Y. Lafoy¹, P. Chavance², D. Buisson³, and A. Rivaton²

1) Office of Regional Cooperation and External Relations, BP M2 – 98849 Nouméa Cedex, Nouvelle-Calédonie

2) Agence de Développement économique de la Nouvelle-Calédonie (ADECAL), BP 2384 Nouméa, Nouvelle-Calédonie

3) Direction des Technologies et des services de l’information (DTSI), 127, rue A. Daly, Ouémo, 98800 Nouméa, Nouvelle-Calédonie

Since 1993 New Caledonia has launched the ZoNéCo programme (for “Zone économique de Nouvelle-Calédonie”) in order to open new avenues for both EEZ governance and economic development.

This multi-disciplinary, on-going programme is divided into three phases:

-        the first phase aims at producing a base map of the ocean floor and includes the analysis of earlier data on the subject;

-        the second phase seeks to identify the mineral and biological resources, and to describe the environment in which they are found;

-        the third phase consists in the assessment of the economic development potential associated with such resources.

The ZoNéCo programme calls upon many scientific disciplines such as, swath mapping and seabed imaging, gravity, magnetism, seismic, physical oceanography, satellite remote sensing, phytoplankton biology, fisheries science, and habitat. From a deep water fisheries point of view, habitat is defined as the physical seafloor condition (e.g. rock, sand, or mud) that allows for sustainability of a targeted species. Distribution and exploitation of resources are mainly linked to: (1) the morphology and depth of the bottom; (2) the nature of the substratum.

Because most of the benthic habitats are defined by their geology, depth, chemistry, and by and other attributes such as, temperature, nutrients, and currents, multidisciplinary techniques are critical in determining habitat structure and lithology.

Between 1993 and 2004, nine surveys have investigated deep sea, mega-scale habitats. Six seafloor mapping, geophysics and physical oceanography cruises were conducted aboard Ifremer’s R/V L’Atalante, followed by three related exploratory fishing surveys in order to survey potential sites of interest for fishery resources. Those nine surveys have swath-mapped and imaged (down to water depths of 2,500 m.) an area of about 500,000 sq. km, i.e. about 35% of New Caledonia’s EEZ.

In addition to the offshore surveys, the ZoNéCo Programme has, since 1999, been focussing on :

-        the understanding of the relationship between Tuna resources and the marine environment variability;

-        the use of remote sensing data for Habitat mapping;

-        the use of an ecosystemic approach on coastal reef fisheries, to understand relationships between the fishing resource and the fishing communities;

-        investigating the Southwestern Lagoon through surveys deploying the Institute for Research and Development / IRD’s Acoustic Ground Discrimination System (AGDS).

Main results and outcomes of the deep-water surveys

Over the last 14 years, the multidisciplinary results of the current ZoNéCo programme have led to an improvement in the knowledge of the marine environment that surrounds New Caledonia.

¤ In terms of deep-sea, mega-scale potential habitats, seafloor-type interpretation of the five swath-mapping surveys has enabled:

•       better understanding of the relationship between Living resources (Marine Biodiversity & Fisheries) and Marine Geology;

•       definition of appropriate fishing strategies, with trawling surveys planned on low reflectivity, flat-topped, shallow areas with smooth slopes;

•       optimization of the fishing grounds exploitation, by using appropriate sampling tools to reduce equipment loss and maximize the gear efficiency.

However, sampling surveys (fishing and dredging cruises) still need to be conducted to ground-truth the newly discovered potential non-living and living resource targets.

¤ In terms of living resource assessment, the three exploratory surveys revealed potentially exploitable species of fish (red snappers, alfonsino (Beryx splendens), black bream (Eumegistus illustris)) and prawns.

In order to formulate a sound resource management policy, specimens of alfonsino were kept for later genetic studies aimed at establishing whether they belong to a single or to several stocks.

Trawl attempts failed to bring back any specimens of orange roughy (Hoplostethus atlanticus), despite the fact that this species is abundant in the New Zealand part of the Norfolk Ridge.

¤ In terms of resource management, the study of samples of commercially exploitable species collected during the programme, particularly alfonsino, has greatly improved the knowledge of the biological parameters (growth rates, reproduction, mortality, etc, ...) required for formulating sound resource management guidelines and policy. Commercial fishing inventories can be thus conducted that could lead to the establishment of a fishing and aquaculture monitoring facility.

¤ Regarding Physical oceanography, measurements obtained, together with satellite observed sea-level altimetry data, make it possible to study the seasonal and interannual variability of climate and rainfall conditions throughout the region, and in New Caledonia in particular.

The unveiled, large (200 km in diameter), deep, counter-clockwise gyre (reaching down to 700 m depth) may play a major role in the dissemination of the larvae and juveniles of species of commercial interest. Moreover, the occurrence of an upwelling along the Southwest coast of New is likely to have an influence on the distribution of marine species of commercial interest.

¤ Finally, regarding the governance of both living and non-living resources of New Caledonia’s EEZ, ZoNéCo’s multidisciplinary data, classified and recorded into specific and topic-oriented data bases, can be cross-cut to generate specific “products” or “decision-making” maps (i.e. integrated maps obtained after compilation of swath bathymetry, acoustic imagery, physical oceanography and geophysical data).

The production of these integrated maps mainly aims at:

-        matching the needs of various professionals (fishermen, geologists, oceanographers, planning authorities),

-        improving the governance of the 1,400,000 sq. km-wide EEZ of New Caledonia, together with its associated non-living and living resources.

Back

A practical morpho-dynamic framework for mapping seafloor environment for the purpose of seabed management in Canadian EEZ

Vladimir E. Kostylev and John Shaw

Natural Resources Canada, Bedford Institute of Oceanography

1 Challenger Drive (P.O. Box 1006), Dartmouth, NS, Canada B2Y 4A2

In recent years some of the largest advances in science have taken place at the intersection between formerly separate disciplines. Habitat mapping – at the intersection between marine ecology, marine geology and physical oceanography – has come to prominence as a necessary tool for ocean management. Habitat mapping recognizes that the physical nature of the sea floor, i.e., surficial geology, is critical to understanding the distribution of marine biological resources that economically sustain coastal communities in Canada. In 2006 Natural Resources Canada commenced a 4-year project to address these questions, with an emphasis on establishing a national perspective of the geo-environment and habitats on Canadian continental shelves. The goal of the project is to describe broadscale patterns in seabed habitat structure and processes in Canadian waters, their impacts on seabed life, and relevance to major issues under Ocean Action Plan. The emphasis is on understanding how geological controls on benthic habitat vary through time, and on the assessment of the relative importance of physical factors at different spatial scales. The unifying idea of the project is to interpret and map emergent, rather than apparent, properties of Canadian seabed habitats based on the integration of knowledge of geologic, oceanographic and ecological patterns and processes on different spatial and temporal scales. The project accommodates a variety of interdisciplinary issues important for Ocean Management, crucial for achieving balance between resource exploitation and preservation of unique seabed habitats.

Back

Quantitative Analysis of Backscatter Data and Automatic Segmentation of Seafloor Physical Properties in Cook Strait, New Zealand

Geoffroy Lamarche¹, Xavier Lurton², Anne-Laure Verdier^1,2, Jean-Marie Augustin², Ian Wright¹, Ashley Rowden¹, Alan Orpin¹ and Miles Dunkin¹

1) National Institute of Water and Atmospheric Research (NIWA) Ltd, Private bag 14-901, Wellington 6021, New Zealand.

2) Institut Français de Recherche pour l’Exploitation de la Mer (IFREMER), BP70, 29280 Plouzane, France

Echo-sounder backscatter intensity is a relative measure of the sound-scattering by the seafloor. Backscatter strength is a function of the seafloor substrate and roughness, i.e. it is associated to both sediment grain size, porosity, and small scale topography. Quantification of the backscattered signal can potentially provide a means to remotely characterize the nature of the seafloor and generate regional scale maps of geological and possibly biological significance.

Our work focuses on the Cook Strait region, central New Zealand, where a wealth of EM300 multibeam bathymetry and backscatter data (~ 30 kHz) are available, augmented by an extensive geological database (seafloor photographs, sediment and rock samples, and high-resolution seismic profiles). This provides an excellent opportunity to ground truth and quantify the backscatter signal. The processing of the backscatter signal, which aim is to remove the effects of the recording equipment, seafloor topography, and the water column, is undertaken using the newly implemented SonarScope software, developed by Ifremer.

The processing includes sonar image mosaicing, signal calibration and compensation, speckle noise filtering, image segmentation and image textural analysis. The backscatter is processed at an enhanced resolution of 5 or 10 m grid depending on water depth. Backscatter profiles have been extracted from the raw data carefully accounting for the angular dependence which is readily available from the co-registered multibeam bathymetry data.

The analysis of the backscatter data resulted in the identification of local geological, sedimentological, topographic, and possibly biological features otherwise not recognised with conventional surveying. Angular backscatter laws have been systematically extracted for characteristic areas, documenting the variety of geological facies in this extremely tectonically and sedimentologically active region. This catalogue will form a generic reference for future investigation at other areas. Examples of detailed analysis of local features include: 1) High reflectivity areas at the top of the continental slope associated with rough micro-topography and carbonate concretions originating from relict cold seeps; 2) Low reflectivity associated with the tops of sand waves and ridges in central Cook Strait, an unprecedented counter-intuitive result; 3) Complex reflectivity patterns associated with active fault scarp which emphasises the potential of backscatter data in submarine seismic hazard studies; 4) Reflectivity in canyons that varies along the continental slope and provides an indication of transported material and activity; and, 5) New statistical compensation of the backscatter data from the Haungaroa volcano allows a proof-of-concept biodiversity mapping exercise. The method utilised ecological theory to predict biodiversity from a knowledge of seabed substrate heterogeneity. The latter could be derived from the segmentation of the backscatter data, now that acquisition artefacts are properly compensated and attenuated. This technique will be of importance for other large-scale mapping initiatives, such as Cook Strait.

Back

“Charting the Yasawa Platform”

Complex morphology and structure with a marine habitat of potentially significant diversity is revealed for the first time in the Fiji Islands

Robert Smith (SOPAC)

Felix Maharaj (Fiji Hydrographic Unit)

The Yasawa Islands platform and its plate tectonic significance has long been the subject of study to unravel its geological evolution. However, until now, much of this shelf platform marking the western boundary of the Fiji platform has never been mapped.

Previous resource surveys were limited to geophysical hydrocarbon surveys undertaken in the early 70s with great difficulty experienced in navigating the shelf waters due to the inadequate coverage of existing charts of the area. Kilometre-long streamers would find uncharted shoals. The occasional reconnaissance surveys by organizations interested in the fisheries potential of the shelf were also not very exhaustive.

Today a large portion of the vessel traffic in the Yasawa and Mamanuca Islands are operated by the tourism industry using modern GPS navigation systems against backdrops of highly inaccurate chart data that has datum incompatibilities and that are certainly not consistent with current mapping standards. The “F – 5”, a chart released by the Fiji Hydrographic Department in 1986, is the only available chart for the Yasawa Islands and clearly indicates how much of the area remains uncharted. Viwa Island, the most western island in the group, lies on the western edge of the platform and outside the western boundary of the chart.

With limited resources to survey an area of 4,000 km² to optimum standards, the Fiji Hydrographic Department sought the assistance of SOPAC for the use of their multibeam mapping system, a RESON 8101. The principal objective of the programme was to produce a new chart for the Yasawa and Mamanuca group of islands. In 2005 four months of survey work was completed; and in 2006 another month. This collaborative effort resulted in the acquisition of 5500 line kilometres of multibeam, sidescan and backscatter data covering an area of approximately 1600 km².

The benefits of this combined mapping exercise are just beginning to unfold along with the almost magical landscape of the Yasawa platform heretofore hidden even from remote imaging satellite sensors. The very complex seafloor morphology with numerous patch reefs, drowned barrier reef systems, a network of structurally controlled valleys and channels, fault scarps, fault-controlled basins dramatic fore reef slopes with 500 m scarps have now yielded their secrets. Offshore submarine canyon development is surprisingly limited. With such a complex morphology, a complex hydrodynamic flow regime is likely to exist across the platform, hence a considerable diversity in the marine biodiversity can be expected. One such habitat explored in this dataset is home to a diverse group of fishes commonly referred to as bottom fish that are found on the fore-reef slopes, pinnacles and seamounts at depths between 100 – 500 m and is highly prized for the quality of its flesh.

Technically the survey was very challenging with excellent lessons learnt about what can be achieved on a shoestring budget – but that is a different story on its own. Apart from producing a new chart for the Yasawa waters, the same dataset can further exploration into the natural resource potential of the area in fisheries, habitats, coral research and exploring climate change and sea-level rise impacts and implications.

Back

Seabed Biodiversity of the Continental Shelf of the Great Barrier Reef Region

Roland Pitcher¹, Peter Doherty², John Hooper³, Peter Arnold³ & Neil Gribble⁴

1) CSIRO Marine Research, Cleveland, Qld. 4163

2) AIMS/CRC-Reef, Townsville, Qld. 4810

3) Queensland Museum: South Brisbane & MTQ Townsville

4) QDPI Northern Fisheries Centre, Cairns, Qld. 4870

Until recently, little was known about the distribution, abundance and diversity of habitats and biota of the deeper shelf seabed between the coral reefs of the Great Barrier Reef (GBR) Marine Park. From 2003 to 2006, the GBR Seabed Biodiversity Project has been mapping these habitats and sampling their biodiversity along the length and breadth of the region.

The Project is now producing comprehensive inventories & maps, developing risk indicators with respect to fisheries sustainability, and assessing the status of biological assemblages. This information is assisting managers to conserve important habitats and rare biodiversity, and to ensure that fisheries within the Park are ecologically sustainable.

The scale of the project is large (>200,000 km²; >1385 sampling sites) and can be achieved only by applying multi-disciplinary skills to acquire the many different data types and describe the diversity of habitats and biota. This was achieved by multiple investigators from several collaborating agencies, with funding support from CRC Reef, FRDC and the National Oceans Office.

Methods include analysis of bio-physical relationships between the physical environment and species and assemblages (eg. large scale datasets such as satellite remote sensing, oceanographic model output, sediments, bathymetry and acoustics), as well as human disturbance, as a basis for biodiversity characterization, prediction and mapping.

Progress and results from fieldwork, biological identifications and analyses will be presented. In particular we will present results on the degree to which biophysical variables can be used to predict the distribution of biodiversity and habitats.

Back

SESSION 2

Benthic habitat mapping in coastal waters of south-east Australia

D. Lerodiaconou¹, A. Rattray¹, L. Laurenson¹, S. Burq² and M. Reston²

1) School of Life and Environmental Sciences, Deakin University, P.O. Box 423, Warrnambool, Victoria, 3280

2) Victorian Partnership for Advanced Computing, 110 Victoria St, Carlton South, Victoria, 3053

The Victorian Marine Mapping Project will improve knowledge on the location, spatial distribution, condition and extent of marine habitats and associated biodiversity in Victorian State waters. This information will guide informed decision making, enable priority setting, and assist in targeted natural resource management planning. This project entails benthic habitat mapping over 500 square kilometers of Victorian State waters using multibeam sonar, towed video and image classification techniques. Information collected includes seafloor topography, seafloor softness and hardness (reflectivity), and information on geology and benthic flora and fauna assemblages collectively comprising habitat. Computerized semi-automated classification techniques are also being developed to provide a cost effective approach to rapid mapping and assessment of coastal habitats.

Habitat mapping is important for understanding and communicating the distribution of natural values within the marine environment. The coastal fringe of Victoria encompasses a rich and diverse ecosystem representative of coastal waters of South-east Australia. To date, extensive knowledge of these systems is limited due to the lack of available data. Knowledge of the distribution and extent of habitat is required to target management activities most effectively, and provide the basis to monitor and report on their status in the future.

Back

Application of swath acoustics to map seabed habitats on the continental shelf of New South Wales, Australia

A. Jordan, P. Davies, T. Ingleton, and T. Pritchard

New South Wales Department of Environment and Conservation, PO Box A2930, Sydney South NSW, Australia, 1232

The mapping and spatial analysis of subtidal seabed habitats and benthic assemblages in the marine environment is being increasingly used to further advance spatial management and monitoring of the seabed. Developments in swath acoustic technology has allowed a considerable increase in the area of the seafloor to be mapped at increasing resolution. An interferometric sidescan sonar that collects geo-referenced depth and sidescan backscatter data has been recently employed in coastal waters off New South Wales, Australia to generate high resolution bathymetric and backscatter mosaics of the seafloor. This key tool is combined with more traditional aerial photography, underwater video surveys and Geographical Information System (GIS) analysis to create a number of spatial products used to derive digital elevation models, seabed habitat maps and data on macro-benthic floral and faunal assemblages.

The detailed bathymetry has revealed large variations in structural complexity of reef habitats, with the morphology of the seafloor examined using parameters of slope and rugosity. The spatial distribution of seabed habitats are digitised as a vector layer using a combination of the bathymetry and backscatter and hill-shading techniques. The backscatter has also revealed significant structuring of unconsolidated habitats (primarily sand) at the scale of 100’s of metres, influenced primarily by the presence of sand waves and variations in particle size and shell content. Further ground-truthing using video and sediment grabs are allowing these habitats to be mapped at a lower hierarchical level improving our understanding of the spatial distribution of soft-sediment assemblages on the continental shelf. Continued video analysis of rocky reef habitat is also providing information on the extent and zonation of sessile assemblages over large geographic scales.

Around 400 sq/km of the NSW continental shelf has been mapped during the first two years of the program, primarily targeted at nearshore and offshore rocky reef habitats within NSW Marine Parks in the Tweed-Moreton, Manning Shelf and Batemans Shelf bioregions. Mapping of seabed habitats is an essential component of Marine Park planning and is particularly important for the process of developing zoning options to ensure representative habitats are included with the highly protected Sanctuary Zones. Zoning plans have recently been prepared for two new marine parks in NSW covering almost 2000 sq/km, and the use of the habitat maps in the zoning process will be discussed.

Protocols are being developed to integrate and display spatial data from the range of remote sensing techniques, including map series at variable scales, spatially referenced video and detailed metadata. The current mapping methods employed in the project will be presented and future research areas, such as the use of digital still photography to improve taxonomic resolution and quantification of morphological structure of sessile fauna and flora will be outlined.

Back

Spatial Analysis of Multiscale Seascapes on Oceanic Island Flanks

Jens C. Krüger

Pacific Islands Applied Geoscience Commission, SOPAC

P.M.B., Suva, Fiji

The most common parameter that is used for benthic habitat mapping is depth. The volume of bathymetric data available in the Pacific region has increased significantly in recent times, with contributions from space, air, and water-based systems, and the dissemination of data through web-based portals. The quality as well as temporal and spatial resolution (both vertical and horizontal) of these data varies greatly as a function of sensor capabilities, initial survey requirements and post-processing efforts. Outputs remain compatible however, as data are commonly distributed as points with depth attributes. Identifying sources and merging multiple datasets for a particular area is becoming increasingly viable and beneficial, as there is a general paucity of bathymetric data within Pacific Island Countries. While creating a continuous raster surface of seafloor topography is a valuable exercise, it requires aggregating, interpolating and edgematching of the original datasets. This presents challenges for subsequent benthic mapping as features are often degraded and quantitative methods of surface analysis rely on scale-dependent morphometric variables such as slope angle.

This study explores spatial analysis techniques using recently acquired multibeam echosounder data from a variety of Pacific Ocean island settings. The resolution of a multibeam echosounder system generally decreases with water depth and distance from the nadir. This has significant implications when mapping the steep slopes (average of 20-25º) of oceanic islands, as resultant grid spacing ranges from 5 m in the nearshore to more than 100 m over abyssal plains. As the resolution of the seafloor data decreases, so does the complexity of the methods that can be used to map and classify potential benthic habitats. The resultant morphological interpretation is generally limited to scale-independent meso- and megascale features such as crests, depressions, and thalweg areas.

Back

Discovery, characterization, and mapping of chemosynthetic habitat along the active East Coast margin of New Zealand

Alan Orpin¹, Ashley Rowden¹, David Bowden¹, Jens Greinert², Peter Hill¹, Cliff Law¹, Geoffroy Lamarche¹, Scott Nodder¹, Arne Pallentin¹, and Ian Wright¹

1) National Institute of Water and Atmospheric Research (NIWA), Private Bag 14-901, Kilbirnie, Wellington, New Zealand

2) Marie Curie Fellow, GNS Science, PO Box 31-368, Avalon, New Zealand

New Zealand and American institutes studying the water chemistry and biodiversity of chemosynthetic habitats discovered numerous new cold seep sites along the East Coast margin of New Zealand during research cruises in October and November 2006 (“Cook Strait Methane”, RENEWZ I – NEW ZEEPS, funded by NOAA and NIWA). Here, oblique convergent tectonics has led to the development of a thick accretionary wedge of deformed tertiary mudstones. Forty years of historical geophysical and fisheries surveys built a library of “unusual phenomena”. This was not fully realized until the nineties as potential evidence of cold seeps. A land-mark study by Lewis and Marshall (1996) compiled the occurrences of seep-related fossils and carbonate concretions collected from dredges and fouled in fishing nets, together with geophysical evidence of suspect seep sites and flares. One of these sites off the Wairarapa has provided NIWA researchers with time-series data that showed persistent methane-enrichment above a flare since September 2005. In isolation these discoveries were somewhat serendipitous in nature, but collectively established a compelling case for cold seeps environments along this region of New Zealand.

Localities published by Lewis & Marshall (1996) formed the framework for the most recent research cruise aimed specifically at precisely locating seep sites and undertaking in situ visual investigations and sampling. Once on location a systematic series of surveys were undertaken to provide increasingly tighter control on the precise occurrence, dimension and geomorphology of the site. In the absence of multibeam bathymetry, regional surveys using an EM300 were initiated to provide adequate coverage (hundreds of kilometres square). In partnership with multibeam binned at 25 m, backscatter imagery was used to identify unusual seafloor (typically a strong intensity return) and authigenic carbonate hard-grounds and chemosynthetic shell hash, characteristic of cold seep sites globally. In some cases, sites were smaller than the resolution afforded by standard backscatter analysis. The broadscale base map data was refined by two echo sounders operated using a series of 1-2 km long figure-of-eight passes over any suspect sites: (1) an ES60 38 kHz echo sounder (primarily for fisheries acoustics) to image the water column above any seep flares and pin-point the seep site; and, (2) a 3.5 kHz sounder to characterize seafloor and sub-seafloor strata, which in most cases clearly identified aureoles of carbonate cementation and hummocky microtopography tens of centimetres to 2 m in scale. Experience showed that the water column soundings were very good at pin-pointing discrete targets while the 3.5 kHz sounder provided excellent spatial control of the seep extent. Once the centre of the seep had been ascertained, a deployment sequence was initiated for instruments lowered off the ship. An HPR 21-32.5 kHz acoustic navigation system was attached to all deployments and the position of the gear was calculated and plotted in real-time over the multibeam data using specifically designed geospatially software (OFOP© v. 3.0.2a, Greinert 2006). These data were also broadcast to the bridge to ensure both scientific and ship’s crews were working together. A series of orthogonal tows followed using a new Deep- Towed Imaging System (DTIS) that recorded high-definition digital video and digital stills. Real-time video footage was relayed to the ship and synchronous sea floor observations were annotated on-the-fly using geospatial software. These data also formed the framework for subsequent quantitative analysis back in the laboratory, detailing faunal assemblage relationships with geomorphology. After 4-8 tows a library of sea floor observations were compiled and reviewed before precision coring, CTD water sampling, grabs or epibenthic sleds were deployed.

The key ingredient to the success of this approach was the continual refinement of observational data within a geospatial environment. The end result was a seafloor, water chemistry, and faunal sampling routine that could isolate targets to within 30 m at 1000 m water depth in the absence of a ROV or dynamic positioning system.

Lewis, K.B, Marshall, B.A. 1996. Seep faunas and other indicators of methane-rich dewatering on New Zealand convergent margins. New Zealand Journal of Geology and Geophysics 39, 181-200.

Greinert, J. (2006). Ocean Floor Observation Protocol (OFOP©) Version 3.0.2a. JGProduction, Days Bay, New Zealand.

Back

Seafloor habitat characterization on the Aleutian Ridge, Alaska

Jennifer R. Reynolds (University of Alaska Fairbanks)

and H. Gary Greene (Center for Habitat Studies, Moss Landing Marine Laboratories)

E-mail: jreynolds@guru.uaf.edu

The Aleutian Ridge extends 1600 km westward from the Alaska Peninsula, and forms the physiographic boundary between the North Pacific Ocean and the Bering Sea. Trawl surveys and fishing bycatch reports indicate that the central and western parts of the Aleutian Ridge may have the highest abundance and species diversity of deep corals in the world. The intersection of deep coral abundance with a high level of fishing activity drives a need for information on the distribution and characteristics of the deep coral, for effective ecosystem management.

The research reported here is part of an multidisciplinary study to identify habitat associations of deep coral and sponges, and to determine the distribution of these habitats. The ultimate goal is to construct a predictive model for the distribution of deep coral and sponge species on a regional scale. This predictive approach is practical because the habitat characteristics of the Aleutian Ridge seafloor vary systematically with the geology and oceanography. Representative areas contain specific assemblages of habitats in relatively predictable patterns. The approach is to characterize representative areas, using new high-resolution sonar mapping and visual groundtruth/sampling, and then extrapolate the regional distribution of those types of areas using existing hydrographic charts and regional geology. While this approach cannot pinpoint the locations of local habitats in areas that have not been mapped at high resolution, it can predict the occurrence and abundance of those types of habitats (habitat assemblages) in new areas. Combining this predicted habitat map with habitat associations of deep coral and sponge species will produce a predicted map of the deep coral and sponge distribution.

The study focuses on a 500km section of the Central Aleutian Ridge, from 50 m to 3000 m water depth. Based on variations in the regional geology, oceanography, coral bycatch data, and fishing pressure, seventeen sites were chosen as representative of seafloor in the region. Bathymetry and backscatter surveys were conducted with Reson SeaBat 8111 (100 kHz) and Reson SeaBat 8150 (24 kHz) multibeam echosounders. Visual observations and sampling were accomplished with the manned submersible Delta and the deep-diving ROV Jason II. The biological aspects of the study are underway by our colleagues at NOAA’s Alaska Fisheries Science Center (Jon Heifetz, Robert Stone) and the Alaska Department of Fish & Game (Doug Woodby). This presentation focuses on the interpretation and habitat classification of the multibeam data, in the context of the regional geology.

Aleutian Ridge forms the forearc and volcanic arc of the Aleutian subduction zone. The crest of the ridge is an eroded, submerged platform, with a chain of islands and active volcanoes along the north edge. The north and south flanks of the ridge face different water masses and current regimes, have different rock substrates, and have distinct patterns of sedimentation and erosion. Along the north flank, the volcanoes act as major sources of detritus, and seafloor characteristics vary with proximity to the active and inactive volcanoes. Submarine volcanic cones also form distinctive habitats. South of the islands, the summit platform has extensive outcrops of volcanic and sedimentary bedrock. Toward the south flank, sediment cover increases with depth and distance from the islands, producing a predictable sequence of outcrops, sedimented shelf with sediment waves, sediment deposition on the upper slope, and a zone of mass wasting located at mid-slope. The Aleutian ridge is also tectonically active, and the south flank within the study region is cut by extensional faults and a deep rift canyon that exposes fresh bedrock, providing substrate for corals. The regional distribution of these general seafloor characteristics is combined with the high-resolution surveys of representative areas, to predict the abundance and distribution of local-scale deep coral habitats across the region.

Back

Benthic habitat variation over tidal ridges

Thaiënne A.G.P. van Dijk¹, Jan A. van Dalfsen², Pieter J. Doornenbal¹, Isabelle Du Four³, Vera R.M. van Lancker³, Sytze van Heteren¹ and Ronnie A. van Overmeeren¹

1) Geological Survey of the Netherlands, TNO Built Environment and Geosciences

2) Institute for Marine Resources & Ecosystem Studies (IMARES)

3) Renard Centre of Marine Geology (RCMG), Ghent University

E-mail: thaienne.vandijk@tno.nl

Marine habitat mapping reveals that benthic habitats vary on continental shelves in relation to the distance to the coast and to offshore morphology. Tidal ridges are expected to accommodate different benthic habitats, which are important to both benthic and pelagic organisms. Some of these ridges are nominated to become marine protected areas, but are also attractive for their marine aggregates and may be designated in part as mining areas. Improving our insight in the benthic habitat variation over tidal ridges and our understanding of the relationship between ecology and abiotic factors, are therefore important not only from a scientific point of view, but also for assessing ecological values and their preservation in the management and use of shallow continental seas. Here, we aim to present the habitat variability within two sites with tidal ridges in the southern North Sea.

Acoustic facies mapping is a spatially continuous method and was carried out at the Brown Bank and Thornton Bank, two tidal ridges of 28 and 12 m height difference respectively. The diagnostic value of facies was validated with seabed samples, obtained using a cylindrical box corer. The selection of sampling locations was based on morphological data and on the acoustic facies of just collected multibeam and sidescan sonar data.

Preliminary results show that the contrasting acoustic facies on sonograms represent differences in geomorphology, sediment grain size and/or macrobenthos (Figure). Expected biological facies are not always corroborated by information from ground truthing. A preliminary classification reveals a general zonation over both tidal ridges in which the sandy tops accommodate poor benthic communities and the adjacent slopes and swales are characterized by rich communities of higher density and diversity.

The tops of the tidal ridges display fine to medium well-sorted and unbioturbated sand and low benthos densities (26-156 ind/m²) with very low diversities (2-6 species per sample). Cluster analysis and multi-dimensional scaling indicate that 4 out of 5 crest samples presently analyzed form a separate cluster, although similarity among these samples is not large. Very coarse sea bed sediments (up to cobbles) occur in the swales to the east of the ridges and bioturbated sandy sediments and sand on clay in the swales to the west of the ridges. All swales have rich communities with densities of >1000 ind/m² up to 4000 ind/m² and diversities of 16-28 species per sample. The dominant species within these communities differ among samples. On the slopes, the average benthos density is 614 ind/m² and diversity is also intermediate.

Both the cluster analysis and the multi-dimensional scaling plot display clear differences between the two areas, for example the near-absence of mollusks at the Thornton Bank, although some annelid worms and crustacean species are common in both areas. The small similarity between samples and the dissimilarity between the two sites may be explained by the large sediment variability and distance between samples both within and between the sites.

As a next step, the effect of the Quaternary geology on facies and benthos communities will be investigated. Also, the analysis of the remaining samples will allow us to test the relationship between sand wave mobility and benthos (e.g. stabilizers/destabilizers).

This study contributes to the understanding of spatial habitat variability and to the improvement of marine habitat mapping techniques, and may be used in the hypothesis-testing of controls on habitat changes in time. Outcomes of this research will help policy makers to manage continental seas in a sustainable manner.

Back

Comparative water residence times of Tarawa lagoon post and pre causeway construction: a hydrodynamic modeling approach

Hervé Damlamian

Ocean and Island program, SOPAC, Private Mail Bag, Suva, Fiji Island

E-mail: herve@sopac.org

During the second half of the 20th century, causeways linking the separate islets of South Tarawa were constructed. Early photographic evidence suggest this was occurring in an ad hock manner from before 1940 (local dry wall construction), through to the early 1980’s when the Bariki / Betio causeway was completed. The impact of the closure of these ocean passages over lagoon water residence times and therefore water quality is unknown. However, the issue of declining water quality within the lagoon is a frequently cited by the Government of Kiribati and broader Tarawa community.

A two dimensional hydrodynamic baseline model was created using MIKE 21 software and once calibrated, this tool was used to calculate spatially-dependant water residence times within the lagoon. The water residence time was defined as the time taken by a water or tracer parcel to leave the domain of interest. Water residence times in turn are a useful tool when considered in terms of inputs, water quality and ecological maintenance of the lagoon ecosystem.

Due to the shallow enclosed nature of the lagoon, water circulation was predominantly regulated by tidal fluctuations over the open western barrier reef and single deeper channel. However, the contribution of flow through the remaining open inter-tidal channels of North Tarawa was also successfully simulated by locating sources in each channel. These flow rates were in turn calibrated using in situ channel flow data collected over a tidal cycle.

Using the calibrated model the closed channels of South Tarawa were removed and the model was run again and allowed to simulate flow through the former intertidal channels. It is recognized that this simulation scenario cannot be calibrated or compared to real data (since no channels remain open) and it is likely that due to the different orientation of South Tarawa, wave setup and tidal flow would be different to that measured in North Tarawa. However, we are confident that such simulations present one of the best opportunities to understand the role that causeway construction has in changing water residence time in Tarawa lagoon.

Back

Evaluation of a wide swath bathymetry system, CUBE modelling, and 3D data visualization for habitat mapping

Peter Davies¹, Douglas Bergersen², Tim Ingleton¹, Alan Jordan¹, Tim Pritchard¹

1) Department of Environment and Conservation, New South Wales, Australia

2) Acoustic Imaging Pty Ltd, 111 Heath Road, Pretty Beach, NSW 2257, Australia

E-mail: dbergersen@tritonimaginginc.com

Geoswath is a shallow water swath bathymetry system designed to provide wide coverage in shallow water depths. Fledermaus is a suite of software designed for the processing and modelling of bathymetry data, and the 3D visualisation of any point or bitmap data. The combined capabilities of these two tools provide an accurate and efficient means of developing digital elevation models of the seabed and providing visualisations used to assist the process of mapping seabed habitats.

The Department of Environment and Conservation, New South Wales has been operating a Geoswath 125 kHz system for 2 years to collect bathymetry and backscatter information for habitat classification in coastal waters. The information is used to assist management of coastal ecological resources and selection of marine protected areas. The project has mapped some 400 km² to date with a further 250 km² planned for the next year.

This paper examines in detail a subset of the NSW dataset, which has been mapped with the Geoswath and groundtruthed with sediment sampling, and underwater towed video. Batch filtering and correction of the raw soundings was accomplished with the Geoacoustics software GS+ and Geotexture. The data XYZ data were then imported to the Fledermaus suite for CUBE modelling and final data validation. This methodology allows large volumes of data to be processed very efficiently. The backscatter data were processed with the Geotexture software into a mosaic and draped over the bathymetry in Fledermaus.

Analysis of the acoustic backscatter and 3D bathymetric model enables reliable seafloor classification to at least three habitat classes from acoustic data alone but further division from seabed morphology is possible. Towed video data is needed to discriminate certain seafloor habitats and to provide data on benthic assemblages. Incorporating all data into the 3D visualization environment of Fledermaus allows subtle relationships between different data types to be easily analysed and understood. The Fledermaus software suite has the additional advantage of being an effective marketing medium for presenting results of habitat studies, either in the form of movies generated from the software or as Scene files of complex data assemblages.

The error budgets of the system and the error analysis from ground truthing will be presented. Improvements to data quality and presentation by through use of the Fledermaus product will be assessed. In our opinion the Geoswath / Fledermaus system provides a suitable option as a fairly low-cost system for shallow water habitat mapping. Advantages and limitations of the combined system plus software will be discussed.

Back

Repeating high-resolution sidescan and multibeam surveys – possible implications for long-term habitat monitoring

Veit Hühnerbach¹, Philippe Blondel², Veerle A.I. Huvenne¹, Olga Gómez Sichi²

1) Geology & Geophysics, National Oceanography Centre, Southampton, European Way, Southampton, SO14 3ZH, UK,

2) Department of Physics, University of Bath, BA2 7AY, UK

E-mail: vhh@noc.soton.ac.uk

During RV Pelagia cruise 250 in the Minch (NW Scotland) in summer 2006, a total of four survey lines of high-resolution deep-towed sidescan sonar (325 kHz) and hullmounted multibeam (30 kHz) were concurrently run over cold-water coral mounds, bedrock and smooth sediment. Two of the tracks each were chosen to run over the same area of seafloor in two different directions (WNW-ESE and NNW-SSE) to see the impact of different ensonification direction on the acoustic backscatter. Extensive ground-truthing complemented the identification of the different facies/potential habitats.

The two acoustic data sets of sidescan and multibeam backscatter were processed to similar resolution, using radiometric and geometric corrections, with the in-house NOCS software suite PRISM. Subsets from the backscatter imagery were then analysed with the University of Bath software TexAn using Grey Level Co-occurrence Matrices (GLCMs) to calculate entropy and homogeneity indices in moving windows across the imagery. Entropy quantifies the amount of local chaos or organisation within an image, whereas homogeneity describes the amount of similarities/ dissimilarities in a chosen neighbourhood around each pixel.

The objective of the study is not only to distinguish between different seabed facies (coral, bedrock, background sediment etc.), using texture analysis, but also to assess the quality of repeated hydro-acoustic surveys over the same terrain in almost identical oceanographic conditions. Our aim is to find out if potential changes in the acoustic data correspond to real changes of the seabed environment, or if they are due to acoustic noise of the equipment and the oceanographic conditions surrounding the habitats.

Back

Seabed classification using acoustic backscatter data

and artificial neural networks

Ivor Marsh and Colin Brown

Department of Earth and Ocean Science

National University of Ireland, Galway, Ireland

E-mail: colin.brown@nuigalway.ie

The MESH (Mapping European Seabed Habitats) programme is an international marine habitat mapping programme whose aim is to produce seabed habitat maps for the seas around north-west Europe. As part of the development of predictive tools for habitat mapping, we have focused on Simrad EM1002 95kHz multibeam backscatter data acquired by RV Celtic Voyager in several shallow-water (<100 m) locations offshore Ireland. In order to prepare backscatter mosaics for seabed classification, we have worked with individual beam data logged in Simrad format datagram. Artifacts associated with data acquisition, including nadir striping, were corrected and beam footprint backscatter data were mosaiced and classified using artificial neural networks.

Unsupervised classifications to delineate acoustically similar seabed types were carried out employing variants of neural network models (Kohonen competitive, self-organising feature map and learning vector quantisation). On a validation data set, the Kohonen competitive network and the self-organising feature map networks out-performed the most commonly utilised statistical classifier, the migrating means (ISODATA) algorithm. Classic Haralick textural features were also computed from the original backscatter mosaic using grey level co-occurrence matrices. Our results indicate that neural network seabed classifications based on these features can be unnecessarily complicated and, on the basis of Occam’s Principle, may be inappropriate in the absence of ground truthing.

These results suggest that Haralick features may be more useful in supervised classifications. These were carried out using a back propagation neural network and a maximum likelihood classification algorithm using ground truthing information collected from 22 sampling stations covering an area of ~140 km². The resultant classifications using the two methods are comparable but it would be possible to improve the accuracy of the predictions if a systematic ground truth sampling programme were undertaken.

The paper concludes with a discussion on the limitations of working with mosaics of backscatter data and outlines an inversion scheme for backscatter ping data to improve seabed classification.

Back

The nature of Tuvalu’s vulnerability to flooding and inundation

revealed by historical reconstruction for 108 years

Hiroya Yamano^1,2, Hajime Kayanne³, Toru Yamaguchi⁴, Yuji Kuwahara⁵, Hiromune Yokoki⁶, Hiroto Shimazaki¹, and Masashi Chikamori^7,8

1) Center for Global Environmental Research, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, Ibaraki 305-8506, Japan

2) UR 128 CoRéUs, Institut de Recherche pour le Développement, BP A5, 98848 Nouméa cedex, New Caledonia

3) Department of Earth and Planetary Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113-0033, Japan

4) Department of Ethnology and Archaeology, Keio University, 2-15-45 Mita, Minato, Tokyo 108-8345, Japan

5) Department of Civil and Urban Engineering, Ibaraki University, 4-12-1 Nakanarusawa, Hitachi, Ibaraki 316-8511, Japan

6) Center for Water Environment Studies, Ibaraki University, 4-12-1 Nakanarusawa, Hitachi, Iba