dimensions and their relatively brief duration in time. This gains expression in both the major subdivisions (formations or biomes) and in the smaller climax categories listed in Tables VI and VII. In following the schema of T. A. and Anne Stephenson (1949, p. 299) the present writer does not deny the probable validity of the formations noted previously (see Column 2 of Table VI). But for the same reason that “community” is used instead of “association”—i.e., that our basic knowledge of the true status of these divisions of the rocky shoreline is far from complete, it seems wiser at present to omit the term “formation” altogether, whilst relating each zone or subzone to its relevant section of the littoral.

To avoid confusion over upper, mid- and lower midlittoral, the Stephensons' (1952, p. 27) practice of naming each belt or zone after its chief characteristic—e.g., oyster zone, muddy zone, is perhaps the most convenient one to follow. On the other hand one finds it hard to avoid using such terms as upper, mid- and lower midlittoral (cf. Trevarthen, 1953, Fig. 3), especially where there is no conspicuous delimiting organism.

The adlittoral—i e., maritime, land vegetation on the most seaward continuous soil (Johnson and Skutch, 1928; Trevarthen, 1953) is beyond the scope of consideration of this survey.

Table VII reflects the change in ecological concepts which has taken place over the past 30 years. All New Zealand writers in this field have adopted a biotic approach, the chief improvements lying in a more precise delimitation of each unit, especially in relation to tide levels and amount of emergence and submergence, as well as in the recognition of seral groups, where they exist (Cooper, unpub.; Carnahan, 1952).

Oliver (1923) made the first genuine attempt to classify the confusing multiplicity of forms occurring between New Zealand tide marks, when the subject of littoral zonation was in its infancy. Regarding an “animal-and-plant formation” as “a biotic community with its principal ecological groups in definite combination and in definite relation to the habitat”, Oliver distinguishes between forma

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

The scheme followed in the present work coincides with that proposed by T. A. and Anne Stephenson (1949, p. 299) except in the use of ‘sub’ for ‘infra’-littoral. (In the first two columns the relation of zones to tide levels is only approximate.)

tions on rock and those on sand and mud. Thus a community on the above definition is equivalent to a wider unit embracing several “associations”, the latter corresponding approximately with the communities described in the ensuing text. Although the information in his paper is presented in such general terms, he describes a number of “formations” and “associations” which are prominent in the Hauraki Gulf, particularly those in Auckland Harbour and at Takapuna. His lists of subordinate species remain of considerable value to the littoral ecologist, but there is a need with our increased knowledge in this field for relating of the communities more precisely to their habitat. This has been achieved to a certain extent by later writers, for restricted localities.

Cranwell and Moore (1938, p. 380) take an association to be composed of species which “form an interdependent unit in equilibrium with the existing environment” the dominants having power to “exclude all invaders of similar life form and requirements”. These writers have contributed a sufficiently detailed account of the Poor Knights littoral communities to provide a basis of comparison with the present survey of the Hauraki Gulf. The concept of single dominance is still upheld; however, in the Nemastoma, Xiphophora, Durvillea, Carpophyllum elongatum and Lessonia associations. Like later writers (cf. Table VII), Cranwell and Moore have not been able to decide with certainty the status of certain groups which fall outside their definition of the littoral, in this case of lichens and Cyanophyceae; and so they have used the noncommital term “community”. Their employment of terms such as biocoenose, epibiose and hypobiose, used also by several European authors, has not been followed here.

Chapman, Beveridge, Dellow and Carnahan have carried the process of classifying the groups further by defining them more accurately in terms commonly used in climax land ecological terminology, while Carnahan and Cooper give due recognition to the seral nature of Zostera and Avicennia, together with several algal “socies” densely clothing mangrove stems and pneumatophores, dominated by Caloglossa, Catenella, Rhizoclonium and sometmes also the barnacle Elminius modestus. The fact that the mangrove swamp is seral in nature is considered sufficiently valid grounds for treating the associated algal groups in a like manner.

Owing to longer periods of observation on each area, the above-mentioned studies have shed more light on seasonal aspect communities. Following on this segregation of different types of community one could proceed almost indefinitely with the recognition of different units, whether one prefers to call them fasciations, sociations, clans, families, or similar terms; in fact the number of subdivisions made to date has appeared to depend at least partly on the length of time each writer spent in studying a particular locality, as well as on his or her chief interest. The fact remains that our knowledge of the majority of littoral communities comes from purely subjective observation; and until dominance is placed on a statistical basis we cannot be sure of the exact ecological status of a given assemblage. To avoid further multiple usage, therefore, the present writer feels it necessary to abandon the previous practice of attaching a narrowly defined term to a seemingly distinct littoral group of species.

Obviously, some communities will be more restricted in time and space than others, and certain dominants will hold sway according to minor variations in the habitat. It is not to be expected that any one community will be found in