Assessment of the application of each multibeam echosounder data product for monitoring of Laminaria digitata in the UK

Amid warming seas, high rates of pollution and declining fish stocks observed around the UK, the vital role of kelp as ecosystem mediators on our coastlines is increasingly significant; currently estimated at £500 billion. Extensive research on the rapid decline of kelp forests and its potential consequences has prompted the initiation of numerous conservation efforts. This research set out to determine the applicability and efficiency of a less invasive, remote sensing technique for monitoring kelp. A high resolution multibeam echosounder (MBES) survey was performed to acquire depths, backscatter and water column data in an area known to have kelp. An evaluation of different combinations of the MBES data products for kelp forest monitoring was carried out. An image-based processing methodology using a random forests algorithm was used to generate classification models, which were trained and tested using ground truth samples obtained through video imagery. This study reports climbing model accuracy scores from 62.2% (±11%, 1σ) to 90% (±10%, 1σ) on consecutive input of data products, indicating MBES as an effective tool with respect to other technologies. When considering practical difficulties associated with simultaneous record of all data products against their individual value, this study suggests that bathymetry and backscatter products deliver greatest value for distinction of small form kelp, while angular response analysis and water column data deliver lesser value but are required for optimised accuracy.