support this hypothesis; i.e., five years after denuding, Chondrus was established under the Fucus canopy 

 in Zone 3 at FS, GN, and WP (Fig. 14). 



Barnacles also settle heavily in low intertidal areas (Fig. 16), more heavily under a Fucus canopy 

 (recolonization strips) than under a Chondrus canopy (controls). However, predatory snails have a greater 

 portion of the intertidal period to feed in the low intertidal, and barnacle mortality is higher than in Zone 

 2. Seasonal variability of barnacles is therefore more pronounced. 



Results from recolonization transect studies conducted under 2-unit operating conditions support the 

 conclusions that: 1) different components of the community recover at different rates, 2) recovery is faster 

 at exposed stations than at sheltered stations, 3) when present, grazing snails can maintain smooth rock 

 surfaces free from algal coloniz.ation, and 4) they are prevented from doing so indefinitely because of 

 surface heterogeneity provided by barnacles. Obviously, spatial and temporal distribution patterns of 

 grazers and predators are important influences on the structure of local intertidal communities. We examine 

 influences of grazing and predation on recolonization in the next section. 



Exclusion cages 



Results from four series of exclusion cage experiments supplement conclusions relating to the effect 

 of .seasonality on rates and patterns of recolonization, and provide information on the effects of grazing 

 and predation. Reduction of interspecific competition separates the physical and biological factors, that 

 influence distribution of intertidal organisms. Thus, it is possible to demonstrate whether rates and patterns 

 of recolonization at Fox Island-Exposed were directly influenced by proximity to the MNPS discharge. 

 Fach series of experiments consisted of nine caged areas and nine control areas at each of four stations; 

 the information is synthesized into a model of community development, outlining common sequences of 

 events and alternate pathways. Complete percent cover data for each experimental area during each 

 denuding are included in Appendix RS Id. 



General patterns of community development are illustrated in Figure 17. In each case, the vertical 

 axis represents relative, not absolute, abundance; absolute values varied among and within stations and 

 zones, but trends were similar. In general, values were higher at exposed stations than at sheltered, and 

 higher in mid and low intertidal areas than in Zone 1 , also higher under the cages than in the control areas. 



45 



