Invertebrate Biology I21(1): 47-54. 0 2002 American Microscopical Society, Inc.

Methane-based symbiosis in a , platifrons, from cold seeps in Sagami Bay, Japan

James P. Barry,'," Kurt R. Buck,' Randall K. Kochevar,',b Douglas C. Nelson,2 Yoshihiro Fujiwara,3 Shana K. Goffredi,' and Jun Hashimoto3

IMonterey Bay Aquarium Research Institute, Moss Landing, California 9.5039, USA 2University of California, Division of Biological Sciences, Davis, California 9.5616, USA "Japan Marine Science and Technology Center, Yokosuka 237, Japan

Abstract. Bathymodiolus platifrons, a chemosynthetic mussel from cold seeps off Japan, relies for its nutrition on the productivity of methylotrophic or methanotrophic endosymbionts. High densities of bacterial symbionts appearing to be type I methanotrophs were observed in trans- mission electron micrographs of gill tissues. Methanol dehydrogenase activity in gill tissue from a single individual was positive compared to non-methanotrophic control samples, indi- cating a high potential for methanotrophy. Stable isotopic ratios of carbon in symbiont-con- taining gill tissue, as well as host tissues, were extremely depleted in IT, and similar to values reported for other methanotrophic species. TEMs of gill tissue showing symbionts in various stages of digestion support the hypothesis that carbon transfer from symbionts to B. platifrons occurs through intracellular digestion of the symbionts. Discovery of methane- or methanol- based symbioses in B. platifrons from cold seeps in Sagami Bay extends the range of such symbioses to include cold seeps and hydrothermal vents, and supports the idea that environ- mental methane levels control the distribution of these symbioses.

Additional key words: chemosynthesis, methanotrophy, methylotrophy, bivalve, biogeography

In chemoautotrophic bacterial-invertebrate symbio- ventional feeding and digestive structures (Fisher ses, the receive much or all of their nutrition 1990; Childress & Fisher 1992; Nelson & Fisher 1995; froin the symbiotic bacteria, while the bacteria pre- Scott & Fisher 1995). sumably benefit from a protected and stable physical Although many reduced compounds are thought and chemical environment favorable for carbon fixa- suitable to support carbon fixation by bacterial sym- tion. Such symbioses span various invertebrate phyla bionts, sulfide oxidation is the basis for most known and are common among molluscs, especially bivalves chemosynthetic symbioses (Cavanaugh 1985). The inhabiting hydrothermal vents, cold seeps, and other predominance of sulfur-based chemosynthesis in in- sub-oxic sediments (Van Dover 2000 and references vertebrate-bacterial symbioses may be related to sev- therein). Morphological and physiological specializa- eral factors. Some invertebrate hosts are able to bind tions for chemosynthetic symbioses vary greatly sulfide, enabling them to elevate internal sulfide levels among host species, ranging from those with epibiotic orders of magnitude above ambient environmental symbionts and few adaptations for nutritional integra- concentrations (Childress & Fisher 1992). Sulfide tion (e.g., alvinellid polychaetes from hydrothermal binding serves several functions, including reduced vents) to species that lack feeding or digestive organs sulfide toxicity, storage of sulfide for use by symbionts and are extremely specialized to enhance and exploit as required, and partial physiological control over the carbon fixation by bacterial symbionts (e.g., vestimen- supply of sulfide to symbionts. tiferan worms). Molluscan hosts generally fall between Conversely, methane, though energetically compa- these extremes, having enlarged gills containing high rable to sulfide in oxidation potential (Anthony 1982; densities of endosymbiotic bacteria, and reduced con- Nelson & Hagen 1995), is less frequently the basis for chemosynthesis in invertebrate-bacterial symbioses. Author for correspondence. E-mail: barry @mbari.org Because metazoans are unable to bind and store meth- Present address: Monterey Bay Aquarium, Monterey, Cal- ane, species with endosymbiotic methanotrophic bac- ifornia, 93940, USA teria may thrive only in environments with stable and 48 Barry et al. elevated methane levels. Methanotrophic species, ob- and have been described in detail by Hashimoto et al. ligate C, oxidizers that use methane as both a carbon (1989) and Tsunogai et al. (1996). Fluid seepage at source and an electron donor, are capable of oxidizing these sites supports a diverse assemblage of chemo- methane to methanol (catalyzed by methane mono- synthetic fauna dominated by vesicomyid clams oxygenase), the first step in methane-based carbon as- (mainly Culyptogenu soyoae), mytilid (Buth- similation. Methylotrophs do not oxidize methane, but ymodiolus platifions and other congeners), vestimen- can assimilate oxidized forms of methane such as tiferan worms (Lamellibruchia sp. and Escarpia sp.), methanol and other C, compounds, or other multi-car- and other less abundant chemoautotrophic species as bon organic compounds (Anthony 1982). Because all well as numerous non-chemosynthetic benthic mega- sites where metazoans with metlianotrophic or meth- fauna. Mytilid mussels are found commonly among ylotrophic endosymbionts have been reported also rocky outcrops with partial sediment cover also inhab- have high environmental methane levels-which very ited by vestimentiferan worms, near sedimentary hab- likely represent the energy basis for chemosynthesis in itats with aggregations of vesicomyid clams. these systems-we use methane-based symbioses to refer to invertebrate-bacterial symbioses involving Collection of specimens methanotrophic, or possibly methylotrophic, symbi- onts. Specimens of Bathyrnodiolus platifrons HASHIMOTO Methanotrophic invertebrate-bacterial symbioses & OKUTANI1994, Culyptogena soynae OKUTAN~1957, have been reported from the Atlantic Ocean for several and associated cold-seep fauna were collected from the species of mytilid mussels in the genus Buthymodiolus Hatsushiina cold seeps during dives of the Shinkai from methane-rich environments (Gulf of Mexico, 2000 manned submersible operated by the Japan Ma- Childress et al. 1986; Fisher et al. 1987; Barbados ac- rine Science and Technology Center (JAMSTEC). cretionary complex, Olu et al. 1996; Mid-Atlantic Specimens were collected from mussel aggregations Ridge, Cavanaugh et al. 1992), a sponge-bacterial on rock outcrops, placed in a sample basket, and cov- symbiosis in the Barbados accretionary complex (Va- ered with sediment to insulate the specimens from celet et al. 1995, 1996), and a pogonophoran (Sibog- warm surface temperatures during recovery of the sub- linum poseidoni) in the Skagerrak (Schmaljohann & marine. Following recovery, several individuals of R. Fliigel 1987). In the Pacific Ocean basin, methano- platifrons and C. soyoue were frozen (-80°C) imme- trophic symbionts have been reported recently from B. diately. Additional specimens of both species were platifrons and B. japonicus from hydrothermal vents preserved for microscopy. in the Okinawa Trough (Fujikura et al. 2000), and are possibly present in a snail (Olgaconchn tujuri) from Electron microscopy thermal springs in the Manus Basin (Gal'chenko et al. Samples of excised gill tissue from B. platifrons and 1992). Methodological limitations concerning the anal- C. soyoue were prepared for ultrastructural and mor- ysis of 0. tufari indicate that additional study is re- phological studies by preservation in cold 0.1 M cac- quired to determine whether this species has methan- odylate-buffered glutaraldehyde (2%) in filtered sea- otrophic symbionts. water. Samples for transmission electron microscopy In this paper, we document a methanotrophic or (TEM) were also dehydrated and infiltrated using mi- methylotrophic bivalve-bacterial symbiosis in the mus- crowave techniques (Giberson et al. 1997). sel B. platifrons from the Hatsushima cold seeps in Sagami Bay, Japan. Evidence is presented demonstrat- Methanol dehydrogenase activity ing the presence of methylotrophic or methanotrophic symbionts in gill tissues, activity of methanol dehy- Methanol dehydrogenase (MDH) catalyzes the oxi- drogenase (MDH), a key enzyme in methane-oxidizing dation of methanol to formaldehyde and is a key en- metabolic pathways, lysosomal features suggesting the zyme in all methanotrophic and methylotrophic organ- dominant mode of carbon transfer from symbiont to isms (Anthony 1982). Although the presence of host, and additional information indicating the nutri- methane mono-oxygenase, which catalyzes the con- tional reliance of B. plutijirons on methane-based car- version of methane to methanol, is a much more direct bon fixation. indication of methanotrophy, this enzyme degrades Methods rapidly following the death of the individual and must be analyzed quickly (Fisher 1990). Because our anal- Study site-Hatsushima cold seeps yses relied on frozen specimens, we were not able to The Hatsushima cold seeps are located at a depth of discriminate between methanotrophy and methylotro- 1100 m in Sagami Bay, Japan (34"59.9'N, 139'13.6'E) phy with this technique. Methane-based symbiosis in a mussel 49

Activity of MDH in tissues of B. platifrons was as- graded cells, and finally, aggregations of membranous sayed from hoinogenates of frozen gill tissues. Ap- elements is evident in some TEM images. Secondary proximately 0.5 g of gill tissue from a single individual lysosomes, vacuoles filled with aggregations of of B. platifrons was suspended in 2.5 ml of chilled whorled membranes or filaments, were most abundant phosphate buffer (0.05 M, pH 6.0), ground in an ice- adjacent to blood vessels. High densities of intact, ap- cold glass tissue grinder, and passed through a French parently healthy symbionts of various sizes were com- pressure cell (15,000 psi). The supernatant (10,000 X mon along the ciliated cell margin of bacteriocytes, g, 5 min) was assayed for MDH activity at 5°C by the where water flow would presumably provide locally method of Anthony & Zatman (1964). This procedure high methane levels. This region of presumed symbi- was repeated using homogenates from 3 individuals of ont development and growth contrasts with the margin Calyptogena kilmeri BERNARD1974, a chemoautotro- of bacteriocytes adjacent to blood vessels, where mem- phic clam that is thiotrophic, rather than methano- branous material and degraded symbionts were most trophic or methylotrophic, and thus should have no common. Several stages of cell division within bac- MDH activity and is a suitable negative control. Cal- teriocytes, indicating in situ growth, were also ob- culated rates are reported after subtracting the rate ob- served. served in the assay mixture without methanol. As a TEMs of gill tissue of C. soyoue revealed high den- positive control, a species of Hyphomicrobium grown sities of sulfide-oxidizing bacteria. anaerobically on a methanolhitrate medium at 30°C was assayed and found to have MDH activity of 95 Methanol dehydrogenase activity nmol min-' mg-' protein, within the expected range. MDH activity for the single individual tested was Protein was assayed by the Coomassie brilliant blue positive compared to control tissues, indicating that dye binding technique as described previously (Nelson methanotrophic or methylotrophic symbionts, or both, et al. 1989). were present in gill tissue. The MDH assay of gill tissue from B. platifrons yielded an activity of 3 nmol Tissue isotopic analyses min-' mg-' protein. MDH activity in gills of C. kil- Tissue samples of B. platifrons and C. soyoae were meri, which lacks methanotrophic or methylotrophic dissected from frozen specimens for analyses of stable symbionts, was below the detection limit (0.1 nmol carbon isotope ratios. After thawing to room temper- min-I mg-I protein) for all individuals assayed. ature, samples of gill, foot, adductor muscle, and man- tle tissues were dissected from individuals of each spe- Carbon isotope ratios cies, rinsed in filtered seawater, dried (60"C), and Carbon isotopes analyzed from several tissues of B. powdered in a mortar & pestle. Powdered samples platifrons were extremely depleted in "C (Table 1). were acidified with dilute HCl to remove inorganic Carbon was isotopically light in all tissues and was carbonate, and combusted in a Finnagen mass spec- within the range (613C <-50%~) expected for nutri- trometer (R. Dunbar, Stanford University). tional reliance on methanotrophic or methylotrophic bacteria (Fisher 1990). Compared to tissues of C. so- Results youe (8'T = -35.6%0), carbon in tissues of B. platif Morphology and cytology rons was considerably lighter isotopically, suggesting a difference in the source of organic carbon. Micrographs (TEM) prepared from hypertrophied gill tissue of Bathymodiolus platifrons (Fig. 1) re- Discussion vealed high densities of coccoid or rod-shaped bacte- Evidence for methane-based chemosynthesis rial symbionts with distinctive, stacked intracytoplas- mic membranes known only from type I Several sources of information support the hypoth- methanotrophic bacteria (Brock 1974). Electron lucent esis that the bacterial symbiosis of Bathymodiolus pla- inclusions in bacterial cells (Fig. lC,D) are identical tifrons in the Hatsushima cold seeps is based on meth- in appearance to poly-hydroxybutyrate-like inclusions ane oxidation. Symbionts with morphologies typical of reported from other methanotrophic bacteria (Anthony type I methanotrophs (Anthony 1982) were found in 1982). gills of B. platifrons. The morphology of these sym- The symbionts appear to represent several stages of bionts is indistinguishable from methanotrophic sym- degradation (Fig. lC,D). A gradient in cell types from bionts reported recently in individuals of B. platifrons robust, intact bacterial symbionts with highly detailed from hydrothermal vents in the Okinawa Trough (Fu- type I methanotroph-like ultrastructure, to partially de- jiwara et al. 2000), and is very similar to methylo- 50 Barry et al.

Fig. 1. Gill tissue from Bathymodiolus platifrons. TEM. A, B. Cross sections of gills showing symbiotic bacteria (S) in the bacteriocyte (B), intercalary cells (IC), their respective nuclei (N), microvilli (MV), cilia (C), and secondary lysosomal bodies (arrows). Water flow is above and blood vessel below. Scale bars, 2 pm. C, D. Detail of a symbiont (S) including stacked intracytoplasmic membranes, poly-hydroxybutarate inclusions (P), bacteriocyte mitochondria (M), and secondary lysosomal bodies (arrow) with residual membrane-like material. Scale bars, 1 pm. Methane-based symbiosis in a mussel 51

Table 1. Stable carbon isotopic values for tissues for B. MDH activity are partially responsible for the lower platifrons. Values for stable isotopic ratios are reported as activity in symbionts of B, p1atifron.s incubated near %O relative to PeeDee belemnite (PDB). Data for each of two the temperature at the Hatsushima cold seeps (-5Q specimens of B. platifrons are given separately, followed by to the Hyphomicrobium culture grown at the mean of the two values. 30°C. Assuming- a Qln~.~of 2.0 for enzyme activity (Kee- ton 1980), the rate estimated for symbionts of B. pla- Tissue SIT S'RC Mean SI3C tifrons at 30°C would be -17 nmol min-' mg-I pro- Foot -63.77 -62.68 -63.22 tein-closer, but still much lower than measured in the Adductor -64.01 -62.43 -63.22 Hyphomicrobium culture. Third, MDH activity is Mantle -67.64 -65.72 -66.68 scaled by protein content in the assay, which for tissue Gill -68.08 -67.54 -67.81 from B. platifrons includes protein from the symbionts and gill tissue, whereas the value for Hyphomicrobium is based on a pure bacterial culture. Because only the trophic or methanotrophic symbionts from species of bacteria are expected to express MDH activity, the Bathymodiolus in Atlantic seep and vent locations (Ca- protein-specific rate calculated for B. platifrons is con- vanaugh et al. 1992; Fisher & Childress 1992). As servative. with B. platifrons from the Okinawa Trough, no other We cannot, however, confirm whether symbionts of symbiont morphologies (i.e., sulfide-oxidizing bacte- B. platifrons from Sagami Bay are methanotrophic or ria, which lack stacked intracytoplasmic membranes) methylotrophic, even though conspecific mussels col- were observed. lected from hot vents in the Okinawa Trough, as well All species of chemosynthetic mussels studied, in- as several congeners from Atlantic sites have symbi- cluding some with both methane-based symbionts and onts with 16s ribosomal RNA sequences that cluster sulfide-oxidizing symbionts, have enlarged ctenidia phylogenetically among known methanotrophs (Distel and reduced digestive systems, representing adapta- et al. 1995; Robinson et al. 1998; Fujikura et al. 2000). tions for nutritional reliance on chemosynthesis rather Additional studies to measure methane uptake (e.g., than filter-feeding, even though the latter feeding mode Cary et al. 1989; Kochevar et al. 1992), coupled with remains possible for some species. The soft anatomy enzyme assays (methane mono-oxygenase activity), of B. platifrons, described in detail by Hashimoto & lipid characterization, and genetic studies could all be Okutani (1994), is similar to that of several congeners, used to discriminate between methylotrophic and with greatly thickened and elongated ctenidia, and methanotrophic pathways and determine whether B. only slight mantle fusion. The gut is short and straight, platifrons from vents and seeps are solely methano- and lacks a recurrent loop found in most other mytil- trophic. ids. Species of Bathymodiolus that feed on particulate The extremely light isotopic content of carbon in material have smaller and thinner ctenidia than B. pla- tissues of B. platifrons (W3C --65%0) and compara- tifrons and other methanotrophic mussels (Gustafson tively heavier carbon in tissues of sympatric vesico- et al. 1998), further supporting a chemosynthetic nu- myid clams (Calyptogena soyoae, 8I3C = -35.6%0) tritional mode for B. platifrons. indicates a carbon source other than CO, for the for- Although only a single individual of B. platifrons mer. Sulfide-oxidizing symbionts of vesicomyid clams was analyzed, the positive assay for methanol dehy- use CO, as a carbon source, which in the top 21 cm drogenase activity indicates the potential for oxidation of sediment at the Hatsushima cold seeps has an iso- of methanol to formaldehyde, an oxidative step found topic composition of S13C = -7 to -45%~(Masuzawa in both methanotrophs and methylotrophs. MDH ac- et al. 1992), with heaviest CO, near the sediment sur- tivity measured for B. platifrons (3 nmol min-' mg-l face. Because thiotrophic bacteria fractionate carbon protein) is within the range for bivalve gills known to isotopes by -25%0 (Ruby et al. 1987), carbon in tissue contain methylotrophic symbionts (Cavanaugh et al. of C. soyoae probably arose from a CO, source just 1992). The much higher MDH activity in a culture of below the sediment surface with values of 613C free-living methylotrophic bacteria (Hyphomicrobium, --l1%0. This carbon source could not yield the iso- 95 nmol min-' mg-' protein), compared to that mea- topic composition measured for tissues of B. platif- sured in B. platifrons, is not surprising and is consis- rons, even if this species had sulfide-oxidizing sym- tent with methylotrophy or methanotrophy in both spe- bionts. Isotopically light, biogenic methane, as found cies. First, MDH activity was detected in B. platifrons at the Hatsushima seeps (Hattori et al. 1996) is a more and Hyphomicrobium sp., but was, as expected, un- likely carbon source for B. platifrons. Carbon values detectable (<0.1 nmol min-l mg-' protein) in all in- from 613C = -90 to -60%0 are generally considered dividuals of C. kilmeri. Second, temperature effects on to originate from biogenic methanogenesis, while val- 52 Barry et al. ues from -50 to -30%0 arise from thermogenic pro- location of fixed radiocarbon from symbionts to non- cesses (reviewed by Martens et al. 1991). Tissues of symbiont containing tissues of Solemya reidi within B. platifrons (i3I3C = -6S%o) should have an isotopic several hours. Because carbon transfer via host diges- composition similar to its carbon source, because iso- tion of symbionts is expected to require much longer, topic fractionation by methanotrophic or methylo- they concluded that carbon transfer in S. reidi occurs trophic bacteria is considered minor (Fisher 1990) and through the direct release of metabolites from symbi- fractionation via heterotrophy (i.e., filter-feeding) is ont to host. Excretion of succinate and glutamate has very small (Parsons et al. 1984). Although its gill mor- been measured from purified symbiotic bacteria from phology is adapted for a chemosynthetic lifestyle (see Riftia pachyptila (Felbeck & Jarchow 1998). above), B. platifrons retains an intact, though reduced, digestive system, and may also filter-feed. Any nutri- Distribution of methanotrophic symbioses tional contribution from filtration must be minor, how- ever, since filtered material would include, in addition Congeners within the genus Bathyrnodiolus have to methanotrophic bacteria, isotopically heavier free- been highly successful in exploiting chemosynthetic living thiotrophic and heterotrophic bacteria. environments, including thiotrophic and methylotroph- ic or methanotrophic associations found in hydrother- Organic carbon transfer from symbiont to host mal vents and cold seeps. Only B. platifrons, however, is now known to rely on methane-based symbioses at Energy transfer between bacteria and host has been both cold seeps and hydrothermal vents. Of -10 investigated for few invertebrate-bacterial symbioses. known methane-based symbioses, only 2 involve spe- Transfer of fixed carbon is thought to occur by intra- cies outside the genus Bathynaodiolus: a sponge-bac- cellular digestion of symbionts in host bacteriocytes or terial symbiosis in the Barbados accretionary complex through translocation of dissolved metabolic products (Vacelet et al. 1995, 1996) and a pogonophoran, Si- from intact symbionts to the host. Digestion of sym- boglinum poseidoni, in the Skagerrak (Schmaljohann bionts may be a primary mechanism for carbon trans- & Flugel 1987). fer, or a consequence of “cellular housekeeping” in Reports of methane-based chemosynthetic symbio- which the host digests aging symbionts (Fisher 1990), ses are much more common from the Atlantic Ocean or both. The presence of primary and secondary ly- than the Pacific (see, e.g., review by Sibuet & Olu sosomes with accumulated membranous material from 1998). In a survey of the literature on metazoan-bac- symbionts in various stages of digestion (Fig. 1C) is terial symbioses, we counted 29 sulfide-based symbi- consistent with symbiont digestion as a mode of me- oses and 8 methane-based symbioses for the Atlantic, tabolite transfer. Similar ultrastructure has been re- and 30 and 2 symbioses, respectively, for the Pacific ported for invertebrate-bacterial symbioses from sev- (including this study). Is the putative rarity of metli- eral taxa (reviewed by Streams et al. 1997; Gustafson ane-based invertebrate-bacterial symbioses in the Pa- et al. 1998). Fisher & Childress (1992) and Streams et cific a result of geochemical limits to the evolution and al. (1997) measured uptake and assimilation of I4C- dispersal of such symbioses, or simply a consequence labeled methane by Bathyrnodiolus childressi, using of limited sampling effort? pulse/chase radiolabel assays. Methane assimilation in Although Barry et al. (1997) speculated that meth- B. childressi was rapid in gill tissues containing sym- anotrophic symbioses might be limited to seeps with bionts, but in tissue lacking symbionts, radiolabeled methane-rich brines associated with evaporite deposits carbon did not accumulate for several days, supporting that are more common in the Atlantic Ocean than the the hypothesis that carbon transfer occurred via sym- Pacific, these symbioses are now known from seeps biont digestion. Electron micrographs of gill tissues and vents in both ocean basins. In all cases, however, corroborated the carbon uptake measurements, show- methane levels of venting or seeping fluids are high. ing evidence of intracellular symbiont digestion simi- Fujikura et al. (2000) reported that methane-based spe- lar to that we observed for B. platifrons. Although cies were common only at sites with high (>2.6 mM) additional studies are required to determine the dom- ambient methane concentrations, compared to those inant mode of carbon transfer, ultrastructural evidence with thiotrophic or duel symbiont types, which were (Fig. lC,D) suggests that energy transfer occurs found in methane levels 10-100 times lower. None- through intracellular digestion of symbionts. theless, while all known methane-based symbioses are We cannot exclude the possibility that some carbon linked to high environmental methane, not all meth- transfer between B. platifrons and its symbionts occurs ane-rich sites have such symbioses, suggesting that through translocation of intermediate carbon fixation other factors may limit their distribution. For example, products. Fisher & Childress (1986) detected the trans- methane-based invertebrate-bacterial symbioses have Methane-based symbiosis in a mussel 53 not been observed near vents along the Juan de Fuca incubation media on the carbon transfer from the bacterial ridge in the Pacific where methane levels up to 3.9 symbionts to the tube-worm Riftia pu- mM have been measured (Lilley et al. 1993), or in chyptila. Cah. Biol. Mar. 39: 279-282. Monterey Bay where methane levels up to -8 mM Fisher CR 1990. Chemoautotrophic and methanotrophic have been measured. Broader exploration and inves- symbioses in marine invertebrates. Aquat. Sci. 2: 399- 436. tigation of the larval ecology and population dynamics Fisher CR & Childress JJ 1986. Translocation of fixed car- of seep and vent species, coupled with identification bon from symbiotic bacteria to host tissues in the gutless of environmental constraints, will likely modify cur- bivalve Solemya reidi. Mar. Biol. 93: 59-68. rent views of the evolution and maintenance of che- 1992. Organic carbon transfer from methanotrophic mosynthetic communities in the world ocean. symbionts to the host hydrocarbon-seep mussel. Symbio- sis 12: 221-235. Acknowledgments. We are grateful to the crew of the RIV Fisher CR, Childress JJ, Oremland RS, & Bidigare RR 1987. Natsushima and support staff and pilots of the submersible The importance of methane and thiosulfate in the metab- Shinkai 2000, and to V. Pearse and an anonymous reviewer olism of the bacterial symbionts of two deep-sea mussels. who provided very helpful comments that improved the Mar. Biol. 96: 59-7 1. manuscript. We acknowledge support from the Monterey Fujikura Y, Takai K, Uematsu K, Tsuchida S, Hunt JC, & Bay Aquarium Research Institute (MBARI) and the Japan Hashimoto J 2000. Phylogenetic characterization of en- Marine Science and Technology Center (JAMSTEC). dosymbionts in three hydrothermal vent mussels: Influ- ence on host distributions. Mar. Ecol. Prog. Ser. 208: 147- References 15.5. Gal'chenko VE Pimenov NV, Lein AY, Galkin SV, Miller Anthony C 1982. The Biochemistry of Methylotrophs. Ac- YM, & Ivanov MV 1992. Mixotrophic feeding of Olga- ademic Press, London. 431 pp. concha tufari Beck (Gastropoda: Prosobranchia) from the Anthony C & Zatman LJ 1964. The microbial oxidation of active hydrothermal field in the Manus Basin (Bismarck methanol. J. Biochem. 92: 6 14-621. Sea). Doklady Akademii Nauk. 323(4): 776-781. Barry JP, Kochevar RE, & Baxter CH 1997. The influence Giberson RT, Demaree RS Jr., & Nordhausen RW 1997. of pore-water chemistry and physiology on the distribu- Four-hour processing of clinical/diagnostic specimens for tion of vesicomyid clams at cold seeps in Monterey Bay: electron microscopy using microwave technique. J. Vet. Implications for patterns of chemosynthetic community Diag. Invest. 9: 61-67. organization. Limnol. Oceanogr. 42: 3 18-328. Gustafson RG, Turner RD, Lutz RA, & Vrijenhoek RC 1998. Bernard FR 1974. The genus Calyptctgena in British Colum- A new genus and five new species of mussels (, bia with a description of a new species (Bivalvia, Vesi- ) from deep-sea sulfidelhydrocarbon seeps in the comyidae). Venus 33: 11-22. Brock TD 1974. Biology of Microorganisms. Prentice-Hall, Gulf of Mexico. Malacology 40: 63-1 12. Hashimoto J & Okutani T 1994. Four new mytilid mussels Englewood Cliffs. 852 pp. Cary SC, Fisher CR, & Felbeck H 1989. Mussel growth associated with deep-sea chemosynthetic communities supported by methane as sole carbon and energy source. around Japan. Ven. Jap. J. of Malacol. 53: 61-83. Science 240: 78-80. Hashimoto J, Ohta S, Tanaka T, Hotta H, Matsuzawa S, & Cavanaugh CM 1985. Symbioses of chemoautotrophic bac- Sakai H 1989. Deep-sea communities dominated by the teria and marine invertebrates from hydrothermal vents giant clam, Culyptogena .soyoae, along the slope foot of and reducing sediments. Biol. Soc. Wash. Bull. 1985: 373- Hatsushima Island, Sagami Bay, central Japan. Palaeo- 388. geogr., Palaeoclimat., Palaeoecol. 7 1: 179-1 92. Cavanaugh CM, Wirsen CO, & Jannasch HW 1992. Evi- Hattori M, Kanie Y, Oba T, Akimoto K, & Kanno S 1996. dence for methylotrophic symbionts in a hydrothermal Environmental conditions of carbonates and chemosyn- vent mussel (Bivalvia, Mytilidae) from the mid-Atlantic thetic communities associated with cold seepage ridge. App. Env. Micro. 58: 3799-3803. zones along the subduction zone in Sagami Bay and other Childress JJ & Fisher CR 1992. The biology of hydrothermal land areas in Japan. JAMSTEC J. Deep-sea Rees. 12: vent animals: Physiology, biochemistry, and autotrophic 117-132. symbioses. Oceanogr. Mar. Biol. Ann. Rev. 30: 337-441. Keeton WT 1980. Biological Science, 3rd. edition. W.W. Childress JJ, Fisher CR, Brooks JM, Kennicutt MC 11, Bi- Norton & Company, New York. 11 3 1 pp. digare RR, & Anderson AE 1986. A methanotrophic nia- Kochevar RE, Childress JJ, Fisher CR, & Minnich E 1992. rine molluscan (Bivalvia, Mytilidae) symbiosis: Mussels The methane mussel: Roles of symbiont and host in the fueled by gas. Science 233: 1306. metabolic utilization of methane. Mar. Biol. 112: 389- Distel DL, Lee HK-W, & Cavanaugh CM 1995. Intracellular 401. coexistence of methano- and thioautotrophic bacteria in a Lilley MD, Butterfield DA, Olson EJ, Lupton JE, Macko hydrothermal vent mussel. Proc. Natl. Acad. Sci. 92: SA, & McDuff RE 1993. Anomalous CH, and NH,' con- 9598-9602. centrations at an unsedimented mid-ocean-ridge hydro- Felbeck H & Jarchow J 1998. The influence of different thermal system. Nature 364(6432): 45-47. 54 Barry et al.

Martens CS, Chanton JP, & Paul1 CK 1991. Biogenic meth- pentis (Bivalvia: Mytilidae), a dual endosymbiotic mussel ane from abyssal brine seeps at the base of the Florida from the mid-Atlantic ridge. Mar. Biol. 132: 625-633. escarpment. Geol. 19: 85 1-854. Ruby EG, Jannasch HW, & Deuser WG 1987. Fractionation Masuzawa T, Handa N, Kitagawa H, & Kusakabe M 1992. of stable carbon isotopes during chemoautotrophic growth Sulfate reduction using methane in sediments beneath a of sulfur-oxidizing bacteria. Appl. Envir. Micro. 53: 1940- bathyal “” giant clam community off Hatsushi- 1943. ma Island, Sagami Bay, Japan. Earth Planet. Sci. Lett. 110: Schmaljohann R & Fliigel HJ 1987. Methane-oxidizing bac- 39-50. teria in Pogonophora. Sarsia 72: 91. Nelson DC & Fisher CR 1995. Chemoautotrophic and meth- Scott KM & Fisher CR 1995. Physiological ecology of sul- anoautotrophic endosymbiotic bacteria at deep-sea vents fide metabolism in hydrothermal vent and cold seep ves- and seeps. In: The Microbiology of Deep-sea Hydrother- icomyid clams and vestimentiferan tube worms. Am. Zool. mal Vents. DM Karl, ed., pp. 125-167. CRC Press, Boca 35: 102-111. Raton. Sibuet M & Olu K 1998. Biogeography, biodiversity, and Nelson DC & Hagen KD 1995. Physiology and biochemistry fluid dependence of deep-sea cold-seep communities at ac- of symbiotic and free-living chemoautotrophic sulfur bac- tive and passive margins. Deep-sea Res. I 45: 517-567. teria. Am. Zool. 35: 91-101. Streams ME, Fisher CR, & Fiala-Mbdioni A 1997. Meth- Nelson DC, Wirsen CO, & Jannasch HW 1989. Character- anotrophic symbiont location and fate of carbon incorpo- ization of large, autotrophic Beggiatoa spp. abundant at rated from methane in a hydrocarbon seep mussel. Mar. hydrothermal vents of the Guaymas Basin. Appl. Envir. Biol. 129: 465-476. Micro. 55: 2909-2917. Tsunogai U, Ishibashi J, Wakita H, Gamo T, Masuzawa T, Okutani T 1957. Two new species of bivalves from the deep Nakatsuka T, Nojiri Y, & Nakamura T 1996. Freshwater water in Sagami Bay collected by the R.V. Soyo-Maru. seepage and pore water recycling on the seafloor: Sagami Bull. Tokai Reg. Fish. Res. Lab. 17: 27-30. Trough subduction zone, Japan. Earth and Planet. Sci. Let. Olu K, Sibuet M, Harmegnies E Foucher J-P, & Fiala-MC- 138: 157-168. dioni A 1996. Spatial distribution of diverse cold seep Vacelet J, Boury-Esnault N, Fiala-MCdioni A, & Fisher CR communities living on various diapiric structures of the 1995. A methanotrophic carnivorous sponge. Nature 377: southern Barbados prism. Prog. Oceanogr. 38: 347-376. 296. Parsons TR, Takahashi M, & Hargrave B 1984. Biological Vacelet J, Fiala-MCdioni A, Fisher CR, & Boury-Esnault N Oceanographic Processes, 3rd. edition. Pergamon Press, 1996. Symbiosis between methane-oxidizing bacteria and Oxford. 330 pp. a deep-sea carnivorous cladorhizid sponge. Mar. Ecol. Robinson JJ, Polz ME Fiala-MBdioni A, & Cavanaugh CM Prog. Ser. 145: 77-85. 1998. Physiological and immunological evidence for two Van Dover CL 2000. The Ecology of Deep-sea Hydrother- distinct C,-utilizing pathways in Bathymodiolus puteoser- mal Vents. Princeton Univ. Press, Princeton. 424 pp.