Community structure and zonation patterns of rocky shores at high latitudes: an interocean comparison
To compare the structure of communities of sheltered rocky shores at high latitudes in widely separated regions in order to search for similarities and dissimilarities. Locations
Three regions with similar physical environments were investigated: south-western Iceland, southern Alaska, and the Magellanic region. Methods
Cover of algae and sessile animals was estimated and animals were counted in quadrats laid along a number of transects in each of the three regions. Transects were divided into 10 height intervals with reference to tidal levels, and comparisons made on the basis of average distributions of species within each of 11 defined guilds. Results
Icelandic and Alaskan shores had a number of species in common, although these were in the minority. No species was common to southern and northern regions, with the possible exception of Mytilus edulis. Most defined guilds were present in all regions, but their taxonomic composition was often radically different. The molluscan filtering guild was the guild showing greatest similarity in the three regions, while canopy-forming algae were absent from the Magellanic region and balanid filterers were virtually absent from Iceland. Crustacean grazers were only prominent in Iceland, and molluscan grazers were more common there than in the other regions. Zonation of species was recognizable and species diversity increased seawards in all regions. Main conclusions
Few, if any, assembly rules could be detected when comparing regions, except for the universal increase in species diversity seawards. The absence of ‘key’ species in one region may affect the distribution of other species. All major opportunities for resource exploitation are not utilized in all regions. In some instances it is possible to recognize ecological equivalents in different regions. No recurrent distinct zones are apparent when regions are compared. A universal descriptive system of zonation, such as developed by Stephenson & Stephenson (1949) and Lewis (1964), was found to be not applicable.