Publication Abstract

Management of the marine environment is progressively adopting an ecosystem-based approach, which necessitates an integrated strategy for monitoring. Collecting the various required data types simultaneously is often challenging, primarily because of the specific gear needed (such as grabs, trawls, and video and acoustic equipment) and discrepancies in their spatial and temporal scales. To accommodate the need for the collection of habitat data from fisheries surveys on the RV Endeavour semi-autonomous multibeam data collection, processing and classifying has been developed. We have tested the influence of the semi-autonomously collected multibeam data, in comparison to seabed data (modelled kinetic energy and depth) gathered by other means, on the fit of negative binomial models of 32 demersal species from the western Channel. The study analysed 508 sampling stations across six habitat classes with varying depths and kinetic energy levels. We found that for the habitat class variable produce by semi-autonomous means was as valuable to the modelling of the demersal as other data sources. A semi-automated approach to broadscale habitat mapping has the potential to inform species distribution modelling required from an ecosystem-based management approach in the future.

Jennifer I fincham * and Jon Barry*