Deep-sea biotope classification using opportunistic sampling: insights for future management

An iterative approach to optimise deep-sea biotope classification using a combination of acoustic data and Remotely Operated Vehicle (ROV) video footage was developed and tested at the Tropic Seamount site in the Northeast Atlantic. Two methods for biotope classification were compared: a top-down approach based on acoustic substrate classification followed by biological characterisation, and a bottom-up approach using multivariate analysis of biological assemblages only. Video transects were analysed at two spatial resolutions (200 m and 50 m segments) to assess scale effects on biotope delineation. Biotopes were classified using a combination of geological and biological data with each biotope representing a distinct combination of substrate types and their associated benthic assemblages. The bottom-up approach using 50 m segments identified 12 distinct biotopes with stronger environmental correlations compared to broader classifications at 200 m scale. This study demonstrates that shorter transects (50 m) combined with bottom-up sampling approaches are preferable for capturing the ecological heterogeneity characteristic of deep-sea seamount environments, with important implications for vulnerable marine ecosystem identification and spatial management.