards until around the early 16th century. Although the details are unclear, some historic documents (for example 'Rekidai Hōan' or 歴代宝案, which recorded the diplomatic activities of the Ryukyu Kingdom from 1424 to 1867 AD in a total of 270 volumes) mentions that the Kingdom sent timber and materials for shipbuilding from the Ryukyu Islands to Fujian and ordered the dockyards there to build shinko-sen-type ships (Okamoto, 2008: 227). By c.1570 AD, the Ryukyu Kingdom had started to build shinko-sen themselves within the Ryukyu Islands, mainly on Okinawa Island, and continued until 1609 AD, when the Ryukyu Kingdom was occupied by the Satsuma Clan (for example Owada, 1993; Yamada, 2007; Okamoto, 2008). Interestingly, the historical records describing these vessels newly built in Fujian and later in the Ryukyu Islands up until c.1510, were much larger than the previously granted Chinese vessels and their loading capacity can be estimated at over 600 kL/m3 with maximum capacity of 366 persons or average capacity of over 200 persons. The use of larger ships by the Ryukyu Kingdom during this period also indicates the expansion of maritime trade with the Ming. After c. AD 1520, however, the size of Ryukyuan-made ocean-going vessels reduced their average loading capacity and seldom exceeded 300 kL/m3, or a capacity of about 100 persons (Okamoto, 2008: 225–6), and such middle-sized vessels seemed to have continued to be built or used after the Early Modern Ryukyu Kingdom Period. During the Early Modern Ryukyu Kingdom period, the Kingdom also had started to purchase old Ming ships, as well as constructing similar ships itself (Tomiyama, 2012: 197–209). The Ryukyuan kai-sen and maran-sen share the same basic ship structure as the shinko-sen and they were used up to the 18th century for local trade by private merchants between Naha and Kagoshima, the capital of Satsuma Domain in the southern part of Kyusyu Island. The construction of maran-sen started in the early 18th century. The actual sizes of these Early Modern Ryukyuan ships are estimated by Tomiyama (2012) from the 19th-century drawings Senpaku Zumen Gomai ('Five ship plans': 船舶図面五枚) owned by Tokyo National Museum (2002). As shown in Table 3, shinko-sen and kai-sen are the largest with their hulls as long as 34.8 m, and 9.7 m in the beam, while maran-sen vessels are much smaller with a maximum hull length of only 14.2 m and beam of 5.4 m. Another historical record Ryukyu han zakki go, dated to 1873 (Minutes of Ryukyu Domain vol. 5: 琉球藩雑記五, originally edited by Ministry of Finance Japan in 1873 and re-edited by the Ryukyu Government in 1965), mentions the loading capacity of shinko-sen and kai-sen as 189 tonnes or 1260 koku (1 koku = 150 kg). The same document also mentions there were eight vessels classified as the largest sized ship of more than 1260 koku (= about 300 kL/m3or a capacity of about 120 persons) or with 15 tan (反) sized sail (1 tan = 991.74 m2). Among those noted in the document, three vessels are classified as shinko-sen and kai-sen types. Two are classified as Ryukyu-made kai-sen-type vessels. The privately owned ships, maran-sen, are more diverse in size. The loading capacity for the largest type is about 51 tonnes (340 koku), while the smallest type is about 14.4 tonnes (40 koku). Another section in the same document also mentions that there was a total of 12 of these smaller-sized vessels, with 7–12 tan sized sails, in 1873. Many much smaller vessels, with sails not exceeding 6 tan, also existed at this time: 77 with 6-tan, 24 with 5-tan, and nine with 4-tan-sized sails (Tomiyama, 2012: 28). No written documents that identify the anchor types that equipped a shinko-sen have been discovered as yet, but there are some drawings. On the folding screen showing the shinko-sen (進貢船の図) during 1830–1844 owned by the Okinawa Prefectural Museum and Art Museum, seven shinko-sen are depicted with their anchors (Figs 9a and b). These anchors are possibly much older Asian-style wooden anchors, rather than iron grapnels. The use of wooden anchors can be seen on other drawings including another folding screen depicting shinko-sen dated to the 19th century (owned by Okinawa Prefectural Museum) and the iconographic record of the voyage under the command of Captain Frederick William Beechey, which visited Ryukyu on May in 1827 (Beechey, 1831). The size of such wooden anchors used with shinko-sen has been estimated by Okinawa Prefectural Museum at 7.6 m in length and 3.62 m in maximum width between each arm (Fig. 10). One of the Ryukyuan kai-sen also appears in the historical document Ōshima zakki (Ōshima Records: 大島筆記) written by Tobe Yoshihiro (戸部良煕), which records that a Ryukyuan kai-sen departed Naha port, but failed in its course and drifted to Ōshima in Tosa Domain (present Kochi Prefecture in Shikoku Island) in 1762, was equipped with both wooden anchors and iron grapnel anchors. Since Ryukyuan kai-sen were often repurposed shinko-sen type ships, it is possible that shinko-sen were also equipped with both wooden anchors and iron grapnel anchors. Another well-known account, Zhongshan chuanxin lu (中山伝信禄) written by Xu Bao Guang (徐保光), who was the deputy envoy of a Chinese mission for the investiture of King Sho Kei and arrived in Ryukyu in 1719, also reports their Chinese-type vessels called hou-sen (封舟) that visited Ryukyu were equipped with both wooden and iron grapnel anchors, and includes a detailed illustration (Fig. 11). Since the Early Modern Ryukyuan ships were constructed following the Chinese shipbuilding tradition, it is presumed that the Ryukyuan ships' anchors were probably identical to the Early Modern Chinese ships' anchors. Although the historical documents and records concerning maran-sen and their anchors are very limited in number, the folding screen Ryukyu kōeki zu byōbufu depicts a maran-sen-type vessel with iron grapnel anchors (see Fig. 8b). Therefore, it is likely that all three types of Early Modern Ryukyuan ships were equipped with both wooden and iron grapnel anchors. Notably, ships from both China (Fujian) and Satsuma that visited the Ryukyu Island were equipped with iron grapnel anchors. According to another drawing of the old Naha port on a 19th-century folding screen (Fig. 12, owned by Okinawa Churashima Foundation Syurijyo Park), some Satsuma ships were also equipped with iron grapnel anchors (Katagiri, 2011). Since the Ryukyu Kingdom was dominated by Satsuma after AD 1609, there are some records that the Ryukyuan people chartered Satsuma ships to sail between Naha and Satsuma or Naha and the Yaeyama Islands (Okinawa Educational Committee, 2005). As shown in Figure 12, the Satsuma ships were depicted as Early Modern Japanese-type vessels in wood, also called kai-sen (廻船), which is different from the Ryukyuan kai-sen (櫂船) described above. The majority of kai-sen used in Satsuma and many other places in Japan were originally constructed as trading ships during the Early Modern times. Since the above-mentioned vessels from both Satsuma and China were equipped with iron grapnel anchors, the anchors found at Yarabuoki could be from either Japan or China. Iron grapnel anchors were originally invented in China possibly during the Ming Dynasty (Matsui, 2013). Some have been excavated in China: the oldest has a total length of 1.36 m and was found in Shandong Province in 1956. The anchor has an inscription on the upper shank that indicates that it was made in 1372 (Wang, 2000; see Fig. 13b). Two other grapnel anchors excavated in China are also as large: one excavated from Shandong Province in 1984 measures 2.15 m in length and weighs 456 kg (Fig. 13a), while another one from Fujian Province, excavated in 1981, is 2.68 m in length and weighs 758.3 kg (Fig. 13c). In Japan, the oldest historical documents depicting iron grapnel anchors are the Shinkō Kōgo Rokuengi Emaki ('Empress Jingū auspicious picture scroll' or 神功皇后縁起絵巻) of 1433 and the Boshi nyūminki ki (戊子入明記) of 1486 (re-edited by Kondo, 1912). Thus these iron anchors were possibly used in conjunction with wooden anchors and anchor stones until the early 15th century (Ishii, 1983: 322–6). Nevertheless, by the 17th-century iron grapnel anchors had become the main anchors used for later Japanese wooden ships such as kai-sen or bezai-sen during the late Edo Period of the Tokugawa Shogunate. Bezai-sen-type vessels that originated in the Setouchi Sea area, in western Japan, had become the major ship type in Early Modern Japan by AD 1700 (Fig. 14). One hokkoku-sen, one of the kai-sen (廻船) type ships of Northern Japan, drawn in 1633 on the votive picture dedicated to the Enkaku-ji Temple in present Aomori Prefecture, Northern Japan, was equipped with iron grapnel anchors (Fig. 15). Among bezai-sen-type ships, the common higaki kai-sen ships, which originated in the Osaka area, started to be used from 1619 for the export of various commercial goods from western Japan to Edo or other eastern parts of Japan (Tomiyama, 2012). In the 17th century there were approximately 1200 to 1300 vessels of this type in Japan (Kojima, 2012: 109). The larger bezai-sen is called sengoku-sen or sengoku-bune (sen = a thousand, goku/koku = basic unit of volume used during the Edo Period, -sen or bune/fune = ship) with loading capacity exceeding 278 kL/m3, a thousand koku. The average loading capacity for a sengoku-sen is around 1200 koku, which is almost the same size as a Ryukyuan shinko-sen. Sengoku-sen were usually equipped with seven different sized iron grapnel anchors identified as the first anchor to the seventh anchor. The largest was over 2 m in length and weighed 80 kan (= 300 kg), with each of the following weighing 5 kan (18.8 kg) less; thus, the seventh anchor weighed 50 kan (= about 180 kg) (Kojima, 2012: 119) (Table 4). This suggests that sengoku-sen (about 1200 koku) class vessels were equipped with anchors weighing from 180 to 300 kg. Much larger vessels (for example the 1500 koku class) would have been equipped with correspondingly larger and heavier anchors. For example, an iron grapnel anchor weighing about 340 kg with a total length of 2.80 m was salvaged from Kōzu Island underwater site, along with a number of Early Modern ink stones, stone lanterns and ceramics dated to AD 1800–1850. Since the Kōzu Island anchor weighs more than 300 kg, it is thought to be the first known anchor for a 1200–1500 koku-class vessel. Based on this information, the larger-sized grapnel anchors at Yarabuoki, which measure about 2 m in length, were possibly those of one or more sengoku-sen. Considering that the jars found at the site are tsuboya-yaki, Early Modern Okinawan ceramics originally produced in Naha City on Okinawa Island, the Yarabuoki grapnel anchors Nos 1, 2 and 3 could have equipped a Japanese-style kai-sen or Ryukyuan-style maran-sen vessel(s), probably from Satsuma or Naha. However, they could have also belonged to Chinese vessels making unintentional visits to Ishigaki Island when voyaging between Naha and Fujian and the other islands in the Yaeyama group. It is also known that some of the Early Modern Chinese vessels (mainly war ships from Fujian according to Dr Tsang, pers. com. 2014) were usually equipped with iron grapnel anchors. In fact, there are two local historic documents dated to 1685 that record the mishaps of possible Chinese vessels in the coastal sea off Yarabuzaki, the promontory close to Yarabuoki underwater site. The Shan-Sei Kafu Hedona Ke ('Genealogy Book of the Shang Clan, the Hedona Family': 尚姓家譜辺土名家), recounts that a small Chinese ship departed Naha port heading to China on 7 December 1685, with the intention of returning to Naha the following July, but was wrecked and sunk 16 July 1686 off the coast of Yarabuzaki, with the death of many of the crew members. Another document Gosei-Kafu Kudaka Ke ('Genealogy Book of the Wu Clan, the Kudaka Family' or 呉姓家譜 久高家) records that a small Chinese ship departed Naha on the 11 March and arrived in China on the 23 March in 1686. It departed again on the 16 July to return to Naha, but was wrecked around Yarabuzaki on the 23 July 1686. The account states that the ship's cargo was salvaged between 29 July and the 4 August, suggesting that the ship did not sink immediately. Both ships sunk off Yarabuzaki in July, possibly during the high typhoon season in 1686. These historical documents explicitly mention that Chinese ships sunk close to the Yarabuoki underwater site, thus the anchors could belong to Ryukyuan or Satsuma vessel(s), or Chinese vessel(s). More detailed analysis, including precise drawing of each anchor for reconstruction and comparative studies with other Early Modern iron anchors in China and Kyushu, particularly in Satsuma, may enable distinguishing features to be established. In order to clarify the characteristics of Yarabuoki underwater site and its artefacts, this section incorporates the results of our survey with the local historic and social context, particularly regarding the history of maritime transportation in the Ryukyu Kingdom. In the Yaeyama Islands including Ishigaki Island, there was no historically recorded trading port under the rule of the Ryukyu Kingdom (AD 1429–1879), and loading and disch