The IMOS AUV Facility participated in a two week deployment of the Autonomous Underwater Vehicle (AUV) Sirius on the Tasman Peninsula in SE Tasmania and in the Huon Marine Protected Area (MPA) to the South West of Hobart.  The objective of the expedition was to document biological assemblages associated with rocky reef systems in deep shelf waters beyond normal diving depths. Detailed multibeam sonar bathymetry data were previously collected by Geoscience Australia using a Simard EM3002 multibeam sonar system, Applannix motion sensor and C-Nav GPS to provide high-resolution DEMs of the study areas. The DEMs were used to determine suitable AUV survey locations and to identify any hazards to operation. At each location, multiple reefs were surveyed at a range of depths from approximately 50 m to 100 m depth. Where distinct ectones (e.g. reef to sand) are present, transects were designed to cross transition zones and help determine the uniqueness of ectonal assemblages. Replication depended upon site logistics, however, dive profiles were designed to provide sufficient replication to quantitatively determine abundances of key species/features within depth strata, within reefs, between reefs (km to 100 km scale), and between differing levels of reef complexity.

While inshore systems are relatively easy to access and describe using methods such as dive surveys, offshore systems have remained relatively unknown, because of the expense and complexity of available survey methods. A recent and very significant development contributing to our understanding of the physical environment of shelf habitats has been multibeam sonar and the interpretation of its associated backscatter. This has opened up opportunities for developing predictive capacity in this field where matching biological datasets are available. However, the effectiveness of this technique has yet to be fully tested as an appropriate surrogate for predicting patterns of biodiversity because of the lack of matching biological datasets collected at the same spatial scales and locations as fine scale acoustic surveys.  In this context, the AUV deployments reported on here were part of a multi-disciplinary experimental program in eastern Tasmania, where the analysis of covariance is to be undertaken on co-located fine-resolution seabed habitat data, provided by the EM3002 multibeam sonar coverages, and biological datasets collected at similar spatial scales by Remotely Operated Vehicles (ROVs), Baited Underwater Video systems (BRUVs), towed video and the AUV. This research is being undertaken by the University of Tasmania and Geoscience Australia as part of the Marine Biodiversity Research Hub, which is a collaborative program funded under the Commonwealth Government's Commonwealth Environmental Research Facilities (CERF) Program. Within this Hub, several interrelated projects are designed to develop and test appropriate surrogates for biodiversity and incorporate these into an advanced predictive framework that covers a range of spatial scales.

The AUV data enables a finer scale coupling of biological datasets with multibeam bathymetry than data collected through the use of ROVs, BRUVs and towed video alone, because of geo-referencing errors associated with USBL tracking systems. This additional data is expected to allow scale matching errors to be examined in more detail and allow surrogacy to be examined at the finest possible scale. Ultimately, CERF researchers expect to be able to  compare the relative efficiency of using AUV, ROV and towed video systems for shelf habitat biological surveys. In addition, the high resolution images produced by the AUV are expected to significantly enhance the ability to identify taxa, adding finer taxonomic resolution, and hence value, to the data collection.