classification (1970:83) separates categories by 

 "combinations of environmental conditions 

 characteristic of various parts of the coastline." 

 An example unit is the Pacific Southwest. Lynch 

 et al. (1976) do not explicitly describe the cri- 

 teria they use to separate units, but the criteria 

 appear to be geological history, tidal amplitude, 

 weather, currents, latitude, and estuarine en- 

 vironments. An example of a unit in the Lynch 

 et al. classification would be the Columbia-North 

 Pacific Region. 



An excellent example of classification of 

 coastal areas on purely physical (chemical, geo- 

 logical, etc.) attributes is Dolan et al. (1972). 

 They use atmospheric and marine climates (cur- 

 rents) as well as coastal materials and configura- 

 tion to separate units. An example of a unit in 

 the Dolan et al. (1972) classification would be 

 Regime VII: Subdominant-Maritime Polar-Marine- 

 Divergent /Convergent. 



Each of the above types of classification may 

 be put to a number of uses, and each is well 

 suited to answering certain types of questions. 

 However, information obtained by applying one 

 type of classification may be useless in trying to 

 solve problems best addressed by application of 

 another type of classification system. A few ex- 

 amples will clarify this. If all coastal areas of the 

 United States were classified according to Odum 

 et al. (1974), then the question, "What is the 

 mixing pattern of estuary X?", could not be 

 answered because their classification only con- 

 sidered energy inputs. If all coastal areas of the 

 United States were classified according to Inman 

 and Nordstrom (1971), then the question, "How 

 many surface hectares of coastline are covered by 

 kelp beds?", also could not be answered because 

 Inman and Nordstrom only considered geologi- 

 cal processes. The information collected for 

 either classification would not be incorrect, but 

 would be inappropriate to answer the types of 

 questions being asked. Thus it is obviously 

 necessary to select a classification which best 

 answers the question or questions being asked. 



The objective of this project is to formulate 

 a hierarchical regional classification scheme for 

 coastal ecosystems of the United States and its 

 territories, based on the physical characteristics 

 of those areas. The question the classification is 

 designed to address is the following: "How can 

 the coastline of the United States be partitioned 

 to best separate ecosystems, when the purpose of 

 defining these ecosystems is to understand and 



subsequently to make predictions about how 

 specific types of perturbations in specific geo- 

 graphical areas will affect those ecosystems 

 hydrologically, structurally, functionally, and 

 biologically?" Structural and functional classi- 

 fications do not adequately address the above 

 stated problem because they are not geograph- 

 ically oriented. The geographic orientation is 

 essential to making predictions about a specific 

 estuarine or marine system. Thus, a regionaliza- 

 tion is necessary. 



Since delineation of ecosystems is the primary 

 interest, a regionalization based on physical 

 parameters is more appropriate than a biogeo- 

 graphical regionalization. Although the argument 

 is frequently made that the biota integrate all 

 the physical attributes of their environment, 

 two factors argue against a biotic regionalization 

 for answering the objective of the study. The 

 first is historical accident of distribution and/or 

 extinction. For example, a group of organisms 

 might be absent from an area which they could 

 inhabit simply because they were never dis- 

 tributed there or had become extinct in that area 

 because of environmental or man-induced pertur- 

 bations. Regionalization with respect to ecosys- 

 tems should not be determined by historical 

 accident. 



The second factor supporting an argument 

 against biogeographical regionalization is the 

 difficulty of selecting the group or groups to 

 represent the whole ecosystem. Questions have to 

 be answered if benthic or motile forms, plants 

 or animals, vertebrates or invertebrates, or vascu- 

 lar or nonvascular plants are the appropriate 

 organisms to consider. A regional scheme based 

 on physical parameters eliminates these problems 

 since physical factors constrain the distribution of 

 ecosystems. Thus a regionalization is most appro- 

 priate to answer the originally stated objective. 



The above argument should not be construed 

 to mean that the distribution of biota should 

 not reflect the distribution of coastal ecosys- 

 tems. If the theory that biota integrate their 

 physical environment is correct, then they should 

 reflect, though perhaps imperfectly by their own 

 distribution, the distribution of coastal ecosys- 

 tems. In fact, the distribution of biota would 

 provide an excellent method for testing a region- 

 alization based on physical parameters. 



The classification proposed by Dolan et al. 

 (1972) is extremely well done and well docu- 

 mented. It was not used to satisfy the objective 



