sign in sign up
<<
Home , Hydrothermal vent, Riftia pachyptila

This article has been published inOceanography , Volume 20, Number 1, a quarterly journal of The Soci- S p e c i a l i ss u e F e at u r e ety. Copyright 2007 by The Oceanography Society. All rights reserved. Permission is granted to copy this article for use in teaching and research. Republication, systemmatic reproduction, or collective redistirbution of any portion of this article by photocopy machine, reposting, or other means is permitted only with the approval of The Oceanography Society. Send all correspondence to: [email protected] or Th e Oceanography Society, PO Box 1931, Rockville, MD 20849-1931, USA. and of Thirty Years of Discovery and Investigations

B y e Va R a m i r e z- L l o d r a , On the oor, Dierent Species T i m o t h y M . S h a n k , a nd Thrive in Dierent Regions c h r i s to p h e r R . G e r m a n Soon after communities were discovered around seafl oor hydrothermal vents in 1977, sci- entists found that vents in various regions are populated by distinct animal species. Scien- Shallow Atlantic vents (800-1700-meter depths) tists have been sorting clues to explain how support dense clusters of The discovery of hydrothermal vents and the unique, often endem- seafl oor populations are related and how on black smoker chimneys. they evolved and diverged over Earth’s • ic fauna that inhabit them represents one of the most extraordinary history. Scientists today recognize dis- scientific discoveries of the latter twentieth century. Not surprising- tinct assemblages of animal species in six major seafl oor regions (colored ly, after just 30 years of study of these remarkable—and extremely dots) along the system of volcanic remote—systems, advances in understanding the and mi- mountains and deep-sea trenches that form the borders of Earth’s crobial communities living around hydrothermal vents seem to tectonic plates. But unexplored Deep Atlantic vent com- munities (2500-3650-me- occur with every fresh expedition to the seafloor. On average, two regions remain critical ter depths) are dominated by missing pieces for assembling swarms of shrimp called Rimica- new species are described each month—a rate of discovery that has the full evolutionary puzzle. •ris exoculata. been sustained over the past 25–30 years (Van Dover et al., 2002; Fisher et al., this issue). Furthermore, the physical, geological, and geochemical features of each part of the ridge system and its asso- ciated hydrothermal-vent structures appear to dictate which nov-

el biological species can live where. Only 10 percent of the ridge Central Indian vent com- Northeast Pacifi c vent munities are populated by system has been explored for hydrothermal activity to date (Baker communities are domi- Western Pacifi c-type fauna, but and German, 2004), yet we find different diversity patterns in that nated by “bushes” of skinny tube- also have North Atlantic-type worms called Ridgea piscesae. •shrimp species. small fraction. While it is well known that species composition var- • ies along discrete segments of the global ridge system, this “bio- geographic puzzle” has more pieces missing than pieces in place (Figure 1, Table 1). In this paper, we start with a general picture of the global ridge Western Pacifi c vent com- Eastern Pacifi c vent com- system and hydrothermal environments, then continue with a de- munities are dominated by munities are dominated barnacles and , as well as by tall, fat tubeworms called scription of known biodiversity—including physiological adapta- •hairy gastropods, shown above. E. Paul Oberlander •. tions—and , which into a discussion of the factors that might drive observed biogeography patterns. We Chile Rise Southern Ocean South Atlantic Caribbean Arctic Ocean Unusual forms may have This region has a full spectrum of This region has a variety of The Drake Passage may act as a Powerful currents and huge In this region, seeping The Arctic Ocean has never conclude with a look toward the future, describing the main efforts evolved under conditions of supporting seafl oor life chemosynthetic habitats and key link or bottleneck for larval seafl oor chasms (fracture zones) from the seafl oor also supports had deep connections with Missing extreme pressure in this 11,000- (hydrothermal vents, cold seeps, geological features in close prox- dispersal between the Atlantic may act as barriers blocking the animal communities. Did other major . It may being made to fill the gaps in our knowledge of vent biogeography. Pieces meter-deep trench, the deepest carcasses, and wood from imity. How do seafl oor popula- and Pacifi c. Whale carcasses and dispersal of vent larvae and dis- animals migrate between “cold harbor fundamentally part of the world’s oceans. shipwrecks and trees) in close tions diverge or converge at this shipwrecks (such as Shackleton’s connecting vent populations in seeps” and nearby hot vents over different vent animals that proximity. How have species triple junction on the “highway” Endurance) may offer refuges or the North and South Atlantic. evolutionary history? evolved in isolation over the evolved in these diverse settings? of mid-ocean ridges? stepping-stones between vents. past 25 million years.

30 Oceanography Vol. 20, No. 1 Figure 1. Schematic illustration of the global mid-ocean ridge system. Scientists today recognize distinct assemblages of animal species in six major seafloor regions (colored dots) along the system of volca- nic mountains and deep-sea trenches that form the borders of Earth’s tectonic plates. However, unexplored ocean regions remain critical missing pieces for assembling the full evolutionary puzzle. From Shank (2004)

On the Sea oor, Dierent Species Thrive in Dierent Regions Soon after animal communities were discovered around seafl oor hydrothermal vents in 1977, sci- entists found that vents in various regions are populated by distinct animal species. Scien- Shallow Atlantic vents (800-1700-meter depths) tists have been sorting clues to explain how support dense clusters of mussels seafl oor populations are related and how on black smoker chimneys. they evolved and diverged over Earth’s • history. Scientists today recognize dis- tinct assemblages of animal species in six major seafl oor regions (colored dots) along the system of volcanic mountains and deep-sea trenches that form the borders of Earth’s tectonic plates. But unexplored Deep Atlantic vent com- munities (2500-3650-me- ocean regions remain critical ter depths) are dominated by missing pieces for assembling swarms of shrimp called Rimica- the full evolutionary puzzle. •ris exoculata.

Central Indian vent com- Northeast Pacifi c vent munities are populated by communities are domi- Western Pacifi c-type fauna, but nated by “bushes” of skinny tube- also have North Atlantic-type •worms called Ridgea piscesae. •shrimp species.

Western Pacifi c vent com- Eastern Pacifi c vent com- munities are dominated by munities are dominated barnacles and limpets, as well as by tall, fat tubeworms called •hairy gastropods, shown above. E. Paul Oberlander •Riftia pachyptila.

Challenger Deep New Zealand Chile Rise Southern Ocean South Atlantic Caribbean Arctic Ocean Unusual life forms may have This region has a full spectrum of This region has a variety of The Drake Passage may act as a Powerful currents and huge In this region, methane seeping The Arctic Ocean has never evolved under conditions of habitats supporting seafl oor life chemosynthetic habitats and key link or bottleneck for larval seafl oor chasms (fracture zones) from the seafl oor also supports had deep connections with Missing extreme pressure in this 11,000- (hydrothermal vents, cold seeps, geological features in close prox- dispersal between the Atlantic may act as barriers blocking the animal communities. Did other major oceans. It may Pieces meter-deep trench, the deepest whale carcasses, and wood from imity. How do seafl oor popula- and Pacifi c. Whale carcasses and dispersal of vent larvae and dis- animals migrate between “cold harbor fundamentally part of the world’s oceans. shipwrecks and trees) in close tions diverge or converge at this shipwrecks (such as Shackleton’s connecting vent populations in seeps” and nearby hot vents over different vent animals that proximity. How have species triple junction on the “highway” Endurance) may offer refuges or the North and South Atlantic. evolutionary history? evolved in isolation over the evolved in these diverse settings? of mid-ocean ridges? stepping-stones between vents. past 25 million years.

Oceanography March 2007 31 Abiotic Setup of the Spreading rates for mid-ocean of the largest and longest-lived hydro- Vent : Physical, ridges—the speed at which two tec- thermal vents discovered thus far occur Geological, and tonic plates are being pulled apart from along the slow-spreading Mid-Atlantic Geochemical Properties each other—vary across the whole sys- Ridge. In contrast, fast-spreading cen- Today, satellite technology provides a tem. Ridges can be classified accord- ters exhibit shallow and narrow (order clear visualization of the globe-encircling ing to their spreading rates into ultra- of tens of meters) summit calderas mid-ocean ridge. It is here, at seafloor- slow spreading (< 20 mm yr-1) such where hydrothermal plumes are not spreading centers, that lava wells up from as the Arctic (see Snow constrained. Along the fastest sections Earth’s interior to generate fresh oce- and Edmonds, this issue), slow spread- of the , vent sites are so anic crust (see Langmuir and Forsyth, ing (20–50 mm yr-1) such as the Mid- geographically close that they can form this issue). While a small proportion Atlantic Ridge, intermediate spreading a continuum along the ridge axis. These of these volcanic systems are found in (50–90 mm yr-1) such as the Central topographic and physical characteristics back-arc basins close to Indian Ridge, and fast (90–130 mm yr-1) have the potential to affect the patterns zones (see Martinez et al., this issue), the to superfast (130–170 mm yr-1) spread- of faunal distribution (see below). The vast majority of all Earth’s mid-ocean ing such as the East Pacific Rise. Spread- western Pacific Ocean seafloor is char- ridges form a single, continuous, globe- ing rates are important in shaping ridge acterized by a complex system of back- encircling volcanic chain that is roughly morphology (MacDonald et al., 1991) arc basins and volcanic arcs with active 60,000 km in length. Mid-ocean ridges and, as a consequence, the hydrodynam- hydrothermal venting. These back-arc typically lie at around 2000- to 5000-m ics of the habitat. At slow-spreading basins are isolated spreading centers that depth and therefore represent environ- rates, the walls of the ridge axis are have been active for less than 10 million ments that experience high pressures, separated by a deep (1–3 km) and wide years—a short geological time com- complete darkness, and ambient tem- (5–15 km) valley that constrains pared to the ages of mid-ocean ridges peratures of only ~ 2°C. dispersing hydrothermal plumes. Some (Hessler and Lonsdale, 1991). Their iso-

Table 1. Dominant fauna of main proposed biogeographical provinces.

Biogeographical Province and Depth Dominating Fauna Azores (shallow north Atlantic, 800-1700 m) Bathymodiolid mussels, amphipods, and caridean shrimp Mid-Atlantic Ridge between Azores Triple Junction Caridean shrimp—mainly Rimicaris exoculata—and bathymodiolid mussels and Equator (deep north Atlantic 2500-3650 m) South Mid-Atlantic Ridge Caridean shrimp, bathymodiolid mussels, and East Pacific Rise and Galápagos Rift Vestimentiferan tubeworms—mainly Riftia pachyptila—and bathymodiolid mussels, vesicomyid clams, alvinellid , amphipods, and crabs Northeast Pacific Vestimentiferan tubeworms excluding Riftiidae, polychaetes, and gastropods Western Pacific Barnacles, limpets, bathymodiolid mussels, “hairy” gastropod, vesicomyid clams, and shrimp Central Indian Ridge Caridean shrimp Rimicaris kairei, and mussels, “scaly” gastropods, and anemones

32 Oceanography Vol. 20, No. 1 lation—both from one another and from precious such as silver, , and thermal vent chemistry and functioning, mid-ocean ridges—and their relatively platinum, as well as those often thought and associated mineral deposits, can be short active geological ages make back of as highly toxic to life (e.g., cadmium, found in an article by Tivey in this issue). arcs particularly interesting habitats for mercury, arsenic and ). The buoyant, It is these vent fluids that provide the biogeographic and gene-flow analyses superheated fluid is expelled into the wa- necessary energy, in the form of reduced (Desbruyères et al., 2007; see below). ter column through polymetallic chemicals, for the development of the occurs at chimneys or “black smokers” that can lush faunal communities found at vents. mid-ocean ridges when dense, cold sea- tower tens of meters above the seafloor The hydrothermal trophic web is based percolates downward through frac- (Figure 3). While approximately half of in the production of chemoautotrophic tured near the ridge crest all hydrothermal fluids are currently be- (Figure 2). The vent fluid is geothermally lieved to be discharged from the seafloor Eva Ramirez-Llodra ([email protected]) heated close to the chamber that in this spectacular black-smoker form, is Research Scientist, National Ocean- feeds the ridge, reaching temperatures about an equal amount mixes with cold ography Centre, Southampton, UK, and that can exceed 400°C. The fluid is also and percolates down into the Institut de Ciències del Mar (CMIMA-CSIC), chemically modified, losing all dissolved highly fractured seafloor as more dilute Barcelona, Spain. Timothy M. Shank and accumulating high con- warm . The latter have precipitated is Associate Scientist, Biology Department, centrations of dissolved reduced their mineral load but, importantly, still Woods Hole Oceanographic Institution, such as methane and ). contain high concentrations of dissolved Woods Hole, MA, USA. Christopher They are also strongly acidic (pH 2–3) methane and hydrogen sulfide, which are R. German is Senior Scientist and Chief compared to the near-neutral character emitted from the in this “diffuse- Scientist for Deep Submergence, Geol- (~ pH 8) of the deep oceans and are rich flow” form, typically at temperatures that ogy and Geophysics Department, Woods in numerous metals—dominantly , are measured in tens of degrees Celsius Hole Oceanographic Institution, Woods manganese, , and , but also (further detailed information on hydro- Hole, MA, USA.

Figure 2. Schematic illustra- tion of hydrothermal fluid cir- culation beneath the seafloor at mid-ocean ridge crests. From German and Von Damm (2004)

Oceanography March 2007 33 microbes (Cavanaugh, 1983; Jannasch, diversity studies in hydrothermal vents The first impression when looking at 1984) that use reduced compounds from takes us back to nine black-and-white a vent field is the profusion of life, with the vent fluids to fix carbon dioxide photographs of large white bivalves on high densities and of exotic ani- from seawater. The microbes are found black , taken by the Deep-Tow mals (Figure 4). This bounty explains free-living as well as in highly success- Camera System on May 29, 1976, on the why hydrothermal vents are often re- ful with many of the macro- Galápagos Spreading Center (Lonsdale, ferred to as “oases of the deep.” While inhabiting the vent habitat 1977). A year later, during Alvin dives re- biomass is extremely high relative to and they are responsible for the very high visiting the region, geologist Jack Corliss the surrounding , biodiversity production at hydrothermal vents. gave astonishing accounts of the first at vents is low, especially in relation to observations of giant tubeworms later the high biodiversity of non-chemosyn- Biogeography of described as Riftia pachyptila (Corliss et thetic, deep-sea (Grassle and Hydrothermal Vents: al., 1979) and thick beds of mussels as- Macioleck, 1992). This relationship is A Knowledge Puzzle sembled around seafloor openings emit- true for all groups, from meio- and mac- Under Construction ting diffuse vent fluids. Only 30 years rofauna found on oceanic basalt and sed- Biodiversity and Adaptations have passed since those initial discover- iments highly modified by hydrothermal A sound knowledge of an ’s ies; we now know of more than 550 vent fluid to macro- and megafauna living species composition is essential for any species and their composition and dis- on beds or tubeworm fields (Van subsequent biogeographic investigations tribution at more than 100 vent sites Dover, 2000). The low diversity and high and to understand the factors driving along the global mid-ocean ridge system densities of individuals in vent commu- distribution patterns. The dawn of bio- (Desbruyères et al., 2006). nities is typical of habitats with high en- ergy availability and environmental con- ditions with high or low values within their range, such as very wide tempera- ture ranges or high levels of toxic chemi- Figure 3. Black smoker from the North cals. The most speciose phyla found at Mid-Atlantic Ridge. Photo credit: hydrothermal vents are the , Deutsche-Ridge Schwerpunkt Program and the MARUM-Quest ROV followed by the molluscs and the anne- lids. However, our knowledge of the di- versity and distribution of vent species is still limited to an extremely small section of the global ridge system, and every new detailed survey and investigation, both on newly discovered sites and on well- explored ones, brings new species and to be identified and described, as well as new insight to the phylogeog- raphy of vent species. One major characteristic of vent bio- logical communities is the high degree (~ 85 percent) of species , with many species showing important physiological, morphological, and eco- logical adaptations to particular environ-

34 Oceanography Vol. 20, No. 1 Figure 4. The giant tubewormRiftia pachyptila, the mussel Bathymodiolus, and crabs from East Pacific Rise vents. Image courtesy of Richard A. Lutz

mental factors. Arguably, one of the most transports unbound or available sulfide. Gebruk et al., 2000) on the chimney remarkable vent animals is the giant A different fascinating walls, feeding on symbiotic cul- tubeworm Riftia pachyptila from the from the East Pacific Rise is the Pompeii tivated within their enlarged gill cham- Galápagos Rift and East Pacific Rise. This worm, . This poly- bers, on their external carapace, and on worm and related species were first de- chaete inside organic tubes that ingested mineral particles. The “” scribed as a new , Vestimentifera, it builds on the exterior walls of active of R. exoculata are considered to have but have now been moved to the family chimneys in close proximity to venting evolved into a broad ocular plate that in the Phylum Annelida, fluid; in contrast to the Riftia habitat, forms part of the shrimp’s dorsal sur- Class Polychaeta (Halanych, 2005). Riftia the chemistry of this fluid is domi- face. The plate’s novel photoreceptors pachyptila lack a mouth and a diges- nated by iron-bound complexes with apparently do not form a distinct image; tive system, and depend completely for hydrogen sulfide (Luther et al., 2001). instead, these “eyes” are highly sensitive growth and reproduction on organic The maximum temperature tolerated to dim light, perhaps an adaptation for matter produced by their endosymbiotic by this species is a continuing subject detecting radiation from the orifices of chemosynthetic bacteria. These bacteria of debate, with values ranging between hot black smokers and for allowing the are found densely packed in an organ 10°C and 80°C (Gaill and Hunt, 1991; shrimp to orient to the chimney walls that spans the interior of the animal’s Cary et al., 1998). (Van Dover et al., 1989). body, called the . The tube- In the , a remarkable ex- Examples of metabolic, morphologi- worm absorbs hydrogen sulfide, carbon ample of sensory adaptation is found on cal, sensory, and symbiotic adaptations dioxide, and oxygen through its well- the caridean shrimp Rimicaris exoculata are abundant at the vent habitat, and ex- irrigated plume and transports them to from the Mid-Atlantic Ridge. These ploration of new sites still results in the the trophosome via the vascular system shrimp live in massive, dense, and dy- discovery of striking animals, such as the along with modified , which namic aggregations (3000 per liter; “scaly foot gastropod” recently found on

Oceanography March 2007 35 the Central Indian Ridge (Figure 5). This science that documents and explains phylogeographic relationships with other gastropod, still awaiting description, har- spatial patterns of biodiversity. The pres- reducing environments, such as cold bors thiotrophic (sulfurous) bacteria in ent distribution of a species represents seeps on continental margins or whale a greatly enlarged esophageal gland and (1) the historical events acting on geo- falls, and (6) identifying faunal provinces has an operculum modified into several logical time scales, or vicariance, that and their boundaries allows us to rap- hundred aligned scales covered by thick have shaped its geographical range, and idly characterize a mechanism respon- layers of iron , whose purpose is (2) the dispersal potential of the species sible for creating and/or maintaining unknown (Warén et al., 2003). determined by its life-history patterns, the province. Our present knowledge of topography, and , and acting hydrothermal vent biology identifies six Known Distribution Patterns on ecological time scales (Tunnicliffe et distinct biogeographical domains char- of Vent Species al., 1996). A number of physicochemi- acterized by specific faunal assemblages: Although we have only explored a cal properties of hydrothermal vent (1) the Azores shallow Atlantic vent com- very limited proportion of the active habitats (described above) make these munities (80–1700 m), (2) the fauna of hydrothermal vents that potentially exist systems interesting for biogeographical the deep Mid-Atlantic Ridge vents be- worldwide, the description of the fauna studies (Tunnicliffe et al., 1991, 1998): tween the Azores Triple Junction and the from the vent sites that we do know is (1) the species are constrained to specific, Equator, (3) the East Pacific Rise com- already sufficient to demonstrate that linear habitats, (2) vent habitats are glob- munities found from 30°N to the Easter different faunal communities character- ally distributed, (3) they are discrete and Micro-Plate, (4) the Northeast Pacific ize different ocean basins/hydrothermal ephemeral environments, (4) there is a vent communities of the Explorer, Juan regions (Figure 1). Biogeography is the high degree of endemism, (5) there are de Fuca, and Gorda Ridges, (5) the west- ern Pacific back-arc basin communities, and (6) the vent fauna of the Central (Figure 1, Table 1) (Van Dover et al., 2002; Shank, 2004). Figure 5. The “scaly foot gastropod” from the Speciation Through Vicariance Indian Ocean ridge. Image courtesy The processes affecting biogeographic of Charles Fisher patterns over geological time scales in- clude vicariant events, where the move- ment of oceanic plates plays a defining role. Vicariance is, for example, used to explain the formation and boundary of the Northeast Pacific and East Pacific Rise biogeographic provinces during the Mid-Tertiary (~ 28 million years ago [Ma]). Subduction of the Farallon Plate under the North American Plate caused the split of the Farallon-Pacific Ridge, resulting in the separation of the Gorda- Juan de Fuca- system, to the north, from the East Pacific Rise, to the south (Tunnicliffe et al., 1996). Subsequent biological isolation permit-

36 Oceanography Vol. 20, No. 1 ted divergent evolution of the fauna on gested as a potential dispersal route for number of small eggs to species that the now-bisected ridges, resulting in an vent siboglinid polychaetes between the produce a lower number of larger eggs absence of the tubeworm Riftia pachyp- Southwest Pacific back-arc basin and the (Ramirez-Llodra, 2002). Typically, lar- tila, the , East Pacific Rise fauna (Tunnicliffe and vae hatching from smaller eggs develop and the polychaete Alvinella spp. from Fowler, 1996; Kojima et al., 2003). during a planktonic phase in which they Northeast Pacific vent sites, and the pro- liferation of vestimentiferan tubeworms on the East Pacific Rise. Another example of a major vicari- With every new discovery and investigation of ant event is the closure of the Isthmus of known sites and communities, our knowledge and Panama 5 Ma and the resulting isolation understanding of the diversity and functioning of Pacific marine fauna from the . Although there was a major deep- of these remote and exuberant increases, ocean gateway in this region prior to the helping us understand the processes driving the closure, there were no ridge connections. deep sea and the global . However, comparisons of cold-seep fau- na from the Gulf of , the margin, and the California margin indi- cate close taxonomic similarities, suggest- Distribution Patterns feed in the , allowing lon- ing a dispersal pathway for seep species and Ecological Processes ger dispersal times but also increasing between the Pacific and Atlantic Oceans While tectonic dynamics affect bioge- mortality risks. In contrast, larvae hatch- via the (now-closed) Isthmus of Panama ography at the geological time scale, a ing from larger eggs exhibit abbreviated, (Tunnicliffe et al., 1996). The only exist- number of ecological and biological fac- lecithotrophic development, usually ing ridge connection between the Pacific tors play major roles in the distribution feeding only on the energetic reserves and the Atlantic was through a complex of species, acting within the life span of provided in the egg. These nonfeeding mid-ocean ridge and subduction system the animal. These processes are related larvae have, in general, lower dispersal between South America and Antarctica, to the life-history patterns of the species potentials but higher survival probabil- which, in the Tertiary, could have served and environmental factors. Marine in- ity. However, recent studies show that as a dispersal pathway between the two ensure gene flow, dispersal, the low temperature of bathyal and abys- ocean basins (Tunnicliffe et al., 1996). and colonization through their larval sal waters can result in long metabolic A similar analysis of vent fauna at the phase. The microscopic larvae of in- life spans for larvae of vent species such genus level indicates a relatively close vertebrates are transported passively in as Riftia pachyptila, allowing for long connection between the western Pacific the water column, and are affected by dispersal distances (Marsh et al., 2001). back-arc and Central Indian Ridge vent a number of factors, both biotic (e.g., Furthermore, studies show that the em- faunas as described by Desbruyères et al., metabolic capacity, larval anatomy, be- bryos of the vent polychaete Alvinella 2007. These authors suggest a potential havior, swimming capabilities, and mor- pompejana can achieve long dispersal past dispersal pathway via the Southeast tality rate) and abiotic (e.g., deep-water times through developmental arrest Indian Ridge, Macquarie Ridge Complex currents, geological barriers, distance when dispersing in cold abyssal waters, (south of New Zealand), and Kermadec between sites, and chemical suitability of only completing their development upon Arc (north of New Zealand) that would habitats) (Cowen et al., 2000; Mullineaux encountering warm water (Pradillon et explain these similarities. The connec- et al., 2005). The life-history traits of al., 2001). In hydrothermal vent species, tion via the Pacific-Antarctic ridges and are very diverse, there are examples of all life-history pat- Macquarie Ridge has also been sug- ranging from species that produce a high terns and larval types, but the relatively

Oceanography March 2007 37 limited data on reproduction of vent tially favorable settlement habitats along by the Iceland and its off-axis species indicates that lecithotrophic lar- the ridge axis that act as stepping stones trace, which generates a shallow-water val development is dominant (Tyler and for genetic communication among sill extending across the full breadth of Young, 1999; Young, 2003). It is the in- populations (or gene flow) through the North Atlantic Ocean. Recent geo- teractions between the life-history traits dispersal and successful colonization. chemical studies indicate the existence of a species and the physical, chemical, Furthermore, hydrothermal plumes are of abundant active vents on the Gakkel and geological environmental factors of not retained within the shallow walls Ridge (Edmonds et al., 2003), but these the larval and juvenile habitat that shape of the axial East Pacific Rise summit habitats and potential associated fauna the dispersal and colonization of a spe- graben, allowing for across-axis plume have not yet been directly observed. A cies; at the same time, ridge spreading movement and potentially greater larval biological investigation of the Gakkel rates, deep-ocean currents, major frac- dispersal (Van Dover et al., 2002). Ridge during the International Polar Year ture zones, and other bathymetric fea- Large fracture zones and other topo- (2007–2008) will provide crucial data tures can all also play important roles in graphic features (e.g., chains, for understanding the evolution of fauna driving biogeographic patterns. depressions) can play important roles in that may have thrived in isolation for the The effect of ridge spreading rate on the gene flow of vent species; the research past ~ 28 million years. larval dispersal is, potentially, directly is currently addressing their In the equatorial region of the Mid- related to the distance between active effects. A recent study on genetic dif- Atlantic Ridge, the 60-million-year-old vent sites, their life span, and the geo- ferentiation of five vent —Riftia Chain and Romanche Fracture Zones morphology of the ridge valley (see, pachyptila, Tevnia jerichonana, Oasi- represent formidable geological features e.g., Baker and German, 2004). On sia alvinae, Alvinella pompejana, and (4-km high and 935 km of ridge offset) slow-spreading ridges such as the Mid- Branchypolynoe symmytilida—in eastern that greatly affect both the linearity of Atlantic Ridge, where the time-averaged Pacific vents revealed barriers to dispersal the ridge system and large-scale ocean magmatic budget is relatively low, vent affecting one or several of these species circulation in this region. It has been sug- sites may be fewer and longer lived. (Hurtado et al., 2004). The 6000-m de- gested that these major facture zones re- Consequently, the distance between po- pression at Hess Deep between the East sult in a geological barrier to the disper- tentially suitable areas may Pacific Rise and Galápagos Rift could sal of vent species along the Mid-Atlantic be greater, and, certainly, the presence limit gene flow ofAlvniella pompejana Ridge axis, isolating certain species north of large fracture zones interspersed with (Hurtado et al., 2004) and of the am- and south of the equator, while others, numerous smaller-scale (≥ 10 km) off- phipod Ventiella sulfuris (France et al., such as the shrimp Rimicaris exoculata sets every few tens of kilometers along 1992), while the 240-km Rivera Frac- whose high pelagic larval and juvenile axis creates potential physical barriers to ture Zone on the northern East Pacific dispersal capacity allows it to exploit larval dispersal. Also, the valleys of slow- Rise may be a barrier to dispersal for phytodetritus in the water column, may spreading ridges have steep walls that O. alvinae, T. jerichonana, and B. sym- not be affected (Shank et al., 1998). retain the hydrothermal plume within mytilida (Hurtado et al., 2004) as well as These hypotheses are being tested by cur- the valley, potentially limiting across- for V. sulfuris (France et al., 1992). rent research programs (see below). axis dispersal. In contrast, magmatically Another controlling factor of bio- Along with their potential as bar- more robust, fast-spreading ridges such geographic patterns may be ridges in riers to north-south, along-axis gene as the East Pacific Rise harbor vent sites hydrographic isolation. An example flow, the Mid-Atlantic Ridge fracture that may be closer in space and are cer- of a potential topographic barrier to zones may serve as a conduit for lar- tainly less likely to be offset from one gene flow is found in the Arctic, where val transport through deep-water cur- another along axis. The absence of mul- the ultraslow-spreading ridges of the rents from west to east in the Atlantic tiple ridge-crossing fractures, in particu- and Arctic Basin are Ocean. The North Atlantic Deep Water lar, creates a virtual continuum of poten- isolated from the Atlantic ridge system (NADW) flows south along the east

38 Oceanography Vol. 20, No. 1 coasts of North and South America as southern East Pacific Rise, the East Scotia Ridge south of the equator (German far as the equator before being deflected Rise near Antarctica, the southern Mid- et al., 2005; Shank, 2006). The geologi- east and crossing the Mid-Atlantic Ridge Atlantic Ridge, and the Southwest Indian cal, geochemical, physical, and biologi- through the fracture zones (Messias et Ridge (Figure 1) (Van Dover et al., 2002; cal investigations of these regions will al., 1999). These fracture zones could be Hurtado et al., 2004). provide essential information toward pathways for migrants between chemo- Finally, depth may also play a signifi- completing the biogeographical puzzle synthetic sites in the Gulf of Mexico to cant role in biogeographic patterns, as of vent biogeography (Figure 1), increas- similar habitats in the Gulf of Guinea. suggested by Desbruyères et al. (2001) ing our knowledge of Moreover, recent physical studies indi- for the differences found near the Azores and distribution, and providing the nec- cate the presence of massive currents or Triple Junction, where depth decreases essary clues to understand what factors deep-water jets in these regions (Schmid from 2400 m at Rainbow to 850 m at drive and shape vent communities and et al., 2005) that could increase the ve- Menez Gwen. This depth variation species distribution. locity of larval transport between the causes changes in fluid toxicity and sus- To address the gaps in our knowledge two sides of the Mid-Atlantic Ridge pended mineral particles, accompanied of biogeography in deep-water chemo- and enhance the possibility of success- by an impoverishment of vent endemic synthetic ecosystems (including hydro- ful communication among popula- species at shallower depths and an in- thermal vents, cold seeps, whale falls, and tions. It is known, for example, that the crease in non-vent bathyal species. These regions of low oxygen and other reduc- seep-dwelling caridean shrimp Alvino- authors also suggest that the existence of ing habitats), the international scien- caris muricola inhabits both Gulf of several biogeographic islands in this re- tific community is joining efforts in the Mexico and Gulf of Guinea cold seeps gion is driven by depth. framework of the ChEss program (www. (Ramirez-Llodra and Segonzac, 2006); noc.soton.ac.uk/chess). ChEss is one of in addition, the siboglinid tubeworm Looking Toward the 14 field projects of the Census of Escarpia southwardae has recently been the Future (www.coml.org), a 10-year described at the Gulf of Guinea seeps, Thirty years after the discovery of hy- global initiative to describe the diver- with morphological and genetic charac- drothermal vents, the investigation of sity, , and distribution of life teristics that indicate a close taxonomic vent habitats and their associated fauna in the oceans. Within this context, the relationship with E. laminata from the is still in a critical exploratory and dis- main goal of ChEss is to describe the Gulf of Mexico seeps (Andersen et al., covery phase. Every new systematic biogeography of species from deep-water 2004). While these findings may suggest survey of a ridge section uncovers new chemosynthetic ecosystems and under- a certain degree of gene flow across the vent sites and their biological commu- stand the forces driving them (Tyler et Mid-Atlantic Ridge, data are still very nities, often yielding species new to sci- al., 2003). To achieve this goal, ChEss limited. Further molecular and larval ence and sometimes new physiological has selected a number of priority study studies are needed in order to or morphological adaptations. The latest areas based on our present knowledge of fully understand dispersal pathways of discoveries span from pole to pole and biogeography (Figure 1). These include: vent species, and particular attention around the globe, including the detec- (1) the Atlantic Equatorial Belt, to ad- should be paid to the effect of larval re- tion of hydrothermal plume signals in dress scientific questions of gene flow tention by local hydrodynamics (Marsh both the Arctic (Edmonds et al., 2003) across and along the Mid-Atlantic Ridge et al., 2001; Mullineaux et al., 2005) and and Antarctic regions (German et al., equatorial fracture zones; (2) the South- the decrease in larval concentration by 2000; Klinkhammer et al., 2001), the dis- east Pacific region off Chile, a unique diffusion and mortality (Cowen et al., covery of new hydrothermal sites along site where the Chile Ridge is subducting 2000). In the southern oceans, the Cir- the Central Indian Ridge (Van Dover beneath the South American Plate and cumpolar Current has been suggested as et al., 2001), and the first investiga- where we find all known chemosyn- a pathway that may link species from the tions of vent sites on the Mid-Atlantic thetic habitat types in close proximity;

Oceanography March 2007 39 (3) the New Zealand region, where vents, creasingly recognized as one technology, links between vent species and other seeps, whale falls, and other reducing in particular, that will be vital to push faunal communities. As an important ecosystems are also found in close prox- back the barriers to our knowledge (see, societal benefit, this research will also imity; and (4) the polar regions, with e.g., Yoerger et al., this issue). undoubtedly continue to provide po- development of field programs along the tentially interesting sources of active potentially isolated Gakkel Ridge and in Conclusion molecules for the biotechnological and the Antarctic. Large-scale national and Exploration and investigation of new biomedical industries. international collaborations and sharing sites at key locations is essential to fill of human and infrastructure resources in important gaps in the biogeographi- Acknowledgements are making such research programs pos- cal puzzle of hydrothermal vents. The E. Ramirez-Llodra is supported by sible, greatly facilitated by coordination existence of such ecosystems was com- the ChEss- pro- of international programs such as ChEss pletely unexpected 30 years ago, but gram (A.P. Sloan Foundation), which and InterRidge (www.interridge.org). their discovery has changed the way we is kindly acknowledged. C.R. German The exploration, investigation, and understand both Earth and the life upon also acknowledges support from ChEss- sampling of remote, dynamic, and it. With every new discovery and inves- Census of Marine Life and further sup- topographically complex ecosystems tigation of known sites and communi- port from the such as hydrothermal vents also requires ties, our knowledge and understand- Research Council (UK) and from the continuing development of state-of- ing of the diversity and functioning of US National Science Foundation (NSF) the-art technologies for fieldwork and these remote and exuberant ecosystems and National Oceanic and Atmospheric laboratory analyses. For example, to increases, helping us understand the Administration (NOAA). T. Shank continue exploration at ever-higher lati- processes driving the deep sea and the acknowledges support from NSF, the US National Aeronautic and Space Admin- istration Program, NOAA- , and the Deep-Ocean Exploration Institute at the Woods Hole Continuing exploration of hydrothermal vents will Oceanographic Institution. The authors no doubt lead to the description of new species, a thank Paul Tyler, Chuck Fisher, Kristen Kusek, and an anonymous reviewer for better understanding of geological and geochemical their comments and suggestions on an processes affecting their biology (and vice versa), earlier version of this manuscript. and the understanding of the phylogenetic links References between vent species and other faunal communities. Andersen, A., S. Hourdez, B. Marie, D. Jollivet, F. Lal- lier, and M. Sibuet. 2004. Escarpia southwardae sp. nov., a new species of vestimentiferan tubeworm (Annelida, Siboglinidae) from West-African cold seeps. Canadian Journal of Zoology 82:980–999. Baker, E.T., and C.R. German. 2004. On the global distribution of mid-ocean ridge hydrothermal tudes, robotic vehicles are increasingly global biosphere. Continuing explo- fields. Pp. 245–266 in Mid-Ocean Ridges: Hydro- important—both remotely operated ration of hydrothermal vents will no thermal Interactions Between the Lithosphere and the Oceans. C.R. German, J. Lin, and L.M. Parson, vehicles and autonomous underwater doubt lead to the description of new eds, Geophysical Monograph Series, Volume 148, vehicles (AUVs). Indeed, with so much species, a better understanding of geo- American Geophysical Union, Washington, DC. Cary, S.C., T.M. Shank, and J. Stein. 1998. Worms bask of the global mid-ocean ridge yet to be logical and geochemical processes affect- in extreme temperatures. Nature 391:545–546. Cavanaugh, C.M. 1983. Symbiotic chemoautotrophic investigated, AUVs that can operated ing their biology (and vice versa), and bacteria in marine invertebrates from sulphide- independently of a mother ship are in- the understanding of the phylogenetic rich habitats. Nature 302:58–61.

40 Oceanography Vol. 20, No. 1 Corliss, J.B., J. Dymond, L.I. Gordon, J.M. Edmond, Grassle, J.F., and N.J. Maciolek. 1992. Deep-sea rich- II 52:409–428. R.P. von Herzen, R. Ballard, K. Green, D. Williams, ness—Regional and local diversity estimates from Shank, T.M. 2004. The evolutionary puzzle of seafloor A. Bainbridge, K. Crane, and T.H. van Andel. 1979. quantitative bottom samples. American Naturalist life. Oceanus 42(2):78–85. Submarine thermal springs on the Galapagos Rift. 139:313–341. Shank, T.M. 2006. Preliminary biological characteriza- Science 203:1,073–1,083. Halanych, K.M. 2005. Molecular phylogeny of sibo- tion of vent sites and the evolutionary relation- Cowen, R.K., K.M.M. Lwiza, S. Sponaugle, C.B. Paris, glinid annelids (a.k.a. pogonophorans): A review. ships of vent fauna on the southern Atlantic Ridge. and D.B. Olson. 2000. Connectivity of marine Hydrobiologia 535:297–307. Exploration and biogeography of deep-water che- populations: Open or closed? Science 287:857–859. Hessler, R.R., and P.F. Lonsdale. 1991. Biogeography of mosynthetic ecosystems on the Atlantic Equatorial Desbruyères, D., M. Biscoito, J.-C. Caprais, A. Colaço, Mariana Trough hydrothermal vent communities. Belt region. First Results and Planning for Future T. Comtet, P. Crassous, Y. Fouquet, A. Khripou- Deep-Sea Research 38:185–199. Research Workshop, Barcelona, Spain. noff, N. Le Bris, K. Olu, R. Riso, P.-M. Sarradin, M. Hurtado, L.A., R. A. Lutz, and R.C. Vrijenhoek. 2004. Shank, T., R.A. Lutz, and R.C. Vrijenhoek. 1998. Mo- Segonzac, and A. Vangriesheim. 2001. Variations in Distinct patterns of genetic differentiation among lecular systematics of shrimps (Decapoda: Bresili- deep-sea hydrothermal vent communities on the annelids of eastern Pacific hydrothermal vents. idae) from deep-sea hydrothermal vents, I: Enig- Mid-Atlantic Ridge near the Azores plateau. Deep Molecular Ecology, 13: 2603–2615. matic “small orange” shrimp from the Mid-Atlan- Sea Research Part I 48:1,325–1,346. Jannasch, H.V. 1984. : The nutritional tic Ridge are juvenile Rimicaris exoculata. Molecular Desbruyères, D., M. Segonzac, and M. Bright, eds. basis for life at deep-sea vents. Oceanus 27:73–78. and Biotechnology 7:88–96. 2006. Handbook of Deep-Sea Hydrothermal Vent Klinkhammer, G., C. Chin, R. Keller, A. Daehlmann, Tunnicliffe, V. 1991. The biology of hydrothermal Fauna, Second edition, Denisia 18, State Museum H. Sahling, G. Sarthou, S. Petersen, F. Smith, and vents: Ecology and evolution. Oceanography and of Upper Austria, Linz, 544 pp. C. Wilson. 2001. Discovery of new hydrothermal Marine Biology: An Annual Review 29:319–407. Desbruyères, D., J. Hashimoto, and M.-C. Fabri. 2007. vent sites in Bransfield , Antarctica. Earth and Tunnicliffe, V., and C.M.R. Fowler. 1996. Influence of Composition and biogeography of hydrothermal Planetary Science Letters 193:395–407. sea-floor spreading on the global hydrothermal vent communities in western Pacific back-arc ba- Kojima, S., S. Ohta, T. Yamamoto, T. Yamaguchi, T. vent fauna. Nature 379:531–533. sins. Pp. 215–234 in Back-Arc Spreading Systems; Miura, T. Fujiwara, and J. Hashimoto. 2003. Mo- Tunnicliffe, V., C.M.R. Fowler, and A.G. McArthur. Geological, Biological, Chemical, and Physical In- lecular of vestimentiferans of the west- 1996. Plate tectonic history and hot vent biogeog- teractions. D.M. Christie, C.R. Fisher, S.-M. Lee, ern Pacific, and their phylogenetic relationship raphy. Pp. 225–238 in Tectonic, Magmatic, Hydro- and S. Givens, eds, Geophysical Monograph Series, to species of the eastern Pacific. III. Alaysia-like thermal and Biological Segmentation of Mid-ocean Volume 166, American Geophysical Union, Wash- vestimentiferans and relationships among families. ridges. C.J. MacLeod, P.A. Tyler, and C.L. Walker, ington, DC. Marine Biology 142:625–635. eds, Geological Society of London, Vol. 118. Edmonds, H.N., P.J. Michael, E.T. Baker, D.P. Connelly, Lonsdale, P. 1977. Clustering of suspension-feeding Tunnicliffe, V., A.G. McArthur, and D. McHugh. 1998. J.E. Snow, C.H. Langmuir, H.J.B. Dick, R. Muehe, near abyssal hydrothermal vents A biogeographical perspective of the deep-sea C.R. German, and D.W. Graham. 2003. Discovery at oceanic spreading centers. Deep-Sea Research hydrothermal vent fauna. Advances in Marine of abundant hydrothermal venting on the ultra- 24:857–863. Biology 34:353–442. slow-spreading Gakkel Ridge in the Arctic Ocean. Luther, G.W., T.F. Rozan, M. Taillefert, D.B. Nuzzio, Tyler, P.A., and C.M. Young. 1999. Reproduction and Nature 421:252–256. C. Di Meo, T.M. Shank, R.A. Lutz, and S.C. Cary. dispersal at vents and cold seeps. Journal of the France S.C., R.R. Hessler, and R.C. Vrijenhoek. 1992. 2001. Chemical speciation drives hydrothermal Marine Biological Association of the United King- Genetic differentiation between spatially-disjunct vent ecology. Nature 410:813–816. dom 79:193–208. populations of the deep-sea, hydrothermal vent- MacDonald, K.C., D.S. Scheirer, and S.M. Carbotte. Tyler, P.A., C.R. German, E. Ramirez-Llodra, and C.L. endemic amphipod Ventiella sulfuris. Marine Biol- 1991. Mid-ocean ridges: Discontinuities, segments, Van Dover. 2003. Understanding the biogeography ogy 114:551–559. and giant cracks. Science, 253:986–994. of chemosynthetic ecosystems. Oceanologica Acta Gaill, F., and S. Hunt. 1991. The biology of Marsh, A.G., L.S. Mullineaux, C.M. Young, and D.T. 25:227–241. worms from high temperature hydrothermal vent Manahan. 2001. Larval dispersal potential of the Van Dover, C.L., E.Z. Szuts, S.C. Chamerlain, and J.R. regions. Review of Aquatic Sciences 4:107–137. tubeworm Riftia pachyptila at deep-sea hydrother- Cann. 1989. A novel ‘eyeless’ shrimp from the hy- Gebruk, A.V., E.C. Southward, H. Kennedy, and A.J. mal vents. Nature 411:77–80. drothermal vents of the Mid-Atlantic Ridge. Na- Southward. 2000. Food sources, behaviour, and Messias, M.J., C. Andrié, L. Memery, and H. Mercier. ture 337:458–460. distribution of hydrothermal vent shrimps at the 1999. Tracing the North Atlantic Deep Water Van Dover, C.L. 2000. The Ecology of Deep-Sea Hydro- Mid-Atlantic Ridge. Journal of the Marine Biologi- through the Romanche and Chain fracture zones thermal Vents. Princeton University Press, Princ- cal Association of the United 80:485–499. with chlorofluoromethanes. Deep-Sea Research I eton, NJ, 424 pp. German, C.R., R.A. Livermore, E.T. Baker, N.I. Bru- 46:1,247–1,278. Van Dover, C.L., S. Humphris, D. Fornari, C. Cavana- guier, D.P. Connelly, A.P. Cunningham, P. Morris, Mullineaux, L.S., S.W. Mills, A.K. Sweetman, A.H. Be- ugh, R. Collier, S.K. Goffredi, J. Hashimoto, M.D. I.P. Rouse, P.J. Statham, and P.A. Tyler. 2000. Hy- audreau, A. Metaxas, and H.L. Hunt. 2005. Vertical, Lilley, A.L. Reysenbach, T.M. Shank, K.L. Von drothermal plumes above the East Scotia Ridge: An lateral and temporal structure in larval distribu- Damm, A. Banta, R.M. Gallant, D. Gotz, D. Green, isolated high-latitude back-arc spreading centre. tions at hydrothermal vents. Marine Ecology Prog- J. Hall, T.L. Harmer, L.A. Hurtado, P. Johnson, Z.P. Earth and Planetary Science Letters 184:241–250. ress Series 293:1–16. McKiness, C. Meredith, E. Olson, L.L. Pan, M. Tur- German, C.R., and K.L. Von Damm. 2004. Hydrother- Pradillon, F., B. Shillito, C.M. Young, and F. Gaill. 2001. nipseed, Y. Won, C.R. Young III, and R. Vrijenhoek. mal Processes. Pp. 181–222 in Treatise on Geochem- Developmental arrest in vent worm embryos. Na- 2001. Biogeography and ecological setting of Indi- istry, Volume 6. The Oceans and Marine Geochem- ture 413:698. an Ocean hydrothermal vents. Science 294:818–823. istry, H. Elderfied, executive ed, and K.K. Turekian Ramirez-Llodra, E. 2002. Fecundity and life-history Van Dover, C.L., C.R. German, K.G. Speer, L.M. Par- and H.D. Holland, eds, Elsevier, Oxford, UK. strategies in marine invertebrates. Advances in Ma- son, and R.C. Vrijenhoek. 2002. Evolution and bio- German, C.R., L.M. Parson, B.J. Murton, S.A. Ben- rine Biology 43:88–170. geography of deep-sea vent and seep invertebrates. nett, A.J. Connelly, D.P. Evans, R.D. Prien, E. Ramirez-Llodra, E., and M. Segonzac. 2006. Reproduc- Science 295:1,253–1,257. Ramirez-Llodra, T.M. Shank, D.R. Yoerger, M. tive biology of Alvinocaris muricola (Decapoda: Warén, A., S. Bengtson, S.K. Goffredi, and C.L. Van Jakuba, A. Bradley, E.T. Baker, and K. Nakamura. Caridea: ) from cold seeps in the Dover. 2003. A hot-vent gastropod with iron sul- 2005. Hydrothermal activity on the southern Gulf of Guinea. Journal of the Marine Biology Asso- fide dermal sclerites. Science 303:1,007. Mid-Atlantic Ridge: Tectonically- and Volcani- ciation of the United Kingdom 86:1–10. Young, C.M. 2003. Reproduction, development and cally-Hosted High Temperature Venting at 2–7°S. Schmid, C., B. Bourlès, and Y. Gouriou. 2005. Impact life history traits. Pp. 381-426 in Ecosystems of the Eos, Transactions, American Geophysical Union 86, of the equatorial deep jets on estimates of zonal Deep Oceans. P.A. Tyler, ed, Elsevier, London, UK. Abstract OS21C-04. transports in the Atlantic. Deep-Sea Research Part

Oceanography March 2007 41