Testing scale variance in species-area and abundance-area relationships in a local assemblage:: an example from a subtropical boulder shore

Constancy or scale variance of species - area and abundance - area relationships has rarely been considered within relatively small spatial domains of a local assemblage. Patterns of species/abundance - area relationships were experimentally investigated in a stone-associated molluscan community on a subtropical boulder shore. In order to systematically examine the effects of variation in habitat area while maintaining other habitat characteristics constant through time, naturally occurring stones were selected and divided into different size classes according to surface area and used as habitat units for regular monitoring of a mobile molluscan community. Species richness and abundance (number of individuals and biomass) of molluscs scaled with stone area, but the power or double-logarithmic regression was not always the best description of the species - area relationship. Seasonal scale invariance was shown by the species - area relationship, whereas scale variance was clearly recognizable in the abundance - area relationships. The latter phenomenon was generated mainly by large stones contributing disproportionately to increases in molluscan abundance in particular. Furthermore, there was a negative effect of small habitat area whereby molluscan abundance was disproportionately reduced on small stones. Some temporal variation in the observed patterns was also recognizable, with higher species richness and abundance in spring than in winter, again with larger stones showing preponderant importance. This study thus demonstrates the significance of scale variance/invariance in species/abundance - area relationships, even within relatively small spatial scales of local habitat.