An expeditious low-cost method for the acoustic characterization of seabeds in a Mediterranean coastal protected area

Posidonia oceanica meadows are ecosystem engineers which, despite their ecological relevance, experiencing habitat fragmentation and area decrease. Cartography information on the status of these habitats is key to an effective maritime spatial planning management for conservation. In this work we apply acoustic methodology map benthic (substrate vegetation) in archipelago Natura 2000 Network close coast Murcia (SE Spain) where dense sparse areas P. oceanica, patches Cymodocea nodosa appear over a sandy had bottom. The uses dual frequency (200 kHz 38 kHz) acquired with single-beam echosounder develop bathymetry, performs sea bottom vegetation supervised classifications, using video scuba diver observations as groundtruthing data. Sea was classified from features first second 200 echoes into 5 substrate classes random forest classifier: sand, fine coarse hard bottoms patches. extracted "above-bottom" part echo (height backscattering intensity) both frequencies, resulting also class classification: C. meadows, dispersed sand patches, no-vegetation; according random-forest Gini index, were most informational variables classification. validation accuracies classifications 85% (substrates) 70% (vegetation), reported literature when similar number classes. results article (including bathymetric, thematic maps), together described used, contributions that can improve continuous monitoring Mediterranean seagrasses.