As described in the previous section, the high intertidal (Zone 1) is mostly barren rock, with seasonal 

 occurrences of barnacles and ephemeral algae. Herbivory contributes to the structure of high intertidal 

 communities (Foertch and Keser 1981; Cubit 1984), but in general, physical factors are the controlling 

 influence (Menge 1975). Thus, the ephemeral algal turf that appears in spring in Zone 1 cages and control 

 areas is removed from control areas by grazing in early summer, and from caged areas by desiccation in 

 late summer. Barnacle settlement and growth are higher under the cages than in controls, but similarly 

 barnacles are usually lost from both areas with autumn desiccation. Degree of exposure had more effect 

 on community structure than did protection from grazing and predation. Growth and survival of barnacles 

 and ephemeral algae were higher at exposed stations (with more available moisture) than at sheltered 

 stations. For example, maximum barnacle coverage in Zone 1 cages at an exposed station exceeds 90%; 

 at a sheltered station less than 10% coverage. The seasonal cycles of settlement, growth, and mortality 

 were independent of the time of year in which denuding occurred (Fig. 17a). 



In mid intertidal cages, initial stages of recolonization were dependent mostly on what reproductive 

 units were in the water column when space was made available, i.e., barnacles in spring, Fucus in summer, 

 ephemeral algae throughout the year (Fig. 17b-f). Unlike the recolonization transects, where a barnacle 

 set was prerequisite for development of a Fucus canopy (crevices between barnacles provided refuge for 

 Fucus germlings that were otherwise vulnerable to grazing), substrata denuded in summer was colonized 

 quickly ( < 1 month) by Fucus, if grazers were excluded. 



Alternatively, exclusion of grazers retarded Fucus colonization, by permitting monopolization of sub- 

 stratum by ephemeral algae. In such instances, opportunistic algae occupied available rock, and Fucus 

 settlement was delayed until the following year. Both patterns are illustrated in the plot representing a 

 summer denuding (Fig. 17c and d). 



In all Zone 2 denudings, it was apparent that exclusion of predators and grazers from mid intertidal 

 areas had more effect on the rate of recolonization than on the ultimate structure of the recolonization 

 community. Fach experiment ran for 15 months, but there was a clear indication that, had they continued, 

 the caged communities would have resembled nearby "undisturbed" communities, i.e., a Fucus canopy 

 over a barnacle understory. These findings support the conclusion that recovery from disturbance is 

 deterministic, even if transition states are variable (Paine 1984). 



48 



