High polychaete diversity in the Gulf of Guinea (West African continental margin): The influence of local and intermediate scale ecological factors on a background of regional patterns

The Tropical East Atlantic is one of the least studied areas in the world's oceans, and thus a blank spot on the map of marine studies. Shaped by dynamic currents and shifting water masses, it is a key region in discussions about marine ecology, biodiversity, and zoogeography, while facing numerous, poorly understood, and unmonitored threats associated with climate change, acidification, and pollution. Polychaete diversity was assessed along four transects along the Ghana coast, from shallow to deep bottoms and distributed along the whole upwelling marine ecoregion. Despite high sampling effort, steep species accumulation curves demonstrated the necessity of further sampling in the region. We observed zonation of fauna by depth, and a decrease in species richness from 25 m to 1000 m depth. Polychaete communities were influenced by sediment type, presence of oxygen minimum zones, and local disturbances caused by elevated barium concentrations. Similar evenness along the depth gradient reflected the importance of rare species in the community structure. Differences in phylogenetic diversity, as reflected by taxonomic distinctness, were small, which suggested high ecosystem stability. The highly variable species richness at small scale (meters) showed the importance of ecological factors giving rise to microhabitat diversity, although we also noticed intermediate scale (50-300 km) differences affecting community structure. About 44 % of the species were rare (i.e. recorded only in three or fewer samples), highlighting the level of patchiness, while one fifth was distributed on all transects, therefore along the whole upwelling ecoregion, demonstrating the influence of the regional species pool on local communities at particular stations. Our study yielded 253 species, increasing the number of polychaetes known from this region by at least 50 %. This casts doubt on previous findings regarding Atlantic bioregionalization, biodiversity estimates and endemism, which appear to have been more pronouncedly affected by sampling bias than previously thought.