GeoHab 2008 Agenda and Abstracts

Back to Sitka Home page

Abstracts

Integrating terrestrial, freshwater and marine processes in a preliminary classification of coastal ecological units in southeastern Alaska: A hierarchical framework and exploratory analysis

David Albert¹, K. Koski¹, Zach Ferdana¹, Jim Baichtal²

1) The Nature Conservancy

2) USFS Tongass National Forest

Complex interactions among terrestrial, freshwater and marine environments exemplify coastal ecosystems in southeastern Alaska. Our objective was to construct a GIS database to characterize coarse-scale ecological processes and develop a preliminary classification of estuarine systems. To describe the nested nature and spatial complexity of these systems, we developed a six-tiered, hierarchical framework that includes (1) marine ecoregions, (2) coastal basins, (3) sub-basins, (4) coastal watersheds, (5) shoreline units and where available (6) across-shore tidal zones and biobands. In addition, a database of individual estuary features was developed that links individual stream systems with shoreline and nearshore habitat features, and also nests within upper levels of the hierarchy. Additional attributes in the database include terrestrial landforms, freshwater fluvial processes, shoreline substrate, wave exposure, intertidal and nearshore vegetation, marine basin geometry and estimated tidal volume. We conducted a series of exploratory analyses to provide a coarse-scale characterization of estuarine systems that may be considered as a working hypothesis to be tested as finer-scale data on physical and biological processes become available.

Refinement and Application of the Coastal and Marine Ecological Classification Standard

R. Allee¹, D. Bamford¹, M. Finkbeiner¹, K. Goodin², C. Madden³

1) National Oceanic and Atmospheric Administration

2) NatureServe

3) South Florida Water Management District

The Coastal and Marine Ecological Classification Standard (CMECS) was first introduced in 2004 as the first nationally applicable habitat classification system developed for the United States. CMECS was conceptually designed to allow coastal and marine habitat managers access to data and information on the physical structure and associated biodiversity of the vast array of habitats found in U.S. coastal and marine waters. Over the years, the hierarchical structure of CMECS has been evaluated and refined to improve its applicability to mapping applications and provide the analytical opportunities requested by coastal managers. The scope of this classification standard extends upstream and landward to where ocean derived salts measure less than 0.5%? during the period of average annual low flow to the deep ocean. This encompasses estuaries, wetlands, rivers, shorelines, islands, the intertidal zone, the entire benthic zone, and the entire water column from the coast to the deep ocean. The standard is hierarchical, extending spatially and conceptually from systems, which are units of large scale, to specific biological assemblages of a very small scale. CMECS Version III addresses the needs expressed by the habitat mapping community, provides better alignment for seamless integration with the U.S. National wetland classification system and the National Vegetation Classification Standard and retains the original intent of the system to provide a common terminology and classification approach to assess the biodiversity of the habitats. This presentation will provide an overview of CMECS Version III and present classified results for various habitat types as derived from several data sources.

Paleogeography of the Late Pleistocene and Quaternary Coastlines of Southeast Alaska and Their Potential Archaeological Significance

J.F. Baichtal¹, R.J. Carlson¹, S.J. Crockford²

1) U.S. Forest Service, Tongass National Forest

2) Pacific Identifications, Inc., British Columbia, Canada

An extensive literature search and years of field reconnaissance have resulted in a dataset of over 300 shell-bearing raised marine deposits throughout Southeast Alaska. It includes site location, elevation, and description when available, and over 170 radiocarbon dates beginning at 14,380 B.P. Interpretation of this data gives insight on the timing and complexity of isostatic crustal adjustments that resulted from glaciation and deglaciation, eustatic sea level change, and subsequent tectonic uplift. Digital bathymetry data of varying resolutions across the region, analysis of marine cores, and geomorphic interpretations from the finding of fish habitat studies on the shelf and its margins were utilized in the paleogeographic modeling. Comparisons with the paleogeographic modeling of the Queen Charlotte Islands/Hecate Strait region of British Columbia suggest a similar response to ice loading during the Last Glacial Maximum (LGM) resulting in a forebulge to the west of the ice front adjacent to Prince of Wales, Baranof, and Chichagof Islands in the Alexander Archipelago. The Alexander forebulge would have created a much larger land mass than previously modeled, providing a nearly ice-free coastal plain available for plants, animals and human occupation as early as 13,500 B.P. This now submerged landform may have provided a refugium for flora and fauna for re-colonizing the islands after the LGM and an explanation for the absence of coastal archaeological sites prior to 10,000 B.P. This interpretation of the paleogeography is preliminary at this time. The extent and timing of the forbulge can only be inferred and estimated from this data. Additional surface sampling, imaging and modeling of the outer coast shelf, and analysis of sediment cores from across the shelf is needed to better define the relationships between the timing and extent glaciation, forbulge development, and the rate of sea level rise. Furthermore, analysis of selected shell-bearing raised marine deposits in southern Southeast Alaska ranging in age from 8170 to 9400 YBP suggest a warmer and dryer climate in the region. The presence of charcoal and the bones of Pacific sardine may be an indicator of a warmer and dryer climate following the end of the Younger Dryas from 10000 to 8000 YBP. This evidence combined with ongoing pollen research in Southeast suggests that this may have been a time of relatively dry climate as well, which would favour the occurrence of wildfires, either natural or man-caused.

The Global Seafloor Atlas Project: Seafloor Geomorphology as Benthic Habitat

Elaine Baker¹, Peter T. Harris², Tina Schoolmeester¹

1) UNEP/GRID Arendal, Postboks 183N-4802 Arendal, Norway

2) Marine and Coastal Environment Group, Geoscience Australia, GPO Box 378, Canberra ACT 2601

Through the efforts of GeoHab together with other marine researchers around the world, knowledge of the geomorphology of the seafloor has improved markedly over the past 10 years. Using multibeam sonar, the morphology of submarine features such as seamounts, canyons, mud volcanoes and spreading ridges has been revealed in unprecedented detail. In particular, case studies are now available for a range of seabed features where detailed bathymetric images have been combined with seabed video and sampling to yield an integrated picture of the benthic communities that are associated with different types of habitat. A picture is emerging that shows patterns of different communities that are consistently associated with certain geomorphic features in combination with other physical environmental variables.

The objective of the atlas project is to contribute to the conservation of marine biodiversity by utilising geomorphological data to identify and characterise global marine benthic habitats. Linking these habitats to marine ecosystems provides a mechanism for mapping the ecological geography of the ocean floor. The goal of the project is to produce an atlas of marine geomorphology that can be used to support the identification of a representative system of Marine Protected Areas - especially in the high seas area where there is currently a scarcity of data and information.

It is envisaged that the atlas will present a collection of up-to-date summaries of the benthic communities that are associated with different types of geomorphic features that have been described from the ocean. The atlas would contain an introductory summary chapter giving an overview of global ocean geomorphology, followed by contributions from a range of authors that combine descriptions of biological community-geomorphic feature associations, illustrated by state-of-the-art imagery of the seabed produced by acoustic and other technologies. The atlas should be inclusive wherever possible and so the number and titles of chapters could be adjusted to suite the material submitted.

The development of a methodology for the atlas has been included as a component of the UNEP project “Development of the Methodology Arrangements for the GEF and Transboundary Waters Assessment Programme (TWAP)”. The objective of this project is to develop the methodologies for conducting a global assessment of transboundary river, lake, and groundwater basins, Large Marine Ecosystems, and ocean areas for GEF purposes and to catalyse a partnership and arrangements for conducting such a global assessment.

New discoveries of coral reefs in the Hola trench off Norway, highlighting challenges in coral reef prediction

Pål Buhl-Mortensen¹, Lene Buhl-Mortensen¹, Margaret Dolan²

1) Institute of Marine Research, Norway

2) Geological Survey of Norway

Reliable predictions of spatial cold-water coral reef occurrence would be a useful tool for the management of deep coastal and offshore areas. Several studies have indicated that the distribution of cold-water coral reefs is correlated with rough topography and slopes exceeding a certain critical angle. In this study we show that this is not always the case. The Norwegian seabed mapping program MAREANO covers areas off northern Norway. Two case study areas within the MAREANO mapping area are used to investigate the potential to predict coral reef occurrence based on multibeam bathymetry, backscatter and topographic variables. Both areas are located within trenches crossing the continental shelf, but have completely different topography. At the Malangen study site the reefs occur on a ridge crossing the trench separating Malangsgrunnen from Fugløybanken, whereas within the Hola area, the reefs occur within the deeper parts of the trench on a relatively level seabed. The reefs of the two sites differ markedly in shape: the Malangen reefs are relatively circular with summits of living corals, whereas the Hola reefs are elongated with a living upcurrent front. The topography of the seabed has no influence on the corals in itself, but influences the environment by modifying the hydrodynamic setting. Currents accelerate over peaks and ridges, and provide environments for enhanced food encounter. At other locations the topography may induce hydrodynamic patterns concentrating food particles. In the Hola area the reefs occur at the side of the trench where the currents flow from the coast towards the shelfbreak. The currents are strong, and local production at the shelf may have increased the nutrient content of the water. Within the range of the coral’s temperature and salinity tolerance the combination of hard bottom substrates for coral larvae settlement and relevant food transport rates are probably more important than the topography per se.

Seamounts, deep-sea corals, and fisheries on the High Seas: what can we do with almost no data?

Malcolm Clark, Derek Tittensor, Alex Rogers

National Institute of Water & Atmospheric Research, Private Bag 14-901, Wellington, New Zealand

Seamounts are widespread features of the world’s underwater topography, and may number tens of thousands in the Pacific Ocean. They can support high biodiversity and unique biological communities. They are often highly productive, and bottom trawl fisheries target deepwater commercial fish species such as orange roughy, oreos, alfonsino, pelagic armourhead and redfishes. However, seamount habitat is ecologically vulnerable to such exploitation.

In this talk we present results of recent studies by CenSeam (the Census of Marine Life programme on seamounts) that examine the relationships between seamounts, deepwater corals, and fisheries. The known distribution of stony corals worldwide is related to their physical environment, and then applied to potential seamount locations derived from satellite altimetry to estimate the likelihood of the seamount having suitable conditions for corals. Habitat suitability is then related to the distribution and depth ranges of deepwater trawl fisheries to assess their vulnerability.

The North Pacific has a broad band of predicted habitat for stony corals at depths down to 250 m, which becomes more restricted with depth. Habitat suitability in the South Pacific is more widespread through the 750 m to 1250 m depth range. This makes deeper seamounts in the South more vulnerable to fisheries targeting orange roughy and oreos, while the North Pacific seamounts are mainly at depths for species like alfonsino, pelagic armourhead and some of the shallower Sebastes spp. Careful management is required for all these seamount fisheries to avoid overexploitation of the fish stocks, and associated damaging effects of trawling on the coral habitat.

Construction of a habitat map for Heceta Bank, Oregon, USA for use in estimates of groundfish assemblages on the bank

Julia E. R. Clemons¹, W. Waldo Wakefield¹, Curt E. Whitmire¹,

Robert W. Embley², Brian N. Tissot³, Susan G. Merle⁴,

Chris Goldfinger⁵, Christopher G. Romsos⁵

1) NOAA NMFS Northwest Fisheries Science Center

2) NOAA OAR Pacific Marine Environmental Laboratory

3) Program in Environmental Science, Washington State University, Vancouver, Washington

4) Cooperative Institute for Marine Resources Studies,

Oregon State University

5) Active Tectonics and Seafloor Mapping Lab,

Oregon State University

Heceta Bank, (offshore Oregon), is one of the largest rocky banks along the US west coast and contains a diverse array of habitats supporting numerous species of commercially important groundfish, including a diverse assemblage of rockfishes (Sebastes sp.). In 1998 we collected high-resolution bathymetry and backscatter imagery of the bank using a Simrad EM 300 multibeam echo sounder, and returned in 2000 and 2001 to conduct strip transect video surveys of habitat, fish, and invertebrates using the remotely operated vehicle ROPOS. These in situ data have been analyzed for fish habitat relationships. One of the critical elements of this project was to create the first comprehensive lithological habitat map of the bank. Polygons of uniform habitat were constructed by analyzing the image data (bathymetry, backscatter, topographic position index and slope) and reconciling with the video data. Habitat areas identified include: high relief ridge sediment complex, heavily eroded ridge complex, pinnacle, boulder/cobble, and unconsolidated sediment (mud and sand). This map, combined with the fish observations made in the ROV video, may be used as a tool to extrapolate groundfish abundances for the entire bank and adjacent areas surveyed by dive transects.

Sea-Floor Character Maps for the

California State Waters Mapping Program

Guy R. Cochrane

USGS Coastal and Marine Geology Program

Santa Cruz, California

A new raster map product has been produced to describe benthic habitat as part of the California State Waters Mapping Program which will be called a sea-floor character map. The map resolution is 2 meters, identical to that of the multibeam-sonar data from which it is derived, and preserves the gradational qualities of the substrate in a marine environment unlike map products based on delineated polygonal regions. Each pixel is given a value, through a sea-floor video supervised numerical classification that combines information about bottom hardness, rugosity, slope, and depth into a single raster of classes based on current standards used in California fisheries management. Both the raster GIS layer, and a digital map (shown in figure) will be published as part of an online publication that will include other digital maps and associated GIS layers. The digital map folio will include imagery derived from the multibeam-sonar data, bottom-video imagery and observations, sediment thickness isopachs, sub-bottom geology from seismics, surficial geology units and structure, and multi-attribute habitat polygons.

Utility of existing single-beam ‘Digital Survey Bathymetry’ for identifying potential Marine Protected Areas

Roger Coggan, Markus Diesing, Koen Vanstaen

The Centre for Environment, Fisheries and Aquaculture Science (Cefas), UK

This poster demonstrates two topographic images of the central English Channel, derived primarily from single beam acoustic data, known as ‘Digital Survey Bathymetry’ (DSB). Each image covers an area about 100 x 50 km (70 x 30 miles), and reveals remarkable detail of the surface expression of the underlying geology and the erosion and transport features that give shape to the current-day seabed.

Such images are particularly useful in broadscale habitat mapping, allowing seabed interpretation on a regional scale and providing confidence in delineating (mapping) the different seabed facies. This confidence can be critical to effective management.

We have used these images to complement site-specific surveys designed to locate and characterise rocky reef habitats, which will soon be afforded a measure of protection under the EU Habitats Directive. The DSB helped us to accurately delineate areas of the seabed containing rocky reefs, providing an important evidence-base in the process of designating Marine Protected Areas.

Figure 1. Seabed topography in the central English Channel, as determined from Digital Survey Bathymetry based on single-beam acoustic data.

Mapping Rocky Reefs in the Central English Channel: a case study using nested surveys and broadscale acoustic proxies

Roger Coggan, Markus Diesing, Koen Vanstaen

The Centre for Environment, Fisheries and Aquaculture Science (Cefas), UK

The EU Habitats Directive requires member nations to protect a representative proportion of certain marine habitats that occur within their waters. Rocky reefs are one of seven habitats listed for special consideration, but in UK waters they are not well mapped. Without knowing the location and extent of rock outcropping at the seabed surface, it is difficult to establish what constitutes a ‘representative proportion’ of the UK’s rocky reefs.

This paper reports on a search for rocky reefs in the central English Channel. We surveyed two areas where rock was indicated on seabed sediment charts. In one area we found hardly any rock at the seabed surface. In the other we found outcrops over a far more extensive area than expected. The reefs were characterised using a combination of acoustic surveys (multibeam and sidescan sonar) and underwater video and photography. Large sponges were a feature at sites > 50 m depth.

Digital Survey Bathymetry (SeaZone Solutions Ltd.) of the central English Channel provides a fascinating insight into the complex topography of the area (Figure 1). Drawn mainly from single-beam acoustic surveys, it places our multibeam surveys and geological knowledge in a wider spatial context and allows us to delineate a reef area in the order of 1,000 sq km (about twice the size of the Isle of Wight).

Our study demonstrates the considerable potential benefits that existing broadscale topographic data can bring to seabed mapping at the regional scale, by using seabed character as a proxy for habitats identified during site-specific surveys. The resolution and extent of the resulting maps are well aligned with the information needs of regional management and implementing policy objectives of the Habitats Directive.

Figure 1. Main study area, ~40 km south of the Isle of Wight, English Channel. Digital Survey Bathymetry (DSB) overlain with initial acoustic survey track. Insets show multibeam detail overlaying DBS, and photo of sponge covered rock outcrop.

Sponge reefs in the Georgia and Queen Charlotte Basins, British Columbia, Canada: a widespread,

readily mapped and sensitive benthic habitat

K.W. Conway¹, J.V. Barrie¹, G.E. Schlagintwiet²

1) Pacific Geoscience Centre, Geological Survey of Canada,

Sidney, BC

2) Canadian Hydrographic Service, Institute of Ocean Sciences, Sidney, BC

Hexactinellid sponge reefs are commonly found on the western Canadian margin during multibeam surveys in two geographically separate basins where seafloor mapping is ongoing. These reefs develop as a result of framework reef construction by rigid skeleton sponges of the Order Hexactinosida and the trapping by sponges of suspended sediments entrained in bottom currents over centuries or millenia. Previous geological surveys documented four very large, contiguous areas of sponge reefs forming large complexes in the Queen Charlotte Basin (QCB) where reefs are up to 21 m in height and hundreds of km² in area. However, ongoing surveys are discovering smaller areas of reefs as inner shelf areas, such as Agassiz Bank, are mapped. These reefs are found as discontinuous clusters of reef mounds along the eastern edge of the QCB indicating that a belt or zone of sponge reefs may exist bordering the eastern QCB, approximately centered on the 200 m isobath, where ice-contact glacial units remain unburied by recent sediments. This zone of potential sponge reef habitat is approximately 300 km in length.

New reefs have been found on glacial promontories in the Georgia Basin (GB) including several locations in Howe Sound and on Ajax Bank. Multibeam bathymetry and backscatter surveys have proved an effective mapping tool in delimiting sponge reefs and also in discriminating between different types of sponge communities such as flat lying sponge spicule mats, found off Parksville, BC and the more commonly observed framework constructed reef mounds. In certain cases discrimination between living reefs, which show undulatory and non-reflective surfaces, and dead reefs, where more reflective and planar surfaces are observed, may be possible. These new discoveries suggest that far from being an anomaly found only in certain locations of the BC margin, sponge reefs form a widespread habitat type on the seafloor in deep shelf and also inshore locations, and thus represent an important, long term habitat for other species on the western Canadian margin. In many areas these reefs have been impacted by mobile fishing gear with resultant loss of benthic habitat complexity. Mapping of these and other sensitive habitat types will permit more effective fisheries management and will enhance the overall decision making framework for ocean management.

Epifauna and infauna associated with shallow cold seeps and carbonate mounds in Hecate Strait, British Columbia, Canada

Sarah E. Cook¹, J. Vaughn Barrie²

1) Archipelago Marine Research Ltd., Victoria, BC

2) Geological Survey of Canada – Pacific, Sidney, BC

Hydrocarbon seeps with carbonate mounds are present at 130 m depth in Hecate Strait, on the continental shelf of British Columbia, Canada. The community associated with these cold seeps has both infaunal and epifaunal components. The epifauna was surveyed using a Phantom ROV and the video was classified using a georeferenced spreadsheet with all organisms enumerated and identified to the lowest taxonomic level possible. All taxa were relatively common species found on the continental shelf, although some, such as the Oregon triton (Fusitriton oregonensis) were in unusually high abundance for a habitat dominated by soft substrate. These taxa were apparently attracted by the presence of hard substrate in an otherwise soft seabed environment, especially those sessile organisms, such as giant plumose anemones (Metridium giganteum) and encrusting sponges that require hard substrate for attachment. No evidence could be found that linked the presence of any epifauna to the presence of hydrocarbons. The infauna at one of the carbonate mounds was sampled using an IKU grab, including a portion of the mound itself. Infaunal biomass was low and was dominated by bivalves, including the gutless bivalve Solemya reidi, which harbour sulphide-reducing bacteria and are found in hypoxic environments.

Another major characteristic of these cold seeps is the presence of significant shell debris accumulated near the carbonate mounds. It was originally suggested that this indicated a significant infaunal clam community; however, further analysis shows that this shell debris must have been transported from other areas of the continental shelf. Under normal oceanographic conditions these carbonate shells would dissolve into the water column. That the shells are persisting in the vicinity of the carbonate mounds suggests that the chemistry of the near-seabed water has been altered, possibly due to the presence of the hydrocarbon seepage.

Use of video classification techniques to describe and map sponge reef habitat on the continental shelf of British Columbia, Canada

Sarah E. Cook¹, Kim W. Conway², and J. Vaughn Barrie²

1) Archipelago Marine Research Ltd., Victoria, Canada

2) Geological Survey of Canada – Pacific, Sidney, Canada

Glass sponge reef complexes have been mapped using multibeam swath bathymetry in the Queen Charlotte Basin (QCB) and Georgia Basin (GB) on the continental shelf of British Columbia, Canada. These reefs are complex habitats and are host to a diverse array of fauna. They have been shown to play a role as nursery habitats for commercially important rockfish species (Sebastes spp.). The reef complexes can be easily damaged, especially by fishing techniques that contact the seabed such as bottom trawling. From compiled submersible and ROV dives, as well as geophysical survey data and groundfish trawl distribution data it is estimated that about 50% of the known reefs in the QCB have been impacted by bottom trawling. It is unknown if heavily trawled reef areas recover. In most cases the status of the reef complexes (damaged or undamaged) and their associated megafaunal community cannot yet be described using acoustic mapping techniques, so Remotely Operated Vehicles (ROV) have been employed to conduct visual surveys of many of the reefs. Depending on the classification methodology the video can be used to qualitatively or quantitatively describe the fish and invertebrate community of each reef complex.

Analysis of the video taken at reef complexes in the GB indicate that four out of seven surveyed reefs have been mechanically damaged and consist mainly of large areas of scattered dead sponge skeleton fragments and few live reef-building sponges. Relative abundance of the fish and invertebrate megafauna associated with the reefs was also assessed. To determine if reef status impacts the associated community, two adjacent reefs, one damaged and one undamaged, were compared. Higher taxonomic richness, and higher abundance and greater variety of rockfish, both adult and juvenile, were recorded on the undamaged reef. These results are compared with a quantitative analysis of video data from a newly discovered sponge reef off Malcolm Island using a structured, georeferenced database. This information was mapped over the multibeam bathymetry of the reef to allow analysis of the spatial patterns in the distribution of fauna. The strengths and weaknesses of both qualitative and quantitative assessment techniques for the sponge reefs are discussed, and their utility for habitat mapping.

Semi-Automated Classification of Acoustic and Optical Remotely Sensed Imagery in the U.S. Caribbean

B. Costa, T. Battista, C. Menza

NOAA/NOS/NCCOS Center Coastal Monitoring and Assessment, Silver Spring, MD

Benthic habitat mapping supports ecosystem-based management objectives by contributing to the development of detailed species utilization models linking physical habitats with biological information. Marine habitats deeper than 30 meters have been successfully characterized by conducting heads-up digitizing of acoustic and optical remotely sensed imagery. These resulting maps, however, are subjective and ultimately irreproducible because they depend on the accuracy and interpretation of the person that is digitizing. Here we semi-automate the seafloor feature extraction and classification process using high-resolution MBES and LiDAR data as well as underwater images collected off the coast of western Puerto Rico. Alternative approaches were used to identify and extract seafloor features at relevant spatial and thematic scales. The accuracy of these mid to deep-water benthic habitat maps were validated using georeferenced underwater imagery. The ability to quickly and objectively create benthic habitat maps would allow scientists and resource managers to better quantify and assess the changing health of mid to deep-water coral reef ecosystems.

Biological communities of the South West approaches (UK)

Jaime Davies¹, Kerry Howell¹, Heather Stewart², Janine Guinan³,

Emma Verling⁴, Neil Golding⁴

1) University of Plymouth, Plymouth, UK

2) British Geological Survey, Edinburgh, UK

3) Marine Institute, Galway, Ireland

4) Joint Nature Conservation Committee, Peterborough

Habitat mapping is the process by which seafloor geological characteristics are integrated with biological information. It can have an important role in sustainable ecosystem-based management of the marine environment, and aid in the selection and management of marine protected areas.

A survey of the canyons on the continental slope south west of the UK was carried out by the RV Celtic Explorer in the summer of 2007. This survey was undertaken as part of the Mapping European Seabed Habitats (MESH) project. The aim of the survey was to identify and map areas of sensitive reef habitat (Annex I reef habitat).

High resolution multibeam and video ground truthing data were obtained from three canyons on the UK continental margin. Multibeam bathymetry and backscatter response were interpreted in terms of topography and broad scale lithology, and ground truthing sites chosen. Forty-four camera tows were undertaken across the sample area. A drop-frame camera system (stills and video) was used to visually sample the benthic faunal communities and their associated seabed habitat. Images were taken along the tow to provide quadrat-like samples for quantitative analysis of the benthos and representation of the different habitats encountered.

Quantitative analysis of the images were undertaken, all organisms >1 cm were enumerated and identified to the lowest taxonomic level. In addition, images were classified on the basis of substrate type using a combination of Wentworth and Folk scales. Biological data were analysed using PRIMER 6. Cluster analysis with group averaged linkage was performed on square root transformed Bray-Curtis similarity matrix using species count and percentage cover data to identify ecologically coherent benthic communities. Thirteen biotopes (biological communities and associated physical habitat) were identified from combined analysis of biological and physical data. Video footage was reviewed and classified according to newly defined biotope. Each tow was coded with the representative biotopes observed and mapped using ArcGIS. Mapped video tow data was combined with classified acoustic data and habitat maps produced. The spatial distribution of biotopes within the canyons is discussed.

Geology meets biology: mapping and prediction of seabed nature types in the MAREANO regional mapping programme

Margaret F.J. Dolan¹, Valérie Bellec¹, Pål Buhl-Mortensen²,

Lene Buhl-Mortensen², Terje Thorsnes¹, Reidulv Bøe¹

1) Geological Survey of Norway, Trondheim, Norway

2) Institute of Marine Research, Bergen, Norway

MAREANO is a regional programme which includes mapping of surficial geology and benthic habitats, or nature types, in the southern Barents Sea, northern Norway. It is a cooperative venture between several institutions with the Geological Survey of Norway, the Institute of Marine Research and the Norwegian Hydrographic Office as major partners.

Mapping of seabed nature types is an important activity in regional mapping programmes like MAREANO and one which lies at the interface between geology and biology. Interpreted maps provide multidisciplinary information which can be essential for the sustainable management of seabed resources.

We provide a summary of ongoing research related to mapping and prediction seabed nature types including mapping of surficial geology, benthic fauna and interpretation of seabed processes. Case studies from the MAREANO area are presented to give examples of the various areas of research. Highlights and ongoing research themes include:

•       The importance of multibeam bathymetry and backscatter data and derived variables in nature type mapping and modelling.

•       Integration of biological information and sediment ground-truth data from video observations.

•       Multivariate modelling of nature types - the roles of supervised and unsupervised classification.

•       Delivering useable map-products - making the transition from paper to web-based mapping, issues related to data, scale and resolution.

Spatial modelling and multivariate prediction of seabed nature types in the MAREANO regional mapping programme

Margaret F.J. Dolan¹, Valérie Bellec¹, Pål Buhl-Mortensen²,

Lene Buhl-Mortensen², Terje Thorsnes¹, Reidulv Bøe¹

1) Geological Survey of Norway, Trondheim, Norway

2) Institute of Marine Research, Bergen, Norway

MAREANO is a regional programme which includes mapping of surficial geology and benthic habitats, or nature types, in the southern Barents Sea, northern Norway. It is a cooperative venture between several institutions with the Geological Survey of Norway, the Institute of Marine Research and the Norwegian Hydrographic Office as major partners. A comprehensive, hierarchical survey strategy has been adopted. Multibeam surveys provide 100% mapping of the seabed bathymetry and backscatter. Targeted seabed sampling is conducted over representative portions of the seabed, but covers a fraction of the total area. Video data provide efficient high-resolution imagery of the sea floor and permit the identification of seabed sediment type (grain size) and the analysis of benthic fauna. These 'ground-truth' data, together with some physical samples and any available historical data provide the basic suite of information from which sediment grain size distribution maps and nature types are developed.

Since they provide full spatial coverage, the multibeam data (bathymetry and derived multi-scale terrain variables, plus backscatter data) provide a good basis for the development of models of the distribution of seabed properties. The data provide basic information related to the seafloor geology which underpins the distribution of benthic habitats or nature types. We summarise current research under the MAREANO project into multivariate segmentation and classification of the seabed using these data in order to map nature types. The roles of supervised and unsupervised classification will discussed, with particular reference to regional mapping and use of classification results in planning ground-truth campaigns.

Addressing issues of scale, we have investigated the use of multibeam data at several resolutions, with derived terrain parameters across a range of spatial scales, in order to find suitable and practical predictor variables for regional mapping of surficial sediments and nature types at scales relevant to management. We compare the results of these automated analyses with 'expert' derived end-product maps and discuss the role of each approach in the context of a regional project like MAREANO, which focuses on the delivery of web-based maps for a wide range of end-users.

Geological basis of large gorgonian coral habitat in Atlantic Canada

Evan N. Edinger¹, Owen A. Sherwood², Kent Gilkinson³,

Vonda E. Wareham³

1) Geography, Memorial University of Newfoundland

2) Biology, Memorial University of Newfoundland

3) Northwest Atlantic Fisheries Centre, Fisheries and Oceans Canada

Deep-sea corals in Atlantic Canada comprise 30+ species of gorgonian, alcyonacean, scleractinian, and antipatharian corals. Most of these corals occur at continental slope depths along the continental margin of Nova Scotia, Newfoundland, Labrador, and Baffin Island. Because most of the longest-lived coral species require specific substrates, surficial geology exerts a primary control on the distribution of the most sensitive deep-sea corals in Atlantic Canada. The geological features that provide coral habitat were observed in-situ from the ROV ROPOS during 2006 and 2007. Broader coral distribution data are derived from DFO trawl surveys and fisheries observer records.

In-situ observations of the surficial geology of coral habitat suggest that upper slope till and till-tongue deposits and margins of submarine fan distributory channels provide hard substrates for deep-sea corals along the upper continental slope. Moraine-derived till provides the primary habitat for corals in the Fundian Channel (1) and the Stone Fence (southwestern margin of the Laurentian Channel, Nova Scotia). Other coral sites with reported stacked tills in Labrador include the Hudson Strait and, probably, the Cartright Saddle. These stacked tills are associated with the locations of principal ice streams during the deglaciation of Atlantic Canada (2). Coral concentrations in the Haddock Channel (Newfoundland) are dominated structurally by the isidid gorgonian Keratoisis ornata, and are developed on deep-water till-tongues derived from glaciomarine debris flows. Other coral concentrations in deep water near shelf-crossing troughs along the Newfoundland and Labrador margin, such as sites along the southwest Grand Banks, northeast Newfoundland shelf-edge, and southern-central Labrador sites including Hawke Channel, Hopedale Saddle, and Ogak Saddle, probably have a similar geological basis.

Large gorgonian coral habitats in the Sable Gully consist primarily of eroded friable Tertiary mudstone, and are dominated by Keratoisis ornata, Primnoa resedaeformis, and Paragorgia arborea. In the upper reaches of the Gully, boulders of crystalline basement rock were observed, with abundant coral growth on them. Shear strength of the mudstone appears to limit size, shape, and species composition of deep-sea corals in many parts of The Gully.

Multibeam bathymetry and backscatter data have been collected for the Fundian Channel, Sable Gully, Stone Fence, and Halibut & Haddock Channels. Further in-situ observations of remaining deep-sea coral concentrations in Newfoundland and Labrador waters will be coupled with multibeam sonar and 3.5 kHz sub-bottom profiling.

(1) Scott, 2003, GSA NE Section, Abstracts with Program 35(3): 27.

(2) Shaw et al. 2006, Quaternary Science Reviews 25: 2059-2081.

Marine habitat mapping in the Gilbert Bay Marine Protected Area, Labrador, Canada

Evan N. Edinger¹, Alison Copeland¹, Rodolphe Devillers¹,

Philippe LeBlanc¹, Joseph Wroblewski²

1) Geography & Biology, Memorial University of Newfoundland

2) Ocean Sciences Centre, Memorial University of Newfoundland

The Gilbert Bay Marine Protected Area in southeastern Labrador, Canada, comprises approximately 60 km² of fjord and related habitats. This MPA was designated to protect a genetically distinct resident population of Atlantic cod (Gadus morhua), the Golden Cod. The management plan for the MPA allows scallop dragging and limited fishing for species other than cod in approximately ½ of the bay. Substrates and marine habitats throughout the bay were mapped using multibeam sonar – derived bathymetry and backscatter, and slope derived from bathymetry, and ground-truthed using drop video and grab samples. Shallow portions of the bay not covered by multibeam were mapped using GPS-single-beam-sonar, drop video, low-speed towed video, and grab samples, with manual interpolation between survey lines. Faunal composition of substrates was compared using multidimensional scaling, analysis of similarity, and similarity percentage analysis in PRIMER, with video data and grab data considered separately.

Six substrate classes were identified that had uniquely identifying combinations of depth, backscatter intensity, and slope based on multibeam sonar. The substrate classes were: muddy-gravel, sandy-gravel, coralline-algal-encrusted gravel, gravelly-sandy-mud, gravelly-mud, and mud. Two additional gravelly substrates were identified based on video: nearshore gravel, and current-swept gravel, but the current-swept gravel did not have unique acoustic characteristics. Bedrock wall substrates were rare, being replaced by coarse, steeply-sloping talus accumulations. Moraine-derived sills, submerged eskers and other glacial features largely control substrate distributions within the bay.

Highest biodiversity was found in the coralline-algal-encrusted gravel substrates. This substrate class included rhodoliths formed by branching Lithothamnion glaciale as well as thin to thick crusts of the encrusting coralline red algae Clathromorphum sp. on coarse gravel. Species composition of biota on this substrate class was statistically distinct from all other substrates. Gravelly substrates all had similar faunal composition, with the exception of the sponge-rich current-swept gravel, which also contained the soft coral Gersemia rubiformis, and the shallow ice-scoured gravel, which was distinct due to its low diversity and prominence of Mytilus edulis. Faunal composition of muddy substrates was statistically indistinguishable. The number of mappable habitats was lower than the number of substrate types.

Marine habitat mapping combined with the current management plan for the MPA shows that more than half of the highly sensitive coralline-algal-based habitat is now protected from fishing activity, but remaining areas of coralline-algal habitat are impacted by scallop dredging. Similarly, most, but not all, of the areas with highest habitat diversity are already protected from fishing activities.

Identifying errors that occur when applying drop video techniques

for marine phytobenthos surveys

Jan Ekebom¹, Mats Westerbom²

1) Metsähallitus, Natural Heritage Services, P.O. Box 94,

FI-01301 Vantaa, Finland

2 Novia University of Applied Sciences, Raseborgsvägen 9,

10600 Ekenäs, Finland

Underwater video surveys have in recent years become a popular method for carrying out marine phytobenthos surveys (inventories of marine benthic macroscopic vegetation). Video surveys make it possible to identify e.g. the seafloor geology, geomorphology, habitats and sessile macroscopic species. The results from video surveys have wide applications ranging from marine ecological research to marine spatial planning which can apply survey maps for planning and management of marine protected area and for sustainable planning of human activities. In recent years the ecosystem approach to management of human activities has been emphasize by the Convention on Biological Diversity as well as the European Union which further underline the importance to acknowledge the marine biota in sea use planning. The spatial plans for marine areas will have a great socio-economic impact that last for a very long time. Consequently, the availability of good biological datasets of high quality becomes crucial for successful marine spatial planning. This study identifies errors that occur when applying drop video techniques for underwater surveys of marine vegetation. The study is based on large datasets from the northern Baltic Sea as well as published papers. Consequences of the identified errors on marine spatial planning are discussed.

Diverse Habitats on Intraoceanic Island Arc Submarine Volcanoes of the Mariana Arc, Western Pacific

Robert W. Embley¹, William W. Chadwick, Jr.², Susan G. Merle²,

Verena Tunnicliffe³, John Dower³, S. Kim Juniper³

1) NOAA/Pacific Marine Environmental Laboratory

2) Cooperative Institute for Marine Resources/Oregon State University

3) School of Earth and Ocean Sciences, University of Victoria

At least 30,000 seamounts have been constructed on the ocean floor over the past 175 million years by submarine volcanic processes. The majority of these are composed of basaltic lavas associated with Mid-Ocean Ridge, hotspot or mid-plate volcanism. A subset of seamounts submarine arc volcanos, which are located within the intraoceanic subduction zones of the western Pacific and Atlantic oceans. Although there are fewer than 1000 of these volcanoes their geologic framework is fundamentally different than other seamounts. The primary differences are: (1) diverse lava types, (2) high amounts of magmatic gases which produce more explosive fragmental eruptions, and (3) long-term volcanic and hydrothermal activity. The volcanic and hydrothermal activity commonly occurs at much shallower water depth (<1000 m) than that occurring on the Mid-Ocean Ridge (almost entirely >2000 m). Three basic types of seafloor habitat are found on these volcanoes; 1) effusive lava flows, (2) fragmental slopes and (3) hydrothermal vents. A 2001-2004 systematic sidescan and multibeam survey of the Mariana Arc provides a basic habitat map of about 50 of these volcanoes and ROV dives on 14 of them provide a preliminary habitat assessment. Lava outcrops on the summits and flanks of the volcanos provide habitat for a diverse benthic taxa. The coarse-grained volcaniclastic deposits have sparse larger sessile taxa although erosion by mass-wasting processes can expose semi-lithified clastic substrata that support fixed benthos. Many of these volcanoes are hydrothermally active (20 out of 50 in the Mariana arc) and support chemosynthetic communities marked by a high diversity between volcanoes and relatively low diversity on individual volcanos. The chemosynthetic communities have evolved in extreme physical and chemical conditions, including those adapted to long-term volcanic activity and low pH induced by high concentrations of CO₂ and Sulfur species. Because the summits of these volcanos are often in shallow depth, many have overlap between photosynthetic and chemosynthesis. In summary, the submarine volcanos of the intraoceanic island arcs comprise a unique suite of open-ocean habitats found nowhere else on Earth.

Explorations of cold-water coral-habitat relationships on Cordell Bank, CA using submersible visual data and multibeam sonar data

Lisa L. Etherington, Pamela van der Leeden

NOAA, Cordell Bank National Marine Sanctuary

This work represents an exploratory analysis of in situ data collected with the Delta submersible on the prominent geological feature of Cordell Bank by relating stylasterid coral (Stylaster californicus) presence to various seafloor features, as measured by multibeam acoustic sampling. It has been recognized that in situ observational data can provide critical information in understanding deep, cold-water coral distribution patterns and habitat associations. These data coupled with fine-scale seafloor mapping provide robust assessments of habitat relationships and allow predictions of species occurrence.

Cordell Bank is situated on the edge of the continental shelf of California and rises from 130m to within 35m of the surface. The Bank contains diverse habitats including high relief rock, flat bedrock pavement, boulders, mixed cobble and sand, sand ripples, and mud. Multibeam acoustic data of Cordell Bank were collected and bathymetric products were created by California State University-Monterey Bay, while substrate analyses were conducted by CSUMB in collaboration with U.S. Geological Survey. These data indicate that 29% of the bank is composed of hard substrates, 48% is mixed substrates, and 23% is soft substrates. The majority (74%) of Cordell Bank is comprised of sloping (1-30°) habitats, while areas with a slope of <1° make up 25% of the Bank, and steeply sloping (30-60°) and vertical (60-90°) habitats make up a very small proportion of the Bank.

Extensive video coverage in representative habitats of the Bank allow for comprehensive assessment of coral distribution and habitat associations. Coral presence is modeled as a function of various habitat features, including depth, slope, rugosity, aspect, topographic position index (tpi), and substrate using data generated from multibeam acoustic sampling. Results indicate that corals are not widely distributed across the Bank, but instead are found in limited habitat types, particularly shallow, hard, high sloping areas. These results provide important information for understanding fine-scale habitat associations of corals and provide the foundation for monitoring these communities. In addition, these habitat models aid in our ability to make informed management decisions regarding these sensitive and diverse communities within a national marine sanctuary.

Predicting suitable habitat for deep-sea coral reefs

in British Columbia

Jessica L. Finney¹, E.J. Gregr², S. Patton³

1) School of Resource and Environmental Management,

Simon Fraser University

2) SciTech Consulting

3) Canadian Parks and Wilderness Society British Columbia

Deep-sea coral reefs provide valuable habitat for fish and other organisms but are highly susceptible anthropogenic threats. The most serious threats are posed by benthic fisheries, most notably bottom trawling, and other human activities that disturb the seafloor. Recent surveys of deep-sea coral reefs suggest that they have been damaged or destroyed in virtually all parts of the world. To date, most deep-sea corals in Canada are unprotected. There is a particularly pressing need to develop a conservation strategy for deep-sea corals in British Columbia (BC), but it is not a simple task. Relatively little coral research has been conducted in BC, and the distribution of deep-sea reefs remains largely unknown. The vast majority of sightings are derived from bycatch data from the trawl fishery. These data are recorded on a taxonomically coarse scale as the on-board observers are poorly trained in coral identification. Furthermore, these data only provide information on the locations of previously damaged corals. Determining the distribution of deep-sea corals and identifying pristine coral aggregations is a crucial first step in establishing effective ocean management strategies to protect these rare and valuable habitats. This study makes a significant contribution toward addressing these information gaps by compiling known locations of corals and using predictive habitat modeling techniques to map locations of potentially suitable coral habitat along the BC coast. Suitable habitat will then be used as a proxy for potential coral locations in previously unsampled areas in BC. The results of this study will be useful in guiding future research and conservation efforts.

Mapping and Lithologic Interpretation of the Territorial Sea, Oregon

C. Goldfinger¹, M. Agapito¹, C. Romsos¹, T. Haddad²,

K. Karageorge¹, R. Dana²

1) Oregon State University College of Oceanic

and Atmospheric Sciences

2) Oregon Department of Land Conservation and Development

On September 18, 2006, the Governors of California, Oregon and Washington signed the “West Coast Governors’ Agreement on Ocean Health”. One of the seven priority areas identified in the agreement is the expansion of ocean and scientific information. Additionally, Oregon’s Ocean Policy Advisory Council (OPAC) in its public workshop held in July 16-17, 2007 proposed seafloor mapping as part of its action plan to expand ocean and coastal scientific information in preparation to establishing marine reserves in Oregon. Moreover, previous high-resolution mapping efforts in Oregon’s territorial sea have mapped only five percent of state waters. In response to this, the Active Tectonics and Seafloor Mapping Laboratory (ATSML) at Oregon State University (OSU) College of Oceanic and Atmospheric Sciences (COAS) initiated mapping of the territorial sea using various thematic GIS layers.

Recently, approximately 11,000 bottom sample labels from National Oceanographic Surveys (NOS) hydrographic smooth sheets in Oregon were digitized. This newly discovered dataset of sample points added a significant amount of information to the previously available 305 bottom samples used to characterize the sedimentary lithology of the territorial sea. These data were collected by the Coast and Geodetic Survey using leadlines and traditional navigation methods, and extend from the late 1800’s to the modern. CGS surveys during that period were more explorational than the purely safety of navigation surveys that NOS performs today, and thus broad areas of coastline between ports were sampled. We collected these data from “smooth sheets”, interim mapping sheets on which field data were compiled. High resolution scans of these sheets were rectified, and the data were captured using heads-up digitizing. Navigation using bearings and horizontal sextant angles was surprisingly good where we could compare the position of offshore rocks to modern data, and typical maximum errors in these cases were ~ 50 m. While not ideal by modern standards, the sheer volume of available sample data have made it possible to construct a preliminary Surficial Geologic Habitat (SGH) map for the Oregon Territorial Sea. In addition to the lithologic point layer, this mapping work uses many other updated thematic GIS layers to interpret the extent of lithologic points such as kelp points and polygon layers, (interpreted as proxies for rocky substrate), exposed rock polygons digitized from 0.5 meter resolution aerial photos, the onshore Oregon digital geologic map, and a TIN model derived from NOS soundings. Rugosity of the TIN layer was considered in the interpretation, along with lithology and the kelp proxy to create the lithologic map. The new inshore map was then merged with the current regional habitat map of the US west coast.

We expect this mapping effort to provide the most detailed surficial geologic interpretation of territorial sea to date. It will also provide a new map product that can be used during the nomination and evaluation process for marine reserves as well as in the overall reserve design process for Oregon.

Surveying deep-water habitats on the UK shelf edge

to inform MPA selection

Neil Golding¹, Heather Stewart², Jaime Davies³, Janine Guinan⁴,

Kerry Howell³, Emma Verling¹, Viv Blyth-Skyrme¹

1) Joint Nature Conservation Committee, Peterborough, UK

2) British Geological Survey, Edinburgh, UK

3) University of Plymouth, Plymouth, UK

4) Marine Institute, Galway, Ireland

A survey of submarine canyons in the South-West Approaches areas of the UK Continental Shelf edge was undertaken in June 2007 as part of the MESH (Mapping European Seabed Habitats) project. This collaborative survey involved the Joint Nature Conservation Committee, the Marine Institute, the British Geological Survey and the University of Plymouth. Starting at around 200 m water depth, the canyons plummet to the abyssal plain over 4000 m below. Previous work (Graham et al, 2001) has indicated the occurrence of ‘potential reef’ in the area surveyed, as defined by Annex I of the EC Habitats Directive. Habitats listed in Annex I of the Directive are considered to be under threat in the European Union because they are in danger of disappearance or have a restricted range in Europe.

The survey successfully acquired high resolution multibeam and sub-bottom profiler data (GeoHAB abstract Stewart et al., 2008) as well as drop-camera data (digital video and stills). A variety of deep-water habitats were observed. The canyon interfluves, or canyon tops, were dominated by mixed substrata and ‘historic’ coral mounds. Living cold water coral reef communities were observed on the flanks of the canyons between 800 m and 900 m water depth. Significant anthropogenic impacts were observed with the camera drop-frame. Evidence of fishing activity (discarded nets and lines) and rubbish was observed across the survey area. Plastic bags were also observed wrapped around the reef structure.

Information provided by surveys such as this can help inform MPA selection. For MPAs designated under the EC Habitats Directive, a number of selection criteria should be met. In addition, such information gathered should confirm that the area hosts the Annex I habitat, enables the identification of the specific Annex I habitat sub-type and provides a picture of the biological community present on the Annex I feature.

A comparison of the habitat structure and ecology associated with cold water coral reefs at the Mingulay Reef Complex (Outer Hebrides) and the Sula Ridge Reef Complex (Norway)

Sophie L. Green¹, Veerle A.I. Huvenne¹, Andrew Davies²,

Veit Hühnerbach¹, J. Murray Roberts², Andre Freiwald³

1) Geology & Geophysics, National Oceanography Centre, Southampton, UK

2) Scottish Association for Marine Science, Dunstaffnage, UK

3) Institut für Paläontologie, Universität Erlangen, Germany

The existence of cold water corals has been identified in a range of latitudinal, hydrodynamic and bathymetric settings. This study involved the application of side scan sonar data and camera footage from a Remotely Operated Vehicle and a manned submersible to establish the similarity of two relatively shallow (~150 – 320 m) aphotic reefs, one at Mingulay in the Outer Hebrides, off Scotland, the other the Sula Ridge Reef Complex on the Norwegian margin.

Combination of the video data and side scan sonar using ArcGIS suggests similar habitat structure at these sites with a characteristic pattern of homogeneous background sediment overlain by coral rubble and topped with dead and living coral framework, dominantely Lophelia pertusa. Similar proportions of each habitat type were recorded within comparable mapped areas at the 2 sites. Ecological assessment also suggests analogous distinctive ecological assemblages may be associated with the individual habitat types recorded at both sites. A tentative estimation of the biodiversity showed that the diversity at the Sula Ridge Reef Complex may be higher than that recorded at Mingulay. However, the relative difference between species richness on and off the coral framework appears lower than expected. This may partly be the result of the limited quality of some of the video data, and of the simplicity of the methodology applied.

Anthropogenic activities can have detrimental impact on cold water coral reefs. Trawling, oil exploration and ocean acidification are all potentially destructive. However, video footage studied in this project did not contain evidence for physical destruction. This led to the suggestion that wider scale studies with a regular sampling period are needed to fully assess impacts. The importance of finding indicators for ecosystem stress, such as key species, was also noted in order to optimise management of cold water coral reef ecosystems.

Geology as a Surrogate to Ecology – Is this Possible?

Examples from Alaska

H. Gary Greene¹, Cleo K. Brylinsky², Victoria M. O’Connell³,

Jennifer Reynolds⁴, Sean Rooney⁴, Jon Heifetz⁵

1) Center for Habitat Studies, Moss Landing Marine Labs, and Tombolo, Orcas Island, WA

2) Alaska Department of Fish and Game, Sitka, AK

3) Coastal Marine Research, Sitka, AK

4) University of Alaska Fairbanks, AK

5) NOAA/NMFS, Auke Bay, AK

Using geology to define and map seafloor conditions that in turn are used to characterize potential marine benthic habitats is becoming a common method in trying to define ecosystems over broad areas of the ocean. With the rapid decline in fisheries worldwide and the intense pressure to establish Marine Protected Areas (MPAs) in an attempt to arrest, or even reverse, this decline, scientists and managers alike are struggling to establish methodologies and standards to map habitats that accurately reflect the existing ecosystems. Driving this effort is the use of modern technology such as acoustic Multibeam Ecosounder (MBES) systems, which can produce high-resolution images of the seafloor over large areas in fairly rapid time. Therefore, the inclination is to map the seafloor geomorphology, lithology, and structure in such a fashion that these can be defined as potential habitats that may represent a particular ecosystem.

The inherent problem with trying to define habitats based on remotely sensed acoustical geophysical data is that only the physical conditions of the seafloor are obtained and the other components (temperature, currents, nutrients, etc.) that comprise a habitat are not readily available. Some of these components can be interpreted from the physical data, but generally the biological and chemical information is lacking. Therefore, certain assumptions are included in any geological surrogate that is produced for a particular ecosystem. In addition, the problems of false heterogeneity and false homogeneity play a role in using geological surrogates as predictors of potential habitats. To overcome such problems it is important to statistically groundtruth the interpreted potential habitats in many areas, a procedure that has been done in the characterization of rockfish habitats in Alaska. Potential marine benthic habitat maps produced in Alaska for the management of demersal groundfish are based on geologically interpreted MBES and side-scan sonar data and groundtruthed using the submersible Delta, which were used to determine the quantity and quality of commercially targeted groundfish habitat. Although these maps are not perfect and may not characterize the diversity of habitats in Alaska waters, they do provide a base upon which ecosystem-based habitat characterization can be built once the biology is better known. It will be shown that geological surrogates for groundfish fisheries in Alaska are useful predictors of habitat types and that the surrogates can be used elsewhere.

An ecological classification of benthic habitat in

Pacific Canadian shelf waters

Edward J. Gregr

SciTech Environmental Consulting

Glen Jamieson, Fisheries and Oceans Canada

The classification of habitat depends on the correct translation of physical characteristics into measures of habitat suitability. Southwood’s (1977) definition of the habitat template was initially proposed for classifying the landscape. The habitat template is intended to group life history traits of species according to their functional role with respect to Disturbance and Adversity, the two axes that define the template. The approach was applied to benthic marine habitat in Atlantic Canada. Kostylev et al. (2005) showed that the template, when divided into four quadrants, was significantly correlated with species diversity. We applied the method to Pacific Canadian waters and, based on available physical data, developed local definitions of Disturbance and Adversity. We assessed how the resulting classification (Figure 1) corresponded to distributions of a number of species including corals, sponges, and commercially important bottom fish. We also explored how the use of discrete categories (i.e., four quadrants) created from the continuous classification influenced the identified biological relationships. The ecological classification of benthic habitat on Canada’s west coast shelf will support ecosystem management efforts by providing a representative view of available habitat, and a means of estimating species diversity and community structure. Our results advance our understanding of Canada’s Pacific coast, and represent a novel, ecosystem-based approach to marine classification.

Southwood, T.R.E. 1977. Habitat, the templet for ecological strategies? Journal of Animal Ecology 46: 337–365.

Kostylev, V.E., et al. 2005. Characterization of benthic habitat on northeastern Georges Bank, Canada. In P. W. Barnes and J. P. Thomas, (eds). Benthic habitats and the effects of fishing. American Fisheries Society Symposium 41: 141–152.

Figure 1: Habitat template for British Columbia shelf waters. Graphs show the distribution of the 500 x 500 m² grid cells across the Adversity and Disturbance axes. Red lines show how the continuous distribution is divided into four quadrats.

Using models to predict deep-sea cold-water coral habitats: case studies from the Irish continental slope

Janine C. Guinan^1,2, Margaret F.J. Dolan^1,3

1) Department of Earth and Ocean Sciences, National University of Ireland, Galway, Ireland

2) Marine Institute, Galway, Ireland

3) Geological Survey of Norway, Trondheim, Norway

Cold-water corals are an important and fragile component of the benthic habitat in the northeast Atlantic Ocean, and the scleractinian corals Lophelia pertusa and Madrepora oculata occur frequently along the Irish continental margin. Following the Irish National Seabed Survey (INSS), multibeam data provide bathymetric mapping of Ireland's entire offshore territory. The occurrence of corals, however, has only been documented over a fraction of this area. Remotely Operated Vehicle (ROV) surveys conducted at several sites of particular scientific interest, including carbonate mounds, have provided much valuable data on coral biology, associated fauna and reef-related sedimentary processes, but such surveys are by nature limited in their extent. The true distribution of the corals remains largely unknown, yet this information is required for sustainable management. Predictive modelling of coral distribution may help to bridge this gap, but models must be based on sound science, and appropriate methodologies.

As a precursor to developing predictive models for the distribution of cold-water corals, we used quantitative methods to investigate the relationships between observed coral distribution and terrain attributes (e.g., slope, aspect, rugosity, and bathymetric position index) derived from multibeam bathymetry data. We demonstrate a method for estimating deep-water coral percentage cover from ROV-based video footage, using data from two carbonate mound provinces on the Irish margin.

Two methods for modelling coral habitat suitability have been assessed: 1) Genetic Algorithm for Rule-set Production (GARP), and 2) Ecological Niche Factor Analysis (ENFA). Models of suitable coral habitat distribution were developed across a range of spatial scales. Our results provide insights into the usefulness of these modelling techniques and highlight several issues relating to the input data. GARP was used to predict cold-water coral habitat in deep water, beyond the extent of existing samples. Its predictive ability at local and regional scales was examined, as was the relative importance of multi-scale predictor variables including terrain and oceanographic parameters. ENFA was used to provide comparisons with the GARP models and gave additional insight on the predictor variables. ENFA was also used in a detailed study to predict locally suitable coral habitat on a single carbonate mound in relation to terrain variables derived from ROV-based multibeam data.

Sitka Sound and Beyond – Serving Up 40,000 km of

Coastal Habitat Mapping and Imagery

Jodi Harney¹, John Harper¹, Steve Lewis², Mandy Lindeberg²,

K. Koski³, Sue Saupe⁴, Mary Morris⁵

1) Coastal & Ocean Resources Inc., Sidney, BC

2) National Marine Fisheries Service, NOAA, Juneau, Alaska

3) The Nature Conservancy, Juneau, Alaska

4) Cook Inlet Regional Citizens Advisory Council, Anchorage, Alaska

5) Archipelago Marine Research Ltd., Victoria, BC

The ShoreZone coastal habitat mapping system was developed in the early 1980s and has been widely applied in the Pacific Northwest, including the entire 40,000 km of British Columbia coastline and 5000 km of the Washington coastline. ShoreZone mapping in Alaska began in 2001 and presently encompasses 40,000 km of shoreline in the Gulf of Alaska, including the xxx km of Sitka Sound. The Gulf Alaska shoreline mapping is expected to be complete in 2010 which will result in a 100,000 km contiguous dataset (Columbia River to the Bering Sea).

A priority of the Alaska ShoreZone program has been to make both the imagery and data widely accessible. At present, the 40,000 km of low-tide imagery is provided via two ArcIMS websites, which reflect the evolution of data delivery. The websites allow users to “fly” most of the Gulf of Alaska shoreline. Benthic habitat characterization is displayed through a variety of themes, including: mapped physical attributes (e.g., substrate type, morphology), mapped biological attributes (e.g., wetlands, eelgrass, canopy kelps) and mapped man-made features (e.g., docks, piers, seawalls) and some derived data layers (e.g., wave exposure, habitat type, oil spill residence). Visibility of the websites has been enhanced not only by presentations at professional meetings but also numerous community presentations (e.g., Kodiak, Homer, Anchorage, Cordova, Ketchikan) as well as interviews with community newspapers (e.g., Sitka Sentinel) and radio stations (e.g., Sitka Raven Radio). A specialized school curriculum was developed to use the web-based Alaska ShoreZone dataset in high school science programs (e.g., Homer).

Revealing Sidney’s Bottom – Seabed Habitat Mapping for the Community of Sidney, BC

John R. Harper¹, Brian D. Bornhold¹, Sheri Ward¹,

Sarah Cook², William C. Austin³

1) Coastal & Ocean Resources Inc., Sidney, BC

2) Archipelago Marine Research Ltd., Victoria BC

3) Marine Ecology Centre, Sidney BC

Nearshore habitat mapping was conducted off the waterfront of Sidney, British Columbia in support of community planning initiatives. The survey area extended from the intertidal zone to depths of approximately 10 m and involved classification of 0.8 km² of seabed along the shoreline of Sidney. Side-scan sonar with 100% overlap was used to delineate seabed sediment texture and man-made features on the seabed. Video imagery was acquired using towed video (the Seabed Imaging and Mapping System or SIMS); the video survey grid resulted in imagery of approximately 2% of the project site. Approximately 35,000 video images were classified for biology, geology and man-made artifacts using the SIMS classification system. Seabed habitat units were defined based on (a) the acoustically-defined substrate units (from side-scan) and (b) the biotic assemblage present in each unit. The five biophysical habitat units mapped are: (1) sand-pebble with eelgrass, (2) intertidal and nearshore rock with rockweed & bladed kelps, (3) sand-pebble with filamentous red algae, (4) rocky reefs with large bladed kelp and (5) dense clay with piddock clams. The use of combined substrate and biota as the habitat descriptor provides a good index of the habitat structure and ecological function.

As part of a 2007 Oceans Day event, the Sidney Marine Ecology Centre planned a Revealing Sidney’s Bottom event to publicly present the habitat maps. In the weeks preceding Oceans Day, the recreational diving community and local whale watching operators were engaged to help identify unresolved targets on the seabed. A newly opened, waterfront hotel hosted the event as part of their opening promotions. Over 150 people, including the mayor and council, attended the presentations with a standing-room only crowd. In addition to the presentation of habitat mapping information, local divers presented pictures of the seabed and biota, and the Canadian Hydrographic Service formally released a new chart of the area, which included the location of a shipwreck that was found during the survey.

The Sidney’s Bottom event was designed to reach and engage people who are seldom exposed to seabed mapping information including, recreational divers and boaters, school children, the mayor and council and local residents, who were just curious as to what the seabed looks like. The diversity of information presented (maps, habitat visualizations, videography, historical photos of the waterfront, new charts, diver photos and projected microscope images) contributed to the “engagement” of the community.

Global ocean conservation priorities for benthic ecosystems identified by GIS analysis of multiple spatial data layers

Peter T. Harris, Tanya Whiteway

Geoscience Australia, GPO Box 378, Canberra ACT 2601, Australia

At GeoHab 2007 in Noumea we presented a global map of seafloor geomorphology and suggested ways it could be used to identify the spatial distribution of broad-scale benthic habitats having high conservation value, namely seamounts and hydrothermal vents. Here we present a new global seabed classification based on the multivariate analysis of bathymetry (depth and seabed slope), sediment thickness, bottom water temperature, bottom water dissolved oxygen and surface primary production. Each of these variables exerts control over the benthic ecology and affects the type of community that might occur in any given location. Multivariate analysis of maps for these 6 variables identifies 11 unique “seascapes” that covers the off-shelf areas of seafloor for the entire globe excluding the Arctic Ocean. The Atlantic, Pacific, Indian and Southern Ocean basins are each characterised by seascapes occurring in different proportions of surface area and there is a broad correspondence between certain seascape assemblages and seabed sediment type and geomorphology. The Arc GIS “focal variety” function was used to carry out an analysis of the seascapes to identify areas having the greatest spatial heterogeneity of seabed types. We suggest such areas of high heterogeneity may be expected to contain the greatest biodiversity within the smallest surface area, which provides useful information for the design of a global network of representative high-seas marine protected areas.

Prediction of Australia’s benthic marine habitat diversity using seascapes: a national map

Andrew D. Heap, Scott L. Nichol, Tanya Whiteway

Geoscience Australia

One of the biggest challenges facing marine scientists is the development of a robust and defensible way to represent potential seabed habitats and ecosystems, based on easily mapped and spatially-abundant biophysical properties. Geoscience Australia is currently developing a statistical approach to integrate multiple spatial biophysical data layers into a single map of seabed habitats or “seascapes” for Australia’s marine region. A total of 13 ecologically-meaningful seascapes were defined on the shelf and nine defined for off-shelf areas. Seascapes were delineated most strongly by variations in seabed sedimentology, rugosity and temperature. Seascapes on the shelf divide into two broad latitudinal groups. The northern group are characterised by muddier, warmer environments and shallow water relative to the southern group, which is characterised by sandy, cooler environments. Interestingly, seascapes dominated by gravel occur only on the southern margin. Seascapes off the shelf lack a distinct latitudinal pattern and are more related to seafloor temperature as a function of depth. For deep ocean environments on the southern and western margins, seascapes are principally defined by rugosity and primary production. For other areas, seascape distribution is more complex, with bathymetry and slope emerging as key drivers. Areas where a network of marine protected areas could maximise biodiversity coverage by protecting the maximum seascape heterogeneity are identified using a focal variety analysis in ARCGIS. Future research includes working with Australia’s marine biologists to correlate the seascapes with high-resolution biological data.

Structural Sponge Communities of the Faeroe-Shetland Channel,

N.E. Atlantic: Preliminary observations

Kerry L. Howell¹, Jaime S. Davies¹, David J. Hughes²,

Bhavani E. Narayanaswamy²

1) Marine Institute, University of Plymouth, UK

2) Scottish Association for Marine Science, UK

A high resolution multibeam survey of the continental slope NW of the UK was undertaken by MV Franklin in September 2006. This was conducted as part of the UK’s Department for Trade and Industry’s Strategic Environmental Assessment programme and the Department for the Environment, Food and Rural Affairs, through their agency the Joint Nature Conservation Committee, Special Area for Conservation programme. The surveyed areas were ground-truthed using video and stills image cameras mounted on a drop frame. One of the aims of the survey was to locate and investigate the occurrence of structural sponge communities on the UK side of the Faroe-Shetland Channel (FSC).

Structural sponge communities have been recorded from the FSC on both the Faroese and UK sides of the Channel. Their occurrence is thought to be a result of the unique hydrography of this region. On the UK side of the channel the inflowing current, North Atlantic Water, is relatively warm (~7°C) and occupies the upper 500 m. Below the surface water mass there is cooler/fresher water (~0°C) which originates in the Nordic Seas and Arctic Ocean. The water temperature in the Channel is known to fluctuate at depths of 350 m – 650 m as a result of internal tides. The incidence of internal waves is thought to result in resuspension of materials leading to increased food supply, which supports the sponge aggregations.

Almost 1500 km² of seabed was surveyed with 28 video ground-truthing tows and 1800 images obtained. Video data were reviewed and a sub-sample of 260 images quantitatively analysed. Fauna were identified to morphospecies and cluster analysis with group averaged linking was performed on a Bray-Curtis similarity matrix using P/A data to investigate patterns in community structure.

This survey identified structural sponge communities associated with various seafloor features including iceberg plough marks, terrace structures and ditch features. Sponge communities were not visible from multibeam bathymetry or backscatter grided at 15m grid squares. Video observations revealed these communities to be characterised by a high diversity of sponge morphospecies including branched, cup, lamellate, globose, erect and encrusting sponges. Distinct species included bright blue and bright yellow encrusting sponge forms, large white erect sponges with multiple chimney like structures, and Geodid species. The distribution of sponges is patchy with some areas supporting dense growths of large sponges and others areas supporting less dense growths of small and encrusting forms. Cluster analysis of the faunal data suggest that the sponge communities are distinct but related to the surrounding faunal communities. Video and image data suggest these communities are present over a very narrow depth range focussed on the 500 m contour. Data from this and previous studies, suggest these communities may form a continuous narrow band on the UK continental slope north of the Wyville–Thomson Ridge.

Challenges and rewards of ROV-based high-resolution

habitat mapping in deep-sea canyons offshore Portugal

Veerle A.I. Huvenne, Douglas G. Masson, Abigail Pattenden,

Paul A. Tyler, Peter J. Mason

National Oceanography Centre, Southampton, UK

Deep-sea canyons are major geomorphological and ecological features of the continental margins and slopes, but are still poorly known and understood. They can be hundreds of kilometres long, cut hundreds of meters in the overall slope, and can act as conduits, but just as well as traps for the transport of sediments (and hence of carbon and pollutants) from the shelf to the deep sea. They form major breaks in the slope, causing important effects on the local physical oceanography and ecosystem distribution. Yet insights in the processes and patterns governing deep-sea canyons are only increasing slowly, because the canyons form such a challenging environment to study.

The Nazaré, Setúbal/Lisbon and Cascais canyons offshore Portugal have been studied for a number of years within a range of national and international projects. Overall bathymetry and sidescan sonar (TOBI – 30kHz) maps have been acquired over the area (e.g. Lastras et al., subm.), together with video data and a number of cores. Still, due to the high spatial variability and complicated nature of the terrain, it was necessary to refine the mapping to a much higher resolution in order to understand the local distribution of the different habitats and communities. During spring/summer 2007, the ROV Isis was used on board the RV James Cook for this objective.

The habitat mapping operations during JC010 included high-resolution multibeam mapping of 4 areas of ca. 0.5 km² each using the Simrad SM2000 system (200 kHz), video surveying and photographing, the continuous recording of temperature and salinity, and the occasional collection of biological and geological samples. The swath data were processed to 0.5 m resolution grids with the IFREMER software suite Caraibes, while overall data integration was carried out within ArcGIS. The main challenges related to the work in the deep canyons included uncertainty in terms of navigation, problems with sound velocity and beam-forming of the multibeam system, the turbidity of the water column in certain parts of the canyons, and strong currents and tides, also related to the steepness and irregularity of the terrain.

However, despite those difficulties, the results significantly increase our insight in the morphologies and processes present in deep-sea canyons. Important features such as sharp bends, gullies, steep (to vertical) walls and ledges, which usually are smoothed in large-scale bathymetry grids, are mapped accurately, and at a scale which can be linked directly to the biological communities at the seabed. The results illustrate the tremendous heterogeneity of canyon habitats and the associated ecosystem spatial distribution. The 4 bathymetrical data sets acquired show how a flat-bottomed thalweg incises the canyon axis over several tens of meters, even at depths of 4300 m. This creates near-vertical to overhanging walls, often the preferred settling grounds for filter feeders such as stalked crinoids, brachiopods, corals etc. On more gentle slopes, sediment transport is evidenced by sediment waves, below the resolution of even TOBI sidescan sonar imagery (which is 6 x 6 m). Terraces are characterised by thick mud deposits, supporting a typical community of deposit-feeders.

Acknowledgements

The authors whish to thank the captain, crew and scientific party of the JC010 cruise for their support during the data collection. Especially the ROV ISIS-team is gratefully acknowledged for their dedication and effort.

References

Lastras, G., Arzola, R.G., Masson, D.G., Wynn, R.B., Huvenne, V.A.I., Huehnerbach, V. And Canals, M. (subm.). Geomorphology and processes in the Central Portuguese submarine canyons, western Iberian margin. Geochemistry, Geophysics, Geosystems.

High-resolution acoustic mapping of Mediterranean Deep Coral areas

C. Lo Iacono¹, E. Gràcia¹, C. Orejas², J.J. Dañobeitia¹,

A. Gori², J.M. Gili²

1) UTM, Marine Technology Unit - CSIC, Spanish National Research Council, Barcelona, Spain

2) ICM, Marine Science Institute - CSIC, Spanish National Research Council, Barcelona, Spain

Deep Coral areas have been detected and mapped off the Almeria Margin and in the Cap de Creus Canyon, along the western and north-western Mediterranean Sea. The study has been carried out by means of an integrated geophysical dataset, comprising large-scale (TOBI) and small-scale sidescan sonar systems, swath-bathymetry, TOPAS high-resolution seismics. In the area of Cap de Creus Canyon, acoustic dataset has been ground-truthed with visual images from ROVs and the man submersible JAGO (IFM-GEOMAR).

Amplitude analysis of backscatter images revealed extremely high reflective patches, punctually distributed along some of the volcanic banks in the Almeria Margin (Chella and Pollux Banks) for depths ranging from 230 to 470 m. High reflective patches coincide in the seismic records with areas where carbonate cold water coral mounds occurs (Fig.1). Very high acoustic strength could be related to the roughness of coral colonies or to associated sediments.

Banks of the cold coral Madrepora oculata and some isolated colonies of Lophelia pertusa have been visually detected along the southern wall of the Cap de Creus Canyon, by 150-300 m depth range. The obtained images offered valuable information for the characterization of the communities (species composition) and abundance of the coral species as well as the conservation stage of them. In both study areas, the position of deep corals banks suggests that the occurrence of strong bottom currents and reduced sedimentary fluxes are environmental factors suitable for their development

The integration of different marine geophysical methods supported by ground-truthing calibrations, allowed to recognize in detail the morphological and sedimentary constrains suitable for the development of deep coral habitats. Furthermore, quantitative analysis of acoustic images could represent a step ahead in the detection of deep coral areas by remote sensing methods.