Given current rates of extinction and potential failure of ecosystem services, it is critical to understand the ecological consequences of biodiversity loss. We designed a manipulative 2-factorial field experiment to test for single and interactive effects of depletion of a key species, the canopy-forming seaweed Fucus serratus (present vs. absent) and disturbance (undisturbed vs. application of 50% of the required force to remove the entire macrobenthic community) on temporal variation of diversity. The experiment was conducted at two intertidal sites (north and west) of different wave exposure at the island of Helgoland between March and December 2006. Diversity was recorded prior to and within 3 days after manipulations (March) and in June, September, and December 2006. The coefficients of variability (100 χ SD/mean) of species richness and evenness in each plot were calculated separately for winter and summer, and analysed as dependent variables. At both sites, Fucus removal increased variability of evenness regardless of disturbance and season. In addition, disturbance had site dependent effects on evenness, destabilising this variable at the northern site, but stabilising evenness at the western site. On the other hand, species richness remained unaffected by Fucus removal, disturbance, and interaction at both sites. Species richness was more variable in winter than summer at the northern site, without seasonal changes at the western side. At both sites, species composition presented seasonal changes that were slightly accentuated by Fucus removal. In this system, Fucus serratus did not provide resistance to disturbance. Our results suggest that canopy-forming species may have lesser importance for persistence of diversity and community structure than previously reported.