Marine Biology (2002) 141: 761–772 DOI 10.1007/s00227-002-0865-y C.L. Van Dover Trophic relationships among invertebrates at the Kairei hydrothermal vent field (Central Indian Ridge) Received: 17 January 2002 / Accepted: 26 April 2002 / Published online: 3 July 2002 Ó Springer-Verlag 2002 Abstract Exploration of hydrothermal vent systems in pic composition in vent communities from locations remote from well-studied sites allows ecolo- geographically disjunct oceanic regions. Given the as- gists to determine the degree of site-specific variation in sumptions associated with interpretations of isotopic trophic relationships among communities. A prelimi- data, there remains a missing pool of carbon (presum- nary outline of the trophic structure of the Kairei hy- ably unsampled bacterial biomass) that contributes to drothermal vent community on the Central Indian the maintenance of the 13C- and 15N-enriched primary Ridge (25°19.23¢S; 70°02.42¢E) is provided here, based consumers in these ecosystems. Electronic supplemen- on analysis of collections from an April 2001 expedi- tary material to this paper can be obtained by using the tion. Invertebrate biomass at Kairei is dominated by Springer LINK server located at http://dx.doi.org/ organic carbon with a d13C isotopic value of about – 10.1007/s00227-002-0865-y. 13&, due to the abundance of primary consumers (shrimp: Rimicaris aff. exoculata) and secondary con- sumers (anemones: Marianactis n. sp.) with this isotopic Introduction composition. Filamentous thiotrophic episymbionts on shrimp have been interpreted to be the major diet items More than two decades have passed since the initial of the shrimp and hence are the dominant primary discovery of invertebrate–bacterial trophic relationships producers within the community. Free-living auto- in hydrothermal-vent tubeworms (Cavanaugh et al. trophic microorganisms are implicated as the dietary 1981; Felbeck 1981). Despite the chemoautotrophic base base for other invertebrate species. Four trophic groups of vent food webs, trophic structure in vent communities are identified within the Kairei invertebrates based on is comparable in many ways to food webs of photo- carbon- and nitrogen-isotope ratios, but these groups synthetically based, shallow-water ecosystems (Van do not always define discrete trophic levels. Ontogenetic Dover 2000). In addition to free-living and symbiotic shifts in diet are documented for R. aff. exoculata and microbial primary producers, there are a variety of pri- brachyuran crabs (Austinograea n. sp.). Diets of sym- mary and secondary consumer types among vent inver- biont-bearing mussels (Bathymodiolus aff. brevior) and tebrate species, including grazers, suspension-feeders, two species of gastropods are isotopically constant deposit-feeders, predators, scavengers, and parasites. throughout the range of sizes analyzed. There is a Regional differences in vent food webs reflect the species consistent but unexplained pattern of increasing nitro- composition of the community and the functional nature gen isotopic composition with increasing carbon isoto- (thiotrophic and/or methanotrophic) of the primary producers. On the East Pacific Rise, for example, tube- worms, clams, and mussels are typical biomass domi- Communicated by J.P. Grassle, New Brunswick nants. They are host to endosymbiotic bacteria that Electronic supplementary material to this paper can be obtained by oxidize sulfide (i.e. are thiotrophic) to generate a cellular using the Springer LINK server located at http://dx.doi.org/ energy source for fixation of an inorganic carbon source 10.1007/s00227-002-0865-y. (CO2) into organic carbon. On the Mid-Atlantic Ridge, shrimp or mussels are typical biomass dominants. The C.L. Van Dover Biology Department, College of William & Mary, shrimp lack endosymbionts; instead they rely at least in Williamsburg, VA 23187, USA part on thiotrophic, filamentous episymbionts that grow E-mail: [email protected] in dense mats on the carapace that lines their branchial Tel.: +1-757-2212229 chambers (Gal’chenko et al. 1989; Casanova et al. 1993; Fax: +1-757-2216483 Polz et al. 1998). The Mid-Atlantic Ridge mussel species 762 (Bathymodiolus azoricus and B. puteoserpentis) host two Southward et al. 1994; Vereshchaka et al. 2000; Colac¸o types of autotrophic symbionts (thiotrophic and met- et al. 2002). Isotopic techniques have also been used to hanotrophic). While the biomass of East Pacific Rise infer the importance of free-living chemoautotrophic and Mid-Atlantic Ridge vent communities is dominated microorganisms in the diet of heterotrophic inverte- by organic material produced by symbiotic bacteria, brates within vent communities (Van Dover and Fry stable-isotope studies suggest that many of the primary 1994). and secondary consumer invertebrates associated with Interpretations of trophic interactions among Kairei biomass dominants rely on an alternative resource, in- invertebrates using isotopic compositions are based on ferred to be free-living autotrophic bacteria (Van Dover empirical and experimental evidence for fidelity between and Fry 1994). d13C values of diet items and direct consumers (Peterson Differences in the degree of dependence on methano- and Fry 1987), a 3–4& shift toward more-positive d15N trophic and thiotrophic microorganisms, such as those values from diet items to direct consumers (Minagawa observed between Mid-Atlantic Ridge and East Pacific and Wada 1984), and the potential for mixing of two or Rise vent systems, point to the potential for food-web more isotopically distinct dietary sources to yield inter- variations in other regions. Discovery of vent communi- mediate isotopic values of carbon and nitrogen in con- ties in the Indian Ocean at the Kairei hydrothermal field sumers (Fry and Sherr 1984). These assumptions are (Hashimoto et al. 2001) provided an opportunity to look valid only where isotopic compositions of inorganic re- for additional trophic novelties within vent ecosystems. sources do not vary significantly within sites and where Invertebrate biomass within the Kairei field is dominated species-specific fractionations of carbon and nitrogen by swarms of shrimp (Rimicaris aff. exoculata) on the isotopes are not anomalous. In this report, isotopic data surfaces of black-smoker chimneys (10–20°C) and by are used in conjunction with other information regard- anemone beds in a lower-temperature peripheral zone ing nutritional modes (including presence or absence of (1–2°C; Hashimoto et al. 2001; Van Dover et al. 2001). A symbionts, behavior, morphology of feeding structures variety of other invertebrates occur in a narrow transi- and digestive systems, and gut contents) to characterize tional zone (1–2 m) between the base of the shrimp a species as either a symbiont-bearing primary con- swarms and the proximal boundary of the anemone beds sumer, a non-symbiont-bearing primary consumer (e.g. (see Appendix 1 at http://dx.doi.org/10.1007/s00227-002- suspension feeder or grazer), or an omnivore. Strict 0865-y, electronic supplementary material, for a photo- predatory or scavenging behaviors within vent inverte- graphic overview of these faunal zones). Many of the brates are difficult to distinguish with confidence. Many species found at this Indian Ocean vent field have evolu- top invertebrate consumers are likely to have mixed diets tionary affinities to western Pacific vent faunas, but the that include microorganisms, primary consumers, and ecologically dominant shrimp closely resembles its other omnivores. Mid-Atlantic Ridge counterpart (Hashimoto et al. 2001; I identified isotopically distinct groups within the Van Dover et al. 2001). Kairei invertebrate community and inferred the dietary An outline of potential trophic interactions or pat- resources and trophic relationships of these species. For terns within a community can be obtained rapidly five Kairei species, I examined the relationship between through a survey of stable-isotope compositions. Stable- individual size and isotopic composition to determine isotope methods are not the only tool for assessment of and interpret shifts in diet during growth. Comparisons trophic relationships among species within a commu- of carbon and nitrogen isotopic ratios for invertebrates nity, and they are subject to site- and species-specific of Indian Ocean, Atlantic, and Pacific vents allow violation of some underlying assumptions regarding evaluation of trophic similarities and contrasts in bi- trophic shifts in isotopic composition. Nevertheless, ogeographically disparate vent systems. isotopic methods have proven valuable in generating or constraining hypotheses about trophic relationships in systems where direct observations of feeding interac- Materials and methods tions and samples are difficult to obtain, as at deep-sea hydrothermal vents (Conway et al. 1994). Carbon and Individuals of 15 invertebrate species (Table 1) were collected from nitrogen stable-isotope compositions have been used to the Kairei hydrothermal field (25°19.23¢S; 70°02.42¢E; 2415– 2460 m depth) on the Central Indian Ridge using the remotely advantage in a variety of marine settings (e.g. Rau and operated vehicle (R.O.V.) ‘‘Jason’’ in April 2001. Voucher speci- Hedges 1979; Rau et al. 1992; Van Dover et al. 1992; mens of undescribed species have been retained in an archival Wainright et al. 1993). Characterization of patterns of collection at The College of William & Mary. In situ observations isotopic compositions among individuals and among of invertebrate distributions and behavior were documented