Developing long-term monitoring programs in offshore waters with little prior knowledge: Applying a novel sampling design to inventory biological assets

Australia’s declaration of a Commonwealth Marine Reserve (CMR) Network, has resulted in a pressing need to 1) provide quantitative inventories of the species and communities represented within reserves, and 2) develop methods that enable trends in their composition, abundance and distribution to be monitored. These objectives require a general but flexible long-term monitoring framework where sampling is conducted using non-destructive methods such as towed video, baited underwater video and still imagery, and informed by habitat data derived from multibeam acoustics. Here we describe the application of a flexible sampling program tailored for assessing the biodiversity associated with continental shelf habitats in the Flinders CMR, north-east Queensland. This program is based on a probabilistic and spatially-balanced sampling design called Generalized Random Tessellation Stratified (GRTS). While GRTS has been used for continental and regional-scale monitoring programmes in the United States, its application to marine systems in Australia is new. Within the Flinders CMR, little previous knowledge existed on the spatial distribution of shelf habitats, necessitating a two-phase sampling approach. Firstly, GRTS was applied to gain an understanding of the distribution of shelf habitats across the CMR shelf; secondly, this information was used to design a biological sampling program. Here we 1) discuss the suitability of this method for conducting ship-based sampling programs with little prior information over large spatial scales, and 2) present the first quantitative assessment of the distribution, composition, and abundance of fish and fish assemblages within the Flinders CMR, as determined with baited underwater videos.