Introduction
Hydrothermal vents are ‘hotspots’ of biology in the typically nutrient-poor deep ocean, powered by microbial chemosynthesis taking advantage of reducing chemicals in the hot vent fluid. The first hydrothermal vent in Japan, the JADE site of Izena Hole in Okinawa Trough, was discovered in 1988 by a joint German-Japanese cruise on-board R/V SONNE (Halbach et al., Reference Halbach, Nakamura, Wahsner, Lange, Sakai, Käselitz, Hansen, Yamano, Post, Prause, Seifert, Michaelis, Teichmann, Kinoshita, Märten, Ishibashi, Czerwinski and Blum1989). Since then, at least 13 active vent fields have been located in the region of the Ryukyu subduction system, mostly in the Okinawa Trough and some on the Ryukyu Arc (Nakamura et al., Reference Nakamura, Kawagucci, Kitada, Kumagai, Takai and Okino2015; Brunner et al., Reference Brunner, Chen, Giguère, Kawagucci, Tunnicliffe, Watanabe and Mitarai2022). These vents are all located in a basin to the west of Ryukyu Islands (Figure 1A) (Mitarai et al., Reference Mitarai, Watanabe, Nakajima, Shchepetkin and McWilliams2016). With only two narrow, deep openings (Tokara Valley and Kerama Gap) along the whole system, this basin is largely enclosed for larval dispersal below 600 m depth (Kizaki, Reference Kizaki1986). This has led to these vents together harbouring a unique set of specialist animal species distinct from other systems in the western Pacific (Brunner et al., Reference Brunner, Chen, Giguère, Kawagucci, Tunnicliffe, Watanabe and Mitarai2022).
The Amami Rift, previously known as the Amami Caldera, is located around 50 km northwest of Amami-Oshima, at the intersecting point between the volcanic front of the central Ryukyu Arc and rifting activity of the Okinawa Trough (Minami et al., Reference Minami, Saitou and Ohara2022). Surveys using ship-based and autonomous underwater vehicle (AUV) acoustic echo-sounding has revealed several rising gas plumes around 630 m depth, a typical feature of vent orifices in the Ryukyu region due to the rich gas content of their vent fluids (Nakamura et al., Reference Nakamura, Kawagucci, Kitada, Kumagai, Takai and Okino2015); further confirmed with smoke-like plumes visualised by the AUV's side-scan sonar imagery plus temperature anomalies (Minami et al., Reference Minami, Saitou and Ohara2022). Known active vent fields in the Ryukyu region mostly occur below 1000 m (Brunner et al., Reference Brunner, Chen, Giguère, Kawagucci, Tunnicliffe, Watanabe and Mitarai2022), making this a shallow site.
Depth is known to be a major factor driving biodiversity in the deep ocean, both directly by being an important physiological boundary for larval dispersal and settlement and also indirectly reflecting the influence of food availability, biotic interactions, among other factors (Rex and Etter, Reference Rex and Etter2010). The species compositions of chemosynthetic ecosystems around Japan appear to be strongly influenced by depth along with vent fluid chemistry and sediment cover (Fujikura et al., Reference Fujikura, Kojima, Fujiwara, Hashimoto and Okutani2000; Kojima, Reference Kojima2002; Nakajima et al., Reference Nakajima, Yamakita, Watanabe, Fujikura, Tanaka, Yamamoto and Shirayama2014; Watanabe and Kojima, Reference Watanabe, Kojima and Ishibashi2015), while other factors such as lava type have also been shown to be major drivers in other vents (Podowski et al., Reference Podowski, Ma, Luther, Wardrop and Fisher2010; Beinart et al., Reference Beinart, Sanders, Faure, Sylva, Lee, Becker, Gartman, Luther, Seewald, Fisher and Girguis2012). Only two Ryukyu vent fields at a similar depth to the Amami Rift acoustic anomalies have been reported with macrofaunal lists, including Minami-Ensei Knoll and Yoron Hole (Watanabe and Kojima, Reference Watanabe, Kojima and Ishibashi2015). The faunal composition of Minami-Ensei at 600–730 m deep has been found to differ from other deeper systems due to the presence of numerous species only found in this site (Hashimoto et al., Reference Hashimoto, Ohta, Fujikura and Miura1995; Okutani, Reference Okutani2001; Watanabe and Kojima, Reference Watanabe, Kojima and Ishibashi2015; Brunner et al., Reference Brunner, Chen, Giguère, Kawagucci, Tunnicliffe, Watanabe and Mitarai2022). Yoron Hole at 580 m deep is also unusual in having very low species richness (10 species) and without mussel assemblages (Watanabe and Kojima, Reference Watanabe, Kojima and Ishibashi2015). The composition of animal communities on the Amami Rift may therefore also be peculiar. Here, we carried out the first submersible dive at the Amami Rift, confirming the presence of an active hydrothermal vent field and report its associated animals. We also compare this with other known nearby shallow sites, and discuss implications in terms of vent biogeography in the Ryukyu region.
Materials and methods
From the acoustic water column anomalies (‘bubble plumes’) detected in the Amami Rift, we targeted a northern cluster (called ‘Area A’ in Minami et al., Reference Minami, Saitou and Ohara2022: figs. 2b, 4d) of signals for ground-truthing. During R/V YOKOSUKA cruise YK23-16S (15–29 September 2023), we conducted a dive using the human-occupied vehicle (HOV) SHINKAI 6500 on this area (dive #1726 on 19 September 2023; Figure 1A). A high-definition video camera (SONY FCB-H11, 1920 by 1080 pixels) and a digital still camera (Olympus E-PL6, 16 megapixels) mounted on the HOV SHINKAI 6500 were used for in situ seafloor imagery. A CTD (conductivity, temperature, and depth) sensor (Seabird SBE-19) on top of the submersible was used to record the ambient water temperature approximately 2–3 m from the seafloor. Temperature of focused flow venting was taken using the HOV SHINKAI 6500's temperature probe, while a RINKO III optical sensor (JFE Advantech Co., Ltd.) was used to measure the temperature of animal communities prior to sampling. Each temperature measurement was taken over a two-minute period. Prior to the dive, seafloor bathymetry was obtained using a multibeam echosounder (Kongsberg EM122) equipped on R/V YOKOSUKA and maps were drawn using the software Generic Mapping Tools (Wessel et al., Reference Wessel, Luis, Uieda, Scharroo, Wobbe, Smith and Tian2019).
Animals were collected using either a six-chambered suction sampler or directly collected using the submersible's manipulator into a bio box. Upon recovery on-board the research vessel, animal samples were immediately taken to a cold room (4°C). The animals were sorted, dissected, and identified morphologically under a stereomicroscope (Leica S9D or MS5). Specimens were fixed and preserved in 10% buffered formalin or directly in 99% ethanol. Macrophotography was carried out using digital single-lens reflex cameras including Canon EOS 5Ds R, Nikon D5600, and Olympus E-M1, and post-processed in Adobe Photoshop CC 2023.
Results
We successfully located a hydrothermally active area on the Amami Rift during HOV SHINKAI 6500 dive #1726 (Figures 1 & 2), presumably responsible for the cluster of acoustic anomalies detected in ‘Area A’ of Minami et al. (Reference Minami, Saitou and Ohara2022). In the earlier part of the dive (Figure 1B), we explored an area around the northernmost acoustic anomalies in Area A (28°36.4′N-28°36.5′N, 128°44.1′E-128°44.2′E) but did not locate any signs of venting. Then we headed towards more southernly points where potential plume signals have been detected, and encountered an area where several tubeworms including Lamellibrachia columna and Alaysia sp. (Figure 2H) were growing on rocks (28°36.2905′N, 128°44.1053′E, 625 m deep). We continued south but soon the tubeworms disappeared, and we turned north instead. Eventually, chains of gas bubbles and a small cluster of ‘Bathymodiolus’ japonicus mussels on pumice (28°36.341′N, 128°44.071′E, 626 m deep) appeared at approximately 100 m northwest of the point where we first saw tubeworms. From here the rocks became overgrown by more sponges, likely a result of increased organic input from the hydrothermal vent, combined with a higher density of tubeworms. Following these sponge-covered rocks, we soon discovered an active hydrothermal vent (Figure 2A) with dense aggregations of animals across an area of about 20 m by 30 m (28°36.360′N,128°44.093′E, 628 m deep). We here name this the Kyorasan (meaning ‘beautiful’ in Amami dialect) site of the Amami Rift hydrothermal vent field. The average bottom water temperature measured by the submersible before arrival at the Kyorasan site was 10.46°C (SD = 0.10), but was elevated to 11.27°C (SD = 0.17) during our stay at the hydrothermally influenced area.
The Kyorasan site lacks discrete chimney structure, and instead vigorous venting takes place directly from a patch of flat seafloor at the centre of activity (about 5 m by 10 m in area). Diffuse flow venting is seen across the entire patch likely from porous substrate (Figure 2B), while focused venting was limited to some fissures within the patch. The highest temperature of the focused venting from a fissure was measured to be 260°C. This region (Figure 2B) appears white in colouration due to both dense aggregations of white nests constructed by the sulphide worm Paralvinella aff. hessleri and bacteria mat on top of what appeared to be thin sulphur crust. Due to vent fluid being emitted from this relatively large area, the entire Kyorasan site is covered in a layer of shimmering water. Other than Paralvinella, the alvinocaridid shrimp Rimicaris leurokolos is the only species abundantly occurring in this central region, both species aggregating around fissures with focused venting.
Further away from the centre of activity (approximately 3–5 m) on weak diffuse flow, dense aggregations of bathymodioline mussels (both ‘Bathymodiolus’ japonicus and ‘B.’ platifrons) covers the surface of rocks (Figure 2C), with limpets (Bathyacmaea nipponica and Pyropelta cf. yamato) living on their shells. Temperature at the mussel aggregation was 12.37°C (SD = 0.43), slightly higher than the ambient temperature. The two bathymodioline mussel species generally tended to cluster separately, though in some aggregations they were mixed. The rocks occupied by the mussels were found to be sulphides, and this was the only area where sulphides were collected during this dive. The vent tonguefish Symphurus thermophilus lives in rather high density around the mussel colonies (Figure 2D) together with crabs (mostly a species of Trichopeltarion, occasionally Gandalfus yunohana). Aggregations of the alvinocaridid shrimp Alvinocaris dissimilis are also common around the mussels (Figure 2E), though many are also seen living inside the mussel clusters.
A third zone further away from the centre of activity (5–10 m) is dominated by the colloniid gastropod Cantrainea jamsteci (Figure 2F), with the sporadic occurrence of the mussel Bathymodiolus aduloides wedged between rock and sediment. The substrate in this third zone was mostly pumice boulders and rubbles scattered on a thin layer of sediment. No clear evidence of venting could be seen at this zone, though the posture of B. aduloides suggests weak diffuse flow is likely present underneath the rocks. Temperature of this zone was measured to be 11.46°C (SD = 0.01). In this zone we also sighted a chiton (Figure 2G) potentially in the genus Deshayesiella (Saito et al., Reference Saito, Fujikura and Tsuchida2008), though this was not collected and the identification is tentative at best without a specimen. Even further away, the tubeworms Lamellibrachia and Alaysia live on rocks (Figure 2H), but we saw few large individuals.
Sorting of the biological material collected revealed a total of 26 macrofaunal species (Figure 3), including 10 molluscs (three bivalves and seven gastropods), 10 annelids, four decapod crustaceans, one sponge, and one fish. Three species including the siboglinid tubeworm Alaysia sp., the chiton, and the bythograeid crab Gandalfus yunohana were seen but not collected, to make a total of 29 species. Among these, 19 are identified as species or genera considered specialists of chemosynthetic habitats around Japan (Fujikura et al., Reference Fujikura, Okutani and Maruyama2012; Brunner et al., Reference Brunner, Chen, Giguère, Kawagucci, Tunnicliffe, Watanabe and Mitarai2022), including: the bathymodioline mussels (‘B.’ japonicus, ‘B.’ platifrons, and B. aduloides); the gastropods Cantrainea jamsteci, Iheyaspira lequios, Puncturella parvinobilis, Bathyacmaea nipponica, Pyropelta cf. yamato, and Provanna clathrata; the shrimps Alvinocaris dissimilis, Rimicaris leurokolos, Lebbeus cf. shinkaiae; the crab Gandalfus yunohana; the tubeworms Lamellibrachia columna and Alaysia sp.; the sulphide worm Paralvinella aff. hessleri; the polynoid scale worm Thermiphione sp.; the amphinomid worm Archinome jasoni; and the tonguefish Symphurus thermophilus.
Discussion
The newly discovered Kyorasan site, Amami Rift hydrothermal vent field is among the shallowest deep-sea hot vent with known faunal composition in the Ryukyu region including the Okinawa Trough and Ryukyu Arc, along with Minami-Ensei Knoll and Yoron Hole (Fujikura et al., Reference Fujikura, Okutani and Maruyama2012; Brunner et al., Reference Brunner, Chen, Giguère, Kawagucci, Tunnicliffe, Watanabe and Mitarai2022). The overall zonation pattern of the Amami Rift field with Paralvinella sulphide worms and Rimicaris shrimps in the warmest region surrounded by bathymodioline mussel colonies and tubeworms in the outskirts is consistent with the overall pattern in Ryukyu vents (Watanabe and Kojima, Reference Watanabe, Kojima and Ishibashi2015; Yahagi et al., Reference Yahagi, Watanabe, Ishibashi and Kojima2015), though Shinkaia squat lobsters are missing. Compared to Yoron Hole at 580 m which only hosts 10 species and without a clear zonation pattern (Watanabe and Kojima, Reference Watanabe, Kojima and Ishibashi2015), the Amami Rift field is certainly more similar to Minami-Ensei and other Ryukyu vents. The occurrence of a zone densely colonised by the colloniid snail Cantrainea jamsteci is atypical for Ryukyu vents but has also been reported from the Minami-Ensei Knoll (Hashimoto et al., Reference Hashimoto, Ohta, Fujikura and Miura1995) at a comparable depth of 600–730 m, indicating this species may prefer shallower depths. The presence of vent-specialist species so far only known from sites shallower than 1000 m like the shrimp Alvinocaris dissimilis (Komai and Segonzac, Reference Komai and Segonzac2005; Methou et al., Reference Methou, Nye, Copley, Watanabe, Nagai and Chen2023) and the tonguefish Symphurus thermophilus (Tunnicliffe et al., Reference Tunnicliffe, Koop, Tyler and So2010) previously also recorded from Minami-Ensei suggests depth is indeed a factor shaping the diversity at the Amami Rift field.
The species composition of the Amami Rift field also exhibit several notable deviations from that of the Minami-Ensei Knoll (Hashimoto et al., Reference Hashimoto, Ohta, Fujikura and Miura1995). Most strikingly, Minami-Ensei Knoll lacks Paralvinella sulphide worms (Figure 3P; also present in Yoron Hole) on its chimneys and ‘Bathymodiolus’ platifrons (Figure 3E) in its dense mussel assemblages composed of only ‘B’. japonicus (Hashimoto et al., Reference Hashimoto, Ohta, Fujikura and Miura1995; Watanabe and Kojima, Reference Watanabe, Kojima and Ishibashi2015). Previously, ‘B’. platifrons has only been sampled from over 900 m depth (Fujikura et al., Reference Fujikura, Okutani and Maruyama2012) and indicated this could be due to depth partitioning – but as the Amami Rift is even shallower this cannot be the case and thus the reason for its absence in Minami-Ensei is unclear. The tubeworm Alaysia has also not been seen in Minami-Ensei, from which only Lamellibrachia was reported. The snail Iheyaspira lequios (Figure 3N), missing in Minami-Ensei, is present in the Amami Rift field, extending its bathymetric range by about 400 m (Fujikura et al., Reference Fujikura, Okutani and Maruyama2012). The vent-specific bythograeid crab Gandalfus yunohana is extremely rare in the Ryukyu region with just one definitive record (Watanabe et al., Reference Watanabe, Chen, Kojima, Kato and Yamamoto2020) before ours at Amami Rift, but we note that ‘Bythograeidae gen. sp.’ recorded in Minami-Ensei (Hashimoto et al., Reference Hashimoto, Ohta, Fujikura and Miura1995) likely also refer to this species.
Among chemosymbiotic species found in Minami-Ensei, the vesicomyid clam Akebiconcha kawamurai and the mussel Gigantidas cf. horikoshii (Hashimoto et al., Reference Hashimoto, Fujikura, Ohta and Miura1993) are conspicuously missing from the Amami Rift field, but this is likely because these are burrowing species requiring thick sandy or muddy sediments that are present in Minami-Ensei but not Amami Rift where only a thin layer of sediment was seen. The presence of such sediment in a number of vents in the Ryukyu region has been suggested as a major contributor to the relatively high species richness there (Watanabe and Kojima, Reference Watanabe, Kojima and Ishibashi2015) and is certainly a factor behind the composition seen in Amami Rift. The absence of the symbiotic squat lobster Shinkaia crosnieri in Amami Rift is more puzzling, since it usually co-occurs between Paralvinella and bathymodioline mussels. In Yoron Hole, not just S. crosnieri but the entire mussel assemblage and its associated species were also missing (Watanabe and Kojima, Reference Watanabe, Kojima and Ishibashi2015). This is not the case in Amami Rift. However, Lepetodrilus limpets which are considered to indicate the same zone as Shinkaia and also abundant in Minami-Ensei (Watanabe and Kojima, Reference Watanabe, Kojima and Ishibashi2015) is lacking in Amami Rift, indicating that environmental conditions of this zone may be simply missing in Amami Rift. This may reflect differences in the vent fluid composition since Minami-Ensei is known to have a high methane output (Chiba, Reference Chiba1993; Brunner et al., Reference Brunner, Chen, Giguère, Kawagucci, Tunnicliffe, Watanabe and Mitarai2022), and methane concentration has been identified as another important driver of species richness at chemosynthetic ecosystems around Japan (Nakajima et al., Reference Nakajima, Yamakita, Watanabe, Fujikura, Tanaka, Yamamoto and Shirayama2014).
The non-vent specific animals recovered from Amami Rift also clearly differ from those found in Minami-Ensei. A number of predators from surrounding normal deep-sea floor have been reported to be common in Minami-Ensei, including the anomuran crab Paralomis, the buccinid snail Neptunea insularis, and the asteroid Ceramaster misakiensis (Hashimoto et al., Reference Hashimoto, Ohta, Fujikura and Miura1995). Furthermore, the outskirts of Minami-Ensei was densely covered in bouquets of a branching sponge in the order Poecilosclerida (Fujikura et al., Reference Fujikura, Okutani and Maruyama2012). All of these are absent from Amami Rift. Instead, the trichopeltariid crab Trichopeltarion sp. (Figure 3L) was numerous in both bathymodioline mussel and Cantrainea snail assemblages. Trichopeltariids have not been recorded from hot vents, and our species is morphologically closest to Trichopeltarion janetae from seamounts and hydrocarbon seeps ranging from eastern New Zealand to Tasmania, Australia between 830–1700 m deep (Ahyong, Reference Ahyong2008; Tavares and Cleva, Reference Tavares and Cleva2010). The presence of T. janetae in several New Zealand seeps indicates members of this genus have some tolerance to chemosynthetic environments, and are able to invade them to take advantage of the increased food availability there. We did recover a species of buccinid snail (Calagrassor cf. aldermenensis; Figure 3D), but it is rare at the Amami Rift field unlike Neptunea which was common in Minami-Ensei (Hashimoto et al., Reference Hashimoto, Ohta, Fujikura and Miura1995). Notably, we also collected two species of sipunculan worms (Figure 3R, 3V), but as they were found in crevices of sulphide deposits and pumice, they may have been simply missed by previous sampling efforts in Minami-Ensei. The cause of these differences in non-vent specific fauna is unclear, but fluid chemistry or species composition of the surrounding non-vent seafloor may be key contributing factors.
Overall, our study reports a new hydrothermal field on Amami Rift whose zonation pattern of dominant species generally resemble those of other vents in the Ryukyu region, but with an unusual species composition. The species richness of 29 is high for a hydrothermal vent in the Ryukyu region and only surpassed by two others (Sakai and Iheya North vent fields) in this region (Brunner et al., Reference Brunner, Chen, Giguère, Kawagucci, Tunnicliffe, Watanabe and Mitarai2022). In addition to first records from vents (e.g., the crab genus Trichopeltarion), our records also represent bathymetric range extensions for many vent-specialist species (e.g., Iheyaspira lequios, ‘Bathymodiolus’ platifrons). The only three taxa shared across Minami-Ensei, Yoron Hole, and Amami Rift are the shrimps Rimicaris and Lebbeus as well as eunicid polychaete worms (Watanabe and Kojima, Reference Watanabe, Kojima and Ishibashi2015), exemplifying that each shallow Ryukyu vent field explored thus far have a unique set of macrofauna. Community analyses have repeatedly recognised shallow sites like Minami-Ensei as distinctive (Nakajima et al., Reference Nakajima, Yamakita, Watanabe, Fujikura, Tanaka, Yamamoto and Shirayama2014; Brunner et al., Reference Brunner, Chen, Giguère, Kawagucci, Tunnicliffe, Watanabe and Mitarai2022), which can also be expected of Amami Rift.
The discovery of the Amami Rift hydrothermal vent field paves the way for follow-up studies toward a better understanding of how disparate community structures arise in nearby shallow vents. Vent fluid composition may be a key driver, and that of Amami Rift could be distinct from other Ryukyu vents due to its unique geological background at the intersection of Ryukyu Arc and Okinawa Trough (Minami et al., Reference Minami, Saitou and Ohara2022). Biogeochemical characteristics of underwater volcanism in the Ryukyu region remains understudied (Shinjo and Kato, Reference Shinjo and Kato2000), and how lava type and depth influence vent fluid chemistry in this region requires further research. Ryukyu vents even shallower than Amami Rift have been hinted by turbidity, water chemistry, and dredges containing bathymodioline mussels between 275–300 m deep on Daiichi-Amami Knoll (Minami and Ohara, Reference Minami and Ohara2016; Wen et al., Reference Wen, Sano, Takahata, Tomonaga, Ishida, Tanaka, Kagoshima, Shirai, Ishibashi, Yokose, Tsunogai and Yang2016) only about 20 km northwest of Amami Rift field, an interesting target for exploration. Just south of Okinawa, the Kueishan Island (also known as Gueishandao or Turtle Island) site off northern Taiwan host very shallow vents between 15–323 m deep (Komai and Chan, Reference Komai and Chan2009; Wang et al., Reference Wang, Chan and Chan2014; Mellado et al., Reference Mellado, Zambrano, Aliaga, Ballesteros and Araya2022). While the sublittoral Kueishan vents comprise mainly non-vent endemics except the crab Xenograpsus testudinatus (Chen et al., Reference Chen, Chan and Chan2018), the upper bathyal (200–323 m deep) vents are dominated by vent endemics not found in any of the three shallow Ryukyu vents such as the mussel ‘Bathymodiolus’ taiwanensis, in addition to X. tetsudinatus (Wang et al., Reference Wang, Chan and Chan2014). An overarching study analysing the links between community structure, fluid chemistry, and underlying geology of these shallow vents in and around the Ryukyu region is warranted in the future.
Acknowledgements
We thank the captain and crew of R/V YOKOSUKA during research cruise YK23-16S for their support of our research activities, as well as the HOV SHINKAI 6500 team for their tireless work to ensure maximum success of the scientific dives. Kenichiro Tani (National Museum of Nature and Science, Japan) is gratefully acknowledged for his diligent leadership as the principle scientist on-board during the same cruise, and for useful discussions that improved this paper including suggesting the name Kyorasan. We extend our gratitude to the on-board scientists, especially Tatsuo Kanamaru (Nihon University) who was the dive scientist of dive #1726. We appreciate the help from Tomoyuki Komai (Natural History Museum and Institute, Chiba) and Pierre Methou (JAMSTEC / Ifremer) in identifying alvinocaridid shrimps.
Authors’ contributions
CC conceived and designed the study. CC, HN, and HK collected, sorted, and identified the macrofauna. CC interpreted the results and drafted the original manuscript which was edited by HN and HK. All authors agreed with the submission and publication of this manuscript in its present form.
Financial support
This study was supported by the Cooperative Research Program of Atmosphere and Ocean Research Institute, The University of Tokyo (R/V YOKOSUKA, cruise YK23-16S) and the integrated research project ‘Extreme Environments’ initiated by the National Museum of Nature and Science, Japan. NH was funded through a Japan Society for the Promotion of Science (JSPS) Postdoctoral Research Fellowship for Young Scientist (23KJ2222). This study was also partly supported by JSPS KAKENHI grants (Grant Numbers 20K06804 and 23H01278) to HK.
Competing interest
None.
Ethical standards
We have followed all applicable international, national, and/or institutional guidelines for the care and use of animals.
Data availability
All data generated or analysed during this study are included in this published article.