The Carnarvon shelf at Point Cloates, Western Australia, is characterised by a series of prominent ridges and hundreds of mounds that provide hardground habitat for coral and sponge gardens. The largest ridge is 20 m high, extends 15 km alongshore in 60 m water depth and is interpreted as a drowned fringing reef. To landward, smaller ridges up to 1.5 km long and 16 m high are aligned to the north-northeast and are interpreted as relict aeolian dunes. Mounds are less than 5 m high and may also have a sub-aerial origin. In contrast, the surrounding seafloor is sandy with relatively low densities of epibenthic organisms. The dune ridges are estimated to be Late Pleistocene in age and their preservation is attributed to cementation of calcareous sands to form aeolianite, prior to the postglacial marine transgression. On the outer shelf, sponges grow on isolated low profile ridges at ~85 m and 105 m depth and are also interpreted as partially preserved relict shorelines.

Explaining spatial variation and habitat complexity of benthic habitats from underwater video through the use of maps. Different methodologies currently used to process and analyse percent cover of benthic organisms from underwater video will be addressed and reviewed.

The northern Australian continental shelf is the focus for an expanding offshore energy industry and is also recognised for its high-value marine biodiversity in regional marine management plans. To reduce uncertainty and risk in the future development and management of the region, Geoscience Australia has an ongoing program to provide integrated marine environmental information to support both activities. The program includes collation of existing marine data and acquisition of new high resolution datasets. In 2009 and 2010, marine surveys in eastern Joseph Bonaparte Gulf were completed to characterise the seabed in representative areas, assess potential for geohazards and identify unique or sensitive benthic habitats. Data acquired included multibeam sonar bathymetry (~1900 km2), shallow (<120 m) sub-bottom profiles, sediment grabs and shallow (2-5 m) cores, towed video and epibenthic sleds. Geomorphic features mapped range from expansive soft-sediment plains, to isolated carbonate banks that rise tens of metres and incised valleys up to 200 m deep. Each feature is characterised by a distinctive biota, ranging from coral and sponge gardens on banks to diverse infaunal communities across plains. Geohazards include potential for localised slumping in valleys and escape of subsurface fluid/gas from plains and valley floors. To facilitate uptake of this information, results are integrated as generalised graphical models representing key spatial patterns of shelf ecosystems. This work has led to further work in targeted areas of the Gulf as part of a new four-year Australian Government program to inform geological and environmental assessments of offshore basins for CO2 storage.

This study tested the performance of 16 species models in predicting the distribution of sponges on the Australian continental shelf using a common set of environmental variables. The models included traditional regression and more recently developed machine learning models. The results demonstrate that the spatial distributions of sponge as a species group can be successfully predicted. A new method of deriving pseudo-absence data (weighted pseudo-absence) was compared with random pseudo-absence data - the new data were able to improve modelling performance for all the models both in terms of statistics (~10%) and in the predicted spatial distributions. Overall, machine learning models achieved the best prediction performance. The direct variable of bottom water temperature and the resource variables that describe bottom water nutrient status were found to be useful surrogates for sponge distribution at the broad regional scale. This study demonstrates that predictive modelling techniques can enhance our understanding of processes that influence spatial patterns of benthic marine biodiversity. Ecological Informatics

Anthropogenic global ocean warming is predicted to cause bleaching of many near-sea-surface (NSS) coral reefs and could make deep-water, mesophotic coral ecosystems (MCEs) into coral reef 'life boats', for many coral species. The question arises: how common are MCE's in comparison to NSS reefs? We used a dataset from the Great Barrier Reef (GBR) to show that only about 37% of available bank surface area is colonised by NSS coral reefs (16,110 km2); the other 63% of submerged bank area (25,599 km2) represents potential MCE habitat and it is spatially distributed along the GBR continental shelf in direct proportion to NSS coral reefs. Out of 25,599 km2 of submerged bank area, predictive habitat modelling indicates that about 52% (13,000 km2) is MCE habitat.

Reliable marine benthic habitat maps at regional and national scales are needed to enable the move towards the sustainable management of marine environmental resources. The most effective means of developing broad-scale benthic habitat maps is to use commonly available marine physical data due to the paucity of adequate biological data and the prohibitive cost of directly sampling benthic biota over large areas. A new robust method of mapping marine benthic habitats at this scale was developed based on a stratified approach to habitat classification. This approach explicitly uses knowledge of marine benthic ecology to determine an appropriate number of stratification levels, to choose the most suitable environmental variables for each level, and to select ecologically significant boundary conditions (i.e. threshold values) for each variable. Three stratification levels, with nine environmental variables, were created using a spatial segmentation approach. Each level represents major environmental processes and characteristics of the Australian marine benthic environment. The finest scale of benthic habitat is represented by seafloor physical properties of topography, sediment grain size and seabed shear stress. Water-column nutrient parameters and bottom water temperature depicted the intermediate scale, while the broadest scale was defined by seabed insolation parameters derived from depth data. The classifications of the three stratified levels were implemented using an object-based fuzzy classification technique that recognises that habitats are largely homogenous spatial regions, and transitions between habitats are often gradual. Classification reliability was indicated in confidence maps. Physical habitat diversity was evaluated for the final benthic habitat map that combines the three classifications. The final benthic habitat map identifies the structurally complex continental shelf break as an area of relatively high habitat diversity. Continental Shelf Research

Inter-reefal (i.e. non-reefal) seabed environments have been much less studied than the coral reefs, however they comprise 95% of the total Great Barrier Reef (GBR) Marine Park area. Regional scale spatial analysis of the sediments and geomorphology in these areas allows for a systematic characterisation of the seabed, where comprehensive biological datasets are lacking. We offer an up-to-date synthesis of inter-reefal environments in the GBR, to better understand the nature and distribution of seabed habitats at a regional scale and within the current planning zone scheme, in support of Marine Park management. New quantitative information about surface sediments and geomorphic features, together comprise a new physical dataset of the GBR seabed. This regional dataset contains over 3,000 sediment samples available in Geoscience Australia's (GA) national marine samples database, MARS (www.ga.gov.au/oracle/mars), substantially improving the coverage of surface sediment data from inter-reefal areas, and; GA's current Geomorphic Features dataset (Harris et al., 2005) of the seabed morphology. This marks the first regional synthesis of the surface sedimentology and geomorphology of the GBR since the pioneering work of Belperio (1983a, 1983b) and Maxwell (1968; 1969a; 1969b; 1973). We present a new quantitative sediment dataset that shows regional trends in surface sediments; refining the existing facies model for the mixed carbonate-siliciclastic GBR margin. Our findings also reveal local scale facies characteristics, within the broader regional trends. Until now these distribution patterns haven't been identified on the GBR shelf and are considered to be an important characteristic of the region. In addition, we have revealed other sedimentary characteristics of the region; - Low gravel concentrations cover extensive parts of the shelf. Patches of high gravel concentration occur locally on parts of the inner and outer shelves, reflecting the input of gravel from reef talus aprons. These areas may also be associated with strong tidal currents. - Sand is the dominant grain size fraction, and highest concentrations occur on the middle and outer shelves. Although continuous regions of high sand concentration occur in the far north (e.g. Cape York) and south (e.g. south of Broad Sound) of the Marine Park, the overall distribution of sand is variable as changes in concentration produce local, small-dimension patches at a scale of 10's of metres. - The patchy distribution of sand may reflect a mixture of; 1) widespread supply of modern skeletal carbonate grains, such as foraminifera, molluscs and Halimeda, and/or restricted supply of relict sand; and, 2) the effects of hydrodynamic irregularities in inter-reef channels. - High mud concentrations predominantly occur along the inner shelf and slope. Mud forms local patches on the inner shelf associated with fluvial point sources, which are spatially discontinuous, producing a regionally variable terrigenous sediment wedge of coalescing mud (and sand) patches. - Surface sediments are carbonate-dominated across the shelf and broadly display a regional north-south, shelf-parallel zonation pattern. Low carbonate concentrations of <40% on the inner shelf denote high terrigenous compositions, which increase to >80% on the outer shelf. Within the regional zonation pattern, carbonate patches locally produce a variable distribution in sediment composition. - Uniformly high concentrations of bulk carbonate and carbonate mud on the outer shelf, reflect the constant supply of skeletal carbonate grains from inter-reefal environments, in areas of high reef density and the negligible influence of fluvial sediments on the outer shelf. Regional variations in seabed sediments and geomorphology across the region are also evident in the physical character of the planning zones.

The central aim of the Comparative Geomorphology of Estuaries Project of the Coastal CRC was to improve our understanding of Australia's near-pristine estuaries. As the title implies, the project had a geomorphic focus in that a major output was mapping of geomorphic habitats of a representative selection of nearpristine estuaries from around Australia.

Geoscience Australia (GA) conducted a marine survey (GA0345/GA0346/TAN1411) of the north-eastern Browse Basin (Caswell Sub-basin) between 9 October and 9 November 2014 to acquire seabed and shallow geological information to support an assessment of the CO2 storage potential of the basin. The survey, undertaken as part of the Department of Industry and Science's National CO2 Infrastructure Plan (NCIP), aimed to identify and characterise indicators of natural hydrocarbon or fluid seepage that may indicate compromised seal integrity in the region. The survey was conducted in three legs aboard the New Zealand research vessel RV Tangaroa, and included scientists and technical staff from GA, the NZ National Institute of Water and Atmospheric Research Ltd. (NIWA) and Fugro Survey Pty Ltd. Shipboard data (survey ID GA0345) collected included multibeam sonar bathymetry and backscatter over 12 areas (A1, A2, A3, A4, A6b, A7, A8, B1, C1, C2b, F1, M1) totalling 455 km2 in water depths ranging from 90 - 430 m, and 611 km of sub-bottom profile lines. Seabed samples were collected from 48 stations and included 99 Smith-McIntyre grabs and 41 piston cores. An Autonomous Underwater Vehicle (AUV) (survey ID GA0346) collected higher-resolution multibeam sonar bathymetry and backscatter data, totalling 7.7 km2, along with 71 line km of side scan sonar, underwater camera and sub-bottom profile data. Twenty two Remotely Operated Vehicle (ROV) missions collected 31 hours of underwater video, 657 still images, eight grabs and one core. This catalogue entry refers to Autonomous Underwater Vehicle (AUV) still image data acquired during survey GA0345/GA0346/TAN1411. Following mapping with the shipboard multibeam, higher-resolution multibeam data were acquired in targeted areas using a Kongsberg Simrad EM2000 system mounted to the Fugro Echo Surveyor V (ES-5) AUV. This instrument had a depth rating of 3000 m, and surveyed the seafloor according to a pre-programmed mission plan. The AUV was fitted with a camera and light system designed to produce images of equal width and height (in the context of this survey, the images comprised 8 m by 8 m of seafloor). The equipment consisted of a light sensitive NEO 11 Megapixel 35 mm monochrome CCD (4008 x 2672) camera and two LED panels, each comprising 360 LEDs. High-resolution multibeam bathymetric data was collected together with side scan sonar and sub bottom profile data at an elevation of 30 m above the seafloor, and at line spacing's of 100 m. Overlapping high-resolution still photographs (captured every second) were then acquired on the survey lines at an elevation of 8 m above the seafloor. The AUV was equipped with an advanced real-time Aided Inertial Navigation System, which calculated the position, velocity and altitude of the vehicle and a HiPAP 500 USBL system was used to acoustically position the AUV. Underwater imagery was collected from two AUV missions in study Areas 3 and 4. During the 2nd AUV mission on 22 October, the vehicle encountered an obstruction on the seabed and became trapped despite commencing an emergency ascent sequence. The AUV was subsequently recovered from the seabed during salvage operations incorporated into the ROV phase of survey operations. A total of 24 877 still images were acquired in Area 3 and 20 743 in Area 4 over 58 and 56 line kilometres, respectively. Still images (.jpg files) are located in folder 'TAN1411_AUV_STILLS' with sub-folders named according to gear code (AUV= Autonomous Underwater Vehicle), mission and study Area (e.g. AUV_M2_A3 = still images acquired during AUV mission 2 in Area 3). USBL (Ultra-short baseline) text files (`TileCam.idx) are located in each sub-folder and provide continuous navigational information on location, time (UTC) and depth of AUV still imagery transect lines.

The Rowley Shoals/Offshore Canning and Roebuck basins survey was conducted on the RV Southern Surveyor survey SS06/2006 (GA-2408) between the 29th May and 22 June 2006. The primary aim of the survey was to identify any sites of natural hydrocarbon seepage, that may provide direct evidence for an active petroleum system within the sub‐surface. A secondary objective was to contribute to the understanding of the modern sedimentary and oceanographic processes influencing this part of the shelf, and to assess the nature of the benthic habitats. Underwater video was captured at 12 sites, with a minimum of 10 minutes to 60 minutes at each site, which totalled approximately 5.5 hours of footage. Video was recorded to mini DV tapes and copied to digital format. Descriptions of footage acquired during the underwater video tows are provided in the post-survey report (GA Record 2007/21 - Geocat # 65453). Please note that the underwater video is unclipped, contains descent and ascent through the water column, laser points in the video are reported to be spaced at 25 cm, and start locations of the underwater video camera stations are found in the Post-survey report.