Detection and characterisation of deep-sea benthopelagic animals from an autonomous underwater vehicle with a multibeam echosounder: A proof of concept and description of data-processing methods

Benthopelagic animals are an important component of the deep-sea ecosystem, yet are notoriously difficult to study. Multibeam echosounders (MBES) deployed on autonomous underwater vehicles (AUVs) represent a promising technology for monitoring this elusive fauna at relatively high spatial and temporal resolution. However, application of this remote-sensing technology to the study of small (relative to the sampling resolution), dispersed and mobile animals at depth does not come without significant challenges with respect to data collection, data processing and vessel avoidance. As a proof of concept, we used data from a downward-looking RESON SeaBat 7125 MBES mounted on a Dorado-class AUV to detect and characterise the location and movement of backscattering targets (which were likely to have been individual fish or squid) within 50 m of the seafloor at ~800 m depth in Monterey Bay, California. The targets were detected and tracked, enabling their numerical density and movement to be characterised. The results revealed a consistent movement of targets downwards away from the AUV that we interpreted as an avoidance response. The large volume and complexity of the data presented a computational challenge, while reverberation and noise, spatial confounding and a marginal sampling resolution relative to the size of the targets caused difficulties for reliable and comprehensive target detection and tracking. Nevertheless, the results demonstrate that an AUV-mounted MBES has the potential to provide unique and detailed information on the in situ abundance, distribution, size and behaviour of both individual and aggregated deep-sea benthopelagic animals. We provide detailed data-processing information for those interested in working with MBES water-column data, and a critical appraisal of the data in the context of aquatic ecosystem research. We consider future directions for deep-sea water-column echosounding, and reinforce the importance of measures to mitigate vessel avoidance in studies of aquatic ecosystems.