EUSeaMap. A European broad-scale seabed habitat map

In order to most benefit from the potential offered by the European marine basins in terms of growth and employment (Blue Growth), and to protect the marine environment, we need to know more about the seafloor. European Directives, such as the MSFD, but also the Horizon 2020 roadmap explicitly called for a multi-resolution full coverage of all European seas including bathymetry, geology and habitats. The present work, following on a suite of past initiatives, has made a big step forward in this direction. It has first boosted the collation of existing maps from surveys by setting up a framework and a procedure to encourage people to submit their maps and data. This resulted in a more attractive EMODnet seabed habitat portal and a snowball effect with more and more people willing to join. However, collation will eventually come to an end and as new creations of seabed habitat maps are so complex and time-consuming, a cost-efficient way to meet the need for a full-coverage habitat map was found to be low-resolution maps and models to predict seafloor habitat types. The broad-scale map referred to as EUSeaMap has been created by this project and after the first two phases it now covers all European basins from the Barents Sea to Macaronesia and to the Black Sea. By harmonising mapping procedures - based on the EUNIS classification - and fostering a common understanding among seabed mappers in Europe, EUSeaMap provides today the community with a comprehensive, free and ready-to-use map that can find applications at regional scale for management and conservation issues. Tables and maps for all basins can be found in section 3 “Results and disciussions”. The project has played a key role in giving feedback to other EMODnet communities dealing with bathymetry, geology and biology, all essential data sources for the broad-scale map. It has also improved the understanding of the EUNIS habitat classification - with a focus on the Adriatic and the Black Sea - by better specifying transitions between classes based on benthic ground-truth data. It has fostered the development of oceanographic variables such as light, waves and currents that have a strong bearing on habitats. Finally it has also been instrumental in developing map confidence assessment methods that account for the broad spatial variation in data sources quality and for uncertain boundaries between habitat classes. The EUSeaMap methods are repeatable and ensure that the predictive maps can continue to be improved in the future, as a result either of EUNIS enhancements or increase in resolution. From today’s 250m resolution it is likely that new deliveries of enhanced source layers due to steady progress in oceanography and geophysics will enable constant refinement of the maps over time.