Twelve sites were chosen from areas of barrens in depths of 5 to 8 m along the coastline of Sydney, and three of these were randomly assigned to each treatment. Sites were separated by between 200 m and 5km and covered an area of reef of 400 to 600m2. Densities of sea urchins prior to manipulation were estimated by counting individuals in ten, haphazardly placed, 10 x 1m transects within each site. In March 1999, the density of sea urchins at each site was reduced to the required levels by removing haphazardly chosen individuals. Sites were revisited approximately every 2 months during the first 12 months of the experiment, and then every 3 months until March 2001, and this maintained the treatments within + 20% of the required density of individuals per m-2.
Continual urchin removal every 2 months was needed. Turf algae returned, but not kelp.
After manipulation, the cover of crustose coralline algae in treatments where density had been reduced by 33% and 66% were both reduced to ~25% of their original cover. Where density was reduced by 100%, crustose coralline algae dropped to ~50% of the original cover. Changes in cover of filamentous algae, Ralfsia and foliose algae were generally complementary to those for crustose coralline algae.