Ecological comparison between artificial reefs with automated procedures

Abstract

Waste materials from metallurgic industry were attempted to be reused for constructions. Unfortunately, white slag showed low performance. Here we present a work on how this industrial by-product was successfully used to manufacture and deploy Artificial Reefs (ARs) in coastal marine environment. The OBSEA cabled observatory cameras monitored two white slag ARs’ prototypes and a concrete biotope with cameras capturing photos each minute from 18th July 2023 to 24th October 2024. The species classification and counting were automated using machine learning algorithms, and ANalyses Of VAriance (ANOVA) were performed on AR Multiparametric Index (ARMI) values to compare ARs ecological performances. Bigger and older ARs proved to perform better. However, camera focus (i.e.; benthic vs benthic-pe-lagic realms) has to be considered in these analyses. The computation of ARMI indices from automated image processing can delineate a pipeline to inform project managers on the status of ARs marine restoration in real-time. This is in accordance with European and global policies goals, as Marine Strategy Framework Directive and United States Decade of Ocean Science for Sustainable Development.