Primary succession on a seasonal tropical rocky shore: the relative roles of spatial heterogeneity and herbivory

Abstract

Hong Kong is within the tropics and has a seasonal climate. In winter, shores support patches of ephemeral macroalgae and areas of seemingly bare rock close to crevices where mollus- can herbivores are abundant. Using a factorial design of herbivore exclusions in areas far and close to crevices, the development of algal assemblages was monitored in mid-shore, cleared areas, in win- ter. To estimate the role of herbivore mucus deposition, half the treatments received a mucus appli- cation. Algal development was estimated from macroalgal and biofilm development and chlorophyll a levels. In all areas, biofilms (diatoms, unicellular cyanobacteria) developed rapidly in herbivore exclusions followed by ephemeral macroalgae (Enteromorpha spp. and Porphyra suborbiculata). In herbivore access treatments, however, the algal assemblage was influenced by treatment location; few macroalgae developed in areas close to crevices, and the rock was dominated by cyanobacteria. A negative relationship between macroalgae and biofilms suggested that ephemeral algae were competitively dominant. In areas distant from herbivore refuges, ephemeral macroalgae did develop, illustrating that the effectiveness of molluscan herbivores was limited to 50 to 100 cm from these refuges. The absence of large herbivorous fish, and the sparse numbers of herbivorous crabs at this site, means that algae can achieve a spatial escape from consumption, and where this occurs compe- tition between producers is important in assemblage development. Mucus appeared to play a limited role, only sometimes stimulating initial stages of unicellular cyanobacteria and macroalgae. With the onset of summer, macroalgae died back, and rock space became available for colonization. Unicellu- lar cyanobacteria developed rapidly but were replaced in all treatments by the encrusting macroalga, Hapalospongidion gelatinosum, which dominated treatments until the end of the experiment. On seasonal, tropical shores processes influencing community structure can, therefore, be temporally variable and their relative importance, even at the same shore level, can change with season.