Predicting the consequences of anthropogenic disturbance: large-scale effects of loss of canopy algae on rocky shores

Anthropogenic disturbances affect natural populations and assemblages by interacting with fundamental ecological processes. Field experiments simulating the effects of human activities at the appropriate spatial and temporal scales are useful to understand these interactions and eventually to predict their ecological consequences. In the Mediterranean, low-shore habitats of rocky coasts are often dominated by canopy algae Cystoseira spp., but these species are frequently replaced by assemblages of turf-forming algae and mussel beds. We propose that anthropogenic disturbance is the proximate cause of loss of Cystoseira in the Mediterranean, and that the disappearance of canopy algae causes an increase in cover of turf-forming species in disturbed habitats. Two hypotheses were investigated to test this proposition: (1) canopy algae will be dominant in relatively pristine habitats while turf-forming algae will be more abundant in urban areas, and (2) removal of canopy algae in unpolluted areas will result in the development of assemblages similar to those found in urban areas. We tested the first hypothesis by comparing patterns in abundance of Cystoseira and turf-forming algae at a number of locations in urban areas and in areas far from distinct sources of anthropogenic disturbance in the northwest Mediterranean. The second hypothesis was tested by conducting a largescale manipulative experiment, involving the experimental removal of Cystoseira and several spatial and temporal repetitions of the manipulation. Turf-forming algae were always dominant in urban areas where Cystoseira was nearly absent. In contrast, canopy algae often dominated rocky shores in relatively pristine areas of the northwest Mediterranean. The removal of Cystoseira generally caused an increase in the percentage cover of turf-forming algae and a decrease in the abundance of invertebrates. These changes were already evident 4 mo after manipulation and were consistent at the spatial and temporal scales examined in the study. Assemblages in cleared patches were qualitatively similar to those occurring where Cystoseira was naturally absent, but quantitative differences in the relative abundance of several taxa were still evident by the end of the study. These results support a causeeffect relationship between anthropogenic disturbance and loss of Cystoseira in the northwest Mediterranean and allow for quantitative predictions of the indirect consequences of disturbing canopy algae for the whole understory assemblage. Management options aimed at conserving these plants should simultaneously preserve other components of the assemblage.