Limnol. Oceanogr., 44(2), 1999, 334–343

Lucky Strike mussels (Bathymodiolus sp.) support two metabolically distinct (methanotrophic and thiotrophic) prokaryotic endosymbionts in their gills. Differences in source inorganic carbon isotope ratios and in carbon fixation pathways between these two symbionts typically result in organic carbon with distinctive d13C values. Site-specific differences in isotopic compositions of host mussels may therefore reflect differences in sulfide and methane availability. Large differences in mean carbon and nitrogen isotopic compositions were observed in adult mussels collected from two chemically distinct vents at Lucky Strike (Sintra: d13C 5 221.3‰, d15N 5 24.5‰; Eiffel Tower: d13C 5 230.7‰, d15N 5 210.5‰). These values are consistent with the hypothesis that Sintra mussels are more dependent on methanotrophy than are Eiffel Tower mussels. Relative abundances of the two types of endosymbionts in mussel gill tissues (Sintra: mean 5 15% methanotrophs; Eiffel Tower: mean 5 6% methanotrophs) provide further support of this hypothesis, suggesting that Lucky Strike mussels express a nutritional response to environmental variations. Within sites, there were small but significant correlations between isotopic composition and mussel size over the shell lengths of 15–80 mm, but these shifts are so small that they are attributable to factors other than ontogenetic shifts in nutritional strategy. In contrast, large nitrogen isotope differences were observed between larval and adult stages. Based on d15N values, mussel larvae appear to rely very little on photosynthetically derived organic material. Observations of the demersal nature of mussel larvae and isotopic similarities between Sintra and Eiffel Tower larvae and Sintra adults suggest that the potential of stable isotopes as useful tracers of larval sources within the Lucky Strike vent field should be considered. The Lucky Strike hydrothermal vent field is dominated by dense beds of mussels belonging to an undescribed species of the genus Bathymodiolus (Craddock et al. 1995; Van Dover 1995; Van Dover et al. 1996; Langmuir et al. 1997). As in other vent communities, Lucky Strike trophic ecology is based on bacterial chemosynthesis rather than photosynthesis, and the mussels depend on internal bacterial symbionts as a significant source of nutrition (Van Dover et al. 1996). Whereas Pacific vent bathymodiolid mussel species usually house only sulfide-oxidizing symbionts (Fisher 1990), Lucky Strike and other Atlantic basin bathymodiolid species exhibit an intracellular dual symbiosis of both sulfide-oxidizing and methanotrophic symbionts (Fiala-Médioni et al. 1986; Cavanaugh et al. 1987, 1992; Fisher et al. 1993; Distel et al. 1995). Dual symbioses allow host mussels to exploit a wider range of chemical environments (Distel et al. 1995) and are assumed to be indicative of exposure to both sulfide and methane substrates over coevolutionary time scales. Because of the difference in chemical substrates used for chemosynthesis by the two symbiont types, a biological response to 1 To whom correspondence should be addressed. Present address: College of William & Mary, Biology Department, P.O. Box 8795, Williamsburg, Virginia 23187. Acknowledgments We thank Andy Bowen and the WHOI Deep-Submergence Operations Group for their skilled operation of Jason during the LUSTRE cruise. We are also grateful to our shipboard colleagues for assistance in collecting these animals, especially Chief Scientists Dan Fornari and Susan Humphris. Coast Seafoods generously donated mussel larvae. The manuscript was improved by the reviews of Ray Highsmith, Don Schell, Mike Simpkins, and two anonymous reviewers. This work was supported with NSF funds. heterogeneous substrate availability within a vent field may be reflected in the relative or absolute abundance of each symbiont species. Unlike other endosymbiont-dependent vent animals that have reduced digestive systems (e.g., adult vestimentiferans and pogonophorans lack mouth and gut; Jones 1981; Southward 1982), filter-feeding capability has been demonstrated in bathymodiolid mussels (e.g., Page et al. 1991). The importance of this heterotrophic nutritional pathway to these mussels has not yet been quantified. Heterotrophic uptake most likely includes some combination of free-living chemoautotrophic bacteria and sedimentary detritus (detritus ultimately derived from surface photosynthetic processes), although selective filtration is possible. Lucky Strike mussels thus have the potential to exploit thiotrophic, methanotrophic, or photosynthetic energy sources, and feeding strategies and the relative importance of nutritional sources may change during growth and development of these mussels. The symbiont–host relationship of at least one species of bathymodiolid mussel (Bathymodiolus thermophilus) is thought to be propagated by vertical transmission, with symbionts passing directly from adults to larvae in eggs (Cary and Giovannoni 1993). This mode of transmission presumably offers larval mussels immediate access to symbiotic energy sources. Histological analysis of the reproductive biology of B. thermophilus (Berg 1985) suggests that they exhibit planktotrophic (yolkless, feeding larvae) development, but it is unknown how long this developmental stage remains in the larval pool or from where their energy is derived. We chose to investigate whether Lucky Strike mussels use a consistent energy source throughout their life or if they switch trophic modes depending on resources available during development and growth. As an a priori model, we hy335 Lucky Strike mussel nutrition pothesized that larvae drift away from vent fields and rely on filter-feeding as their initial nutritional source, implicating a significant contribution of photosynthetically derived carbon. As postmetamorphic juveniles settle at vents, nutrition would switch quickly to reliance primarily on symbiotic chemoautotrophic energy sources. By this model, larvae would not be dependent on vent fluids to sustain symbiotic nutrition, and widespread dispersal could be routine. An alternative model suggests that larvae instead depend wholly on their symbionts or maternally derived reserves for energy. Unless the larvae can survive for long periods on stored resources, this strategy would presumably restrict larval dispersal primarily to within areas of active venting. A combination of these strategies is also possible, incorporating both facultative use of photosynthetic and symbiotic energy sources as resources become available. A first-order assessment of the relative importance of chemoautotrophic production in the nutrition of vent mussels throughout their life history can be made using stable isotope analyses. Differences in isotopic fractionation of carbon during metabolism result in distinct variations in isotope ratios of photosynthetically and chemosynthetically derived organic carbon. Carbon isotope values are relatively conservative in trophic interactions—marine food webs typically express ,1‰ increases in d13C per trophic level (DeNiro and Epstein 1978). This carbon isotope fidelity between trophic levels allows one to implicate specific food resources in the diet of a consumer when putative foods are isotopically distinct. Differences in nitrogen isotope ratios can help resolve trophic relationships because there is typically a 2–4‰ increase in d15N per trophic level (Minagawa and Wada 1984; Rau 1985). Given large differences in the isotopic compositions of putative nutritional resources of Lucky Strike mussels, we undertook a survey approach to the study of their nutrition, examining both sizeand site-related differences in isotopic composition. We also analyzed the mussel commensal polychaete Branchipolynoe seepensis to explore the fidelity of commensal and host isotopic composition between sites. Significant differences in stable isotope ratios between mussel populations at Lucky Strike suggested site-specific differences in endosymbiont associations. To assess the degree of variation in relative abundance of methanotrophic and sulfide-oxidizing symbionts in the Lucky Strike mussel population, we used transmission electron microscopy to determine relative abundances of gill-housed symbionts at each site.