Hotspot Ecosystem Research on Europe ’ s Deep-Ocean Margins

132 Europe’s deep-ocean margin stretches over a distance of 15,000 km along the Atlantic Ocean from the Arctic to the Iberian margin and from western to eastern Mediterranean, and to the Black Sea. The margin extends from the shelf edge at about 200 m depth until about 4000 m depth where the abyssal plain or oceanic basins begin, and covers three million square kilometers, an area about onethird of that covered by Europe’s landmass. Most of this deep-ocean frontier lies within Europe’s Exclusive Economic Zone (EEZ) and is therefore of direct interest for the exploitation of biological, energy, and mineral resources. A major European policy aim is to develop these resources in an ecologically sustainable manner. This requires a profound knowledge of the structure and dynamics of ocean margin ecosystems incorporating a wide variety of complex environments, such as deep-sea corals, cold seeps, and canyons. The knowledge reHotspot Ecosystem Research on Europe’s Deep-Ocean Margins

be distinguished from each other before man's activities make this distinction impossible (Danovaro et al., 2001).In some areas, notably deep-water coral reefs, man's impact on the environment has already been considerable (Freiwald et al., 2004).Since the mid-1980s, the socio-economic value of cold-water coral ecosys-tems has risen tremendously.In many areas of the European EEZ, major trawling areas overlap with occurrences of coral ecosystems (Freiwald et al., 2004).
Trawling over cold-water coral reefs with demersal trawls is comparable to forest clear-cutting, but the coral systems might take much longer to recover.The documentation of ongoing damage to the benthic ecosystem and a risk analysis of planned human activities along Europe's continental margin is a major issue that will be addressed by the HERMES research consortium.

Cold Seep and Microbially Driven Ecosystems
Microbes occur in every niche in the ocean and comprise a signifi cant part of the global biomass.In some continental margin ecosystems, they dominate life almost exclusively, generating a great diversity of bacteria, archaea, and some single-cell eukaryotes.Natural chemical laboratories occur in areas of subsea discharge of fl uids and gas (e.g., methane) (Boetius et al., 2000).The free living and symbiotic microbial communities associated with some invertebrates are nourished by the chemical energy rising from these sources and form the basis of cold seep ecosystems (Sibuet and Olu-Le Roy, 2003).These often take the form of dense and endemic benthic communities, in which the high production of organic carbon sustains large size or typical animals and very high biomasses.
In high methane fl ux areas, the benthic biomass produced through chemosynthetic processes can be 1,000 to 50,000 times greater than the deep-sea biomass resulting indirectly from photosynthetic production.The remarkable abundance Determining the distribution as well as the resilience of these ecosystems is fundamental to producing plans for their sustainable management.provides evidence for a variety of these ecosystems, which hold a great diversity and biomass of bacteria and archaea (Boetius et al., 2000).Our current understanding of the distribution of methane within sediments, both regionally and on small scales, is poor.The relationship of hydrate to microbial activity, and to venting and support of chemosynthetic communities, is also poorly known.Additionally, gas hydrates pose a potential threat because submarine landslides could be initiated by their rapid melting.This process would also release large volumes of the greenhouse gas methane into the atmosphere.We will monitor fl uid release at a variety of sites (including some known, overlying areas of methane hydrate), to determine (1) the contribution of this source of carbon to the hy-

Canyon Ecosystems
Canyons are deep incisions of the continental shelf and slope.They dissect much of the European ocean margin.
Were these canyons on land, they would present some of the most dramatic mountain scenery in the world.Hidden by the ocean, they have been ignored.This is largely because of diffi culties in exploring their complex terrain.Yet, canyons are known as (1) hotspots of high faunal biomass, (2) major pathways for transportation and burial of organic carbon in the oceans, and (3) fast-track corridors for material transported from the land to the deep sea (Rogers et al., 2003).It is only now-with advances in technology such as ROVs, swath bathymetry, side-scan sonar, and defi nitive position-fi xing systems-that progress is being made in their study.Some canyons are closely connected to major river out- fl ow systems while others funnel large quantities of sediment from the continental shelf into deep water.Canyons act as temporary depots for sediment and carbon storage.However, rapid, episodic fl ushing of canyons may mobilize large amounts of sediment carrying it to the abyss and overwhelming benthic ecosystems over a wide area (Thomsen et al., 2003).The frequency of these potentially catastrophic events and the fl uxes of particles produced are largely unknown, as are the rates of recolonization and restoration of the canyon ecosystems.
Canyons are complex systems in terms of their hydrography, sedimentology, biogeochemistry, and biology.As more is learned about canyons, it becomes increasingly obvious that there is great variability both within individual canyon systems and between different canyons.
Individual canyons have very different environmental characteristics that determine the diversity and the ecology of their fauna (Vetter and Dayton, 1998).
This makes it diffi cult to reach generalizations that will be useful in creating policies for whole ecosystem management, without ( 1  microcosm studies suggest that biodiversity peaks at intermediate levels of productivity (Kassen et al., 2000).Inverse relationships between biodiversity and ecosystem functioning have also been observed, suggesting the key role of a few species (Loreau et al., 2001)

INTEGR ATION OF GEOSCIENCE WITH ECOSYSTEM STUDIES
Over the past decade, marine geologists have become increasingly involved in the application of marine geoscience to biological issues, such as characterization of habitat structure and dynamics (Wefer et al., 2003).This is a direct result of sig-

Hotspot
Ecosystem Research on Europe's Deep-Ocean Margins B Y P H I L I P P.E .W E A V E R , D A V I D S .M .B I L L E T T , A N T J E B O E T I U S , R O B E R T O D A N O V A R O , A N D R É F R E I W A L D , A N D M Y R I A M S I B U E T S T R ATA F O R M AT I O N O N E U R O P E A N M A R G I N Squired must be generated in an integrated way that ties research on biodiversity and biological processes intimately to the physical factors that control ecosystems (geology, sedimentology, physical oceanography, biogeochemistry).In addition, it is important to set present-day ecosystems in an historical framework by studying the sediment record to determine long-term environmental changes and the potential response of ecosystems to global change over decadal to millennial scales.Changes due to large-scale natural forcing (e.g., climate oscillations, sea-level change) or to more local human effects (e.g., resource exploitation, inputs of pollutants and nutrients) must

A
consortium of 45 partners, including 9 small companies from 15 European countries (Box 1), are being funded under the European Union's Sixth Framework Research Programme to study benthic ecosystems on Europe's continental margins.The project-HERMES (Hotspot Ecosystem Research on the Margins of European Seas)-will begin in early 2005.It will study "hotspot" ecosystems-discontinuous environments that are constrained by chemical, physical, topographic, and geological factors and that contain a wealth of unknown species that thrive in insular habitats.Determining the distribution as well as the resilience of these ecosystems is fundamental to producing plans for their sustainable management.HERMES takes a major leap forward from previous, smaller research projects because it coordinates research efforts along the whole European margin.
The project-HERMES (Hotspot Ecosystem Research on the Margins of European Seas)-will begin in early 2005.It will study "hotspot" ecosystemsdiscontinuous environments that are constrained by chemical, physical, topographic, and geological factors and that contain a wealth of unknown species that thrive in insular habitats.
Ecosystems on continental margins that are least understood include canyon ecosystems, microbially driven ecosystems in anoxic environments, and chemosynthetic ecosystems associated with methane seeps.In addition, deep-water coral ecosystems require urgent study as they occur at depths where deep-water trawlers are active; these trawlers have already caused considerable destruction of these fragile habitats.Despite their fragmented distribution, these ecosystems have important functions: (1) cold seep and anoxic ecosystems act as fi lters for methane and sulfi de, (2) deep-water corals play a role in CO 2 sequestration, and (3) canyon systems are preferential conduits and deposition centers for carbon and are thought to be important nursery areas for deep-water fi sh stocks.HERMES will also study open-slope ecosystems adjacent to the hotspots so that biological systems can be studied in the context of the wider continental slope.Here, large environmental gradients (temperature, pressure) and major environmental perturbations, such as recent landslides, play an important role.Cold-Water Coral Ecosystems and Carbonate Mounds Cold-water coral ecosystems create reeflike frameworks and contribute to the formation of carbonate mounds.The colonial stone corals Lophelia pertusa and Madrepora oculata (Figure 1) occur on the deep shelves along 4500 km of the northwestern European continental margin, and in Scandinavian fjords.Despite intense mapping, progress achieved during the Fifth Framework Programme's ACES, ECOMOUND, and GEOMOUND projects, and various national seabed mapping surveys off Norway and Ireland, researchers still do not know how many reefs and mounds exist.(ACES is the Atlantic Coral Ecosystem Study.ECOMOUND is Environmental Controls on Mound Formation along the European Continental Margin.GEO-MOUND is a project focusing on the geological evolution of giant, deep-water carbonate mounds off western Ireland

Figure 1 .
Figure 1.Scleractinian (Lophelia pertusa) and red actiniarians at around 850 to 900 m depth in the Pelagia Mound province, southeastern Rockall Trough off western Ireland and the United Kingdom.Image courtesy IFREMER-CARACOLE cruise in 2001 with the ROV Victor in the Northeast Atlantic.

Figure 2 .
Figure 2. Sampling of microbial mats at the Haakon Mosby mud volcano, located on the Norwegian margin west of the Barents Sea at 72°N.Two joint French/German expeditions with the research vessels Atalante (IFREMER) and Polarstern (Alfred Wegener Institute) and the deep-diving ROV Victor 6000 (IFREMER) were carried out in 2001 and 2003 to study the biogeochemistry of this active mud volcano.Th e white patches are mats of the giant sulfi de-oxidizing bacterium Beggiatoa.Th ese bacteria profi t from high fl uxes of sulfi de produced by anaerobic methane-oxidizing communities in the subsurface sediments.
drosphere, (2) its rate of use by seabed communities, and (3) the variation in this fl uid escape with time.HERMES will integrate geological techniques with biogeochemical fl ux measurements and biological data to achieve a quantitative understanding of ecosystems.The aim of future research on these ecosystems is to understand (1) fl uid generation mechanisms and how the fl ux rates vary through zones and time, (2) the historical development of mud volcanoes and pockmarks which include the characterization and dating of authigenic carbonates and associated sediment, and (3) how physical and chemical characteristics of fl uids in the sediment, crust, and at the sediment-water interface control the community diversity, the dynamics of the system, and the biological production based on chemosynthesis through free and symbiotic bacteria.

Figure 3 .
Figure 3. Microbial biofi lms at a cold seep on the Nile deep-sea fan at 2970 m depth.Image courtesy IFREMER-Nautinil cruise (EUROCORE Euromargin project MEDIFLUX) in 2003 with the manned submersible Nautile in the eastern Mediterranean.
) a concerted effort to compare canyons from different biogeochemical provinces and different topographic settings and (2) coordinated, multidisciplinary projects relating the fauna to the environmental variables that regulate their distributions.HERMES will study specifi c canyons in four different biogeochemical provinces: (1) off Ireland, (2) off Portugal, (3) in the western Mediterranean, and (4) in the eastern Mediterranean.The physical processes in canyons will be studied with particular regard to the transport of particulate material and the distribution of key fauna.Physical processes in and around canyons can be highly complex and are diffi cult to study because many of the more important processes are episodic in nature.Understanding physical processes, such as the focusing of internal waves and storm events, are critical in understanding the production of nepheloid layers by resuspension and the enhancement of primary productivity at canyon heads.There is great temporal variation in the creation of nepheloid layers (from days to years), and fl uxes can vary over several orders of magnitude.Our view of biological processes in canyons has changed considerably in the last few years because of the increased use of submersibles and ROVs.The results indicate the importance of various zooplankton groups acting as a link to fi sh and mammal populations.The species and their abundances differ from canyon to canyon and appear to be related to downward particle fl uxes, topography, and the hydrographic features of individual canyons.Canyons appear to be important in the channeling of macrophyte debris, which may have a signifi cant effect on the relative abundance of some species.Few studies of the chemistry of canyons have been carried out, even though canyons play a crucial role in the redistribution of carbon and anthropogenic materials derived from marine primary production and terrestrial runoff.Because canyons channel and focus sediment distribution, anthropogenic tracers are relatively high in relation to surrounding slope areas.Canyons are being considered as potential disposal sites for various wastes, including carbon dioxide.These plans assume that canyons are isolated from the adjoining continental slope.We will test this assumption and determine the degree of interconnectivity between canyons and the open slope.Open Slope Ecosystems Broad open slope ecosystems are strongly infl uenced by current fl ow, seabed character, and sediment instability.Landslides, in particular, have destroyed large areas of habitat in single events.Investigating the regeneration of areas subject to recent landslides (e.g., the Nice airport slide of 1979) will provide important information on the resilience of sediment ecosystems on the continental slope and the interdependence of species.Apart from habitat loss, landslides have the potential to devastate offshore installations.Despite considerable progress in previous research programs, such as COSTA (the Continental Slope Stability program) (see Mienert et al., this issue), the causes of seabed instability (particularly landslides) are still not fully understood, and, yet, they pose signifi cant threats to coastal communities through associated tsunamis.For example, the Storegga slide and its resulting tsunami devastated So far, marine ecosystem assessment suffers from a lack of models integrating biology into element cycles and global change issues.Denmark 8,200 years ago.Slopes are ideal systems for investigating benthic patterns: the decrease of benthic faunal biomass with increasing depth is one of the best-known patterns in marine ecology.However, there is considerable variability in the abundance and biomass of benthic fauna along the same isobath, and upwelling regions and coastal trenches may lead to "hotspots" of life.It is becoming increasingly evident that we are not able to predict the spatial distribution of deep-sea benthic ecosystems using a limited set of variables.Deep-sea hotspots of benthic biomass suggest that the "paradigm" of decreasing life with increasing depths is not universal and that detailed knowledge is needed to explain anomalies.Determining spatial heterogeneity is one of the most signifi cant challenges in the study of continental margin ecosystems.The distribution of benthic organisms on macro-scale (>1000 km) is assumed to be dependent on physical parameters (temperature, water masses), while at mesoscale (1-100 km), export of primary production, sediment heterogeneity, oxygen availability, and catastrophic events play major roles.At smaller scales the distribution of benthic organisms is infl uenced by interactions between organisms (competition, predation) and microhabitats (sediment micro-topography, chemical interactions, food distribution).Several key questions remain unanswered: Is spatial variability of deep-sea benthos dependent upon the characteristics of the system?Is the mosaic of distribution (and community composition) of deep-sea fauna explained by the biotic/abiotic interaction at a specifi c spatial scale?Is energy source a factor relevant at all spatial and temporal scales?Are benthic processes (e.g., ecological effi ciency in exploiting resources) related to spatial variability?The deep sea is the largest ecosystem on Earth and is the largest reservoir of (yet undiscovered) biodiversity.However, most of what we know about the diversity of life on Earth comes from large-scale studies of terrestrial ecosystems (Waide et al., 1999).It is unlikely that biodiversity paradigms, evident in terrestrial ecosystems, can be applied to marine ecosystems.For instance, it is clear that on open continental slopes there is greater species richness at midslope depths (Figure 4).While changes in species and diversity do occur with increasing altitude on land, the effects on biodiversity are also related to reduction in total land area at each altitude.Such a reduction in area with increasing altitude and its effects on diversity are not apparent with increasing depth in most deepsea ecosystems.It is not clear how high local species richness in deep-sea sediments is related to ecosystem functioning.In terrestrial ecosystems there are reports of a linear relationship between biodiversity and ecosystem functioning, but marine

Figure 4 .
Figure 4. Biodiversity patterns in the deep sea: this illustration shows the depth-related pattern of benthic biodiversity, obtained by summarizing all the information available in literature.It is evident from the hump-shaped curve that highest biodiversity values occur at about 2000 m depths.Open slopes are expected to host most of the undiscovered biodiversity of the globe.

Figure 5 .
Figure5, which also displays the distribution of key features (deep-water coral sites, landslides, canyons, cold seeps).While HERMES will concentrate its research effort on a few specifi c sites, it aims to compare similar systems within each area, where possible, so that general principles on the interaction of biodiversity, the environment, and ecosystem functioning might be generated.The HERMES study areas represent a range of environments:1.The Nordic margin is a cold-water end member with environmentally stressed ecosystems from intensive