A qualitative model had earlier been proposed to explain why ''good competitors that are poor recruiters'' (GC) are more abundant in the sessile assemblages on large substrata than ''Poor competitors that are good recruiters'' (PC), the PC being relatively more abundant on small isolated patches of substratum. One consequence of this model was that the composition of the fauna, and nature of differences between patch sizes, would vary with larval recruitment rates. In this study, sessile fauna were recorded on shells of the bivalve Pinna and on pier pilings at three sites, and an experiment on colonization of plates of five sizes was done at two sites, chosen from earlier work to have different recruitment rates. In the comparison of Pinna and pilings, predictions were confirmed for colonial ascidians, bryozoans, the lumped groups PC (bryozoans, tubeworms and barnacles) and ''very poor competitors'' (tubeworms and barnacles), partially for GC, but not for sponges. In the case of colonial ascidians and GC, the proposed mechanism did not fully explain the results. In the plate experiment, predictions were confirmed for ascidians, PC, barnacles and tubeworms, partially for GC but not for sponges or bryozoans. Over the timescale of the experiment, the effect of heavy recruitment of PC was more important than their subsequent dominance by GC. It is suggested that with more time the effect postulated in the model would have become clearer, but also that it would have required very much greater resources to make a more effective test of the model underlying this work, which involves complex population structures and spatial, temporal and spatio-temporal variation. Meanwhile, the study raises questions worthy of study at smaller scale; these concern the population structure and dynamics of particular species, and the responses of major animal groups to variations in food and physical conditions.
