Patterns of disturbance and recovery in littoral rock pools:nonhierarchical competition and spatial variability in secondary succession

Abstract

Patterns of &sturbance and recovery were investigated in algal-dominated littoral rock pools on the west coast of Italy from February 1991 to November 1993. The assemblage was a mosaic of canopyforming species interspersed amongst patches of turf-forming and encrusting algae. The effects of natural disturbance were Investigated by monitoring the size distributions of patches of open space (mainly consisting of encrusting corahnes and bare rock) In fixed plots over 2 yr. These plots were distributed in pools maintained at natural and reduced densities of herbivores. In both cases, the mean percentage of open space did not change greatly from February 1991 to February 1993, nor were there differences attributable to grazing by the end of the monitoring period Shifts in size distributions of natural patches of open space occurred durlng the 2 yr of observations The area accounted for by large size-classes (> 1200 cm2) increased In pools exposed to natural densities of herbivores and decreased In the others. Howevcr, grazers never affected size distributions significantly. Overall, results suggested that the provlslon of open space in this system was mainly due to physical disturbances. A multifactorial experiment was used to assess the effects of herbivores (natural vs reduced densities), size of patch (17 X 17 vs 32 X 32 cm) and time of clearance (February vs August) on algal colon~zation In expelimentally cleared plots. An additional experiment tested for the effects of algal turfs on the establishment of canopy-forming algae m pools where grazers had been removed. The results of the recolon~zation experiment ~n d ~c a t e d that herbivores, slze of patches and time of clearance exerted important but variable effects on patterns of algal recolonization. Herbivores reduced but could not prevent algal recovery. Their effect was particularly evident for the algal turfs throughout the study, but it was not significant for Cystoseira spp. by the end of the experiment. There were significant size X time of clearance interactions affecting patterns of algal colonizat~on early in the study. The percent cover of turf-forming and canopy-forming plants was greater in the small plots produced in February 1991 than in the other treatments. These effects were more evident in pools exposed to natural densities of grazers and during the early stages of colonization. The results also suggested that competltlon among Cystosejra spp. and algal turfs was nonhierarchical, the outcome of the interaction depending primarily upon the relative density and life stacjc of the interacting organisms, rather than on their ident~ty. This view was supported by the fact that estdbl~shed mats of algal turfs drastically inhibited the recruitment of Cystoseira, whilst no strong interaction occurred if recruitment by these species was synchronous. Thus, variability in the intensity and timing of colonization could have important effects for subsequent patterns of algal dominance in thls system, but only when grazing was unimportant and algal colonization plentdul. Alternative models were formulated to explain spatial and teniporal patterns in pools where these conditions did not occul-