Broad-scale spatial factors outweigh the influence of habitat structure on the fauna associated with a bioengineer

Mechanisms operating at regional scales, such as nutrient supply or temperature, are often modified within biogenic habitats via changes in habitat structure created by the bioengineer. By examining beds of the brown mussel Perna perna and its associated fauna, we unravelled the relationship between local (structural) and regional effects. We made specific predictions based on the structure of mussel beds from either the eastern or western region of the warm-temperate (Agulhas) biogeographic province of South Africa. We created patches of mussels based on the structure of mussel beds from both of these regions on each of the 2 shores within each region. We sampled the fauna associated with these biogenic habitats after 3 mo, and at the same time sampled undisturbed beds of mussels as a control to determine if artificially created mussel beds were similar to natural mussel beds. There were some experimental artefacts and the assemblages associated with the east structure were generally not the same as those from naturally occurring eastern mussel beds. Overall, our hypothesis was not supported; we found no evidence of the influence of the structure of mussel beds on assemblages of associated fauna. We found that within the same biogeographic province, assemblages only tens of km apart had very different groups of species. Contrary to previous studies, there was a de-coupling between the habitat-forming taxon and its associated fauna. The existence of increased productivity through upwelling may reduce the importance of facilitation in both our study regions, so that larger-scale (shore or regional) effects had a stronger influence on associated fauna than habitat structure. Nevertheless, we suggest that, regardless of upwelling, shore and regional level effects can override those of habitat structure.