Answers are grounded in this paper and cite their sources.
One of the few generalizations that we can make about vegetation is that it is spatially heterogeneous. Its constituent species show pattern, i.e. departure from randomness of distribution. This is obvious where pattern is intense, as in the contrast between evidently different communities or where what we commonly accept as one community shows mosaic structure. The explanation of such obvious pattern was the challenge that gave rise to plant ecology. The increasing application of numerical methods to the study of plant distribution over the last quarter of a century has made it clear that randomness is the exception even over very small areas. Ordinational studies have exposed larger-scale patterns of less intensity, and the simple technique of nested-block analysis of variance and covariance of data from transects and grids has permitted the description of the patterns of species and the relation between them at smaller scales (Greig-Smith 1952, 1961a, 1964; Kershaw 1957, 1960; Hill 1973). The patterns of environmental factors can be analysed in the same way. Such pattern analyses are conveniently presented as graphs of mean square or covariance against block size, in which peaks indicate the scales of pattern present. If data are of a form for which random distribution has an expected variance, e.g. density (Poisson), cover estimated from point quadrats and local frequency (binomial), scaling of the observed variance as mean square/expected variance allows the calculation of confidence limits for random distribution (Thompson 1958). Although some less abundant species in a community may occasionally show no evidence of departure from randomness, I know of only one case where a major component is apparently randomly distributed, apart from a slight tendency towards regularity at the smallest scale indicated by the only points outside the confidence limits being below the lower limit. This is for Shorea albida Sym. in lowland swamp forest in Borneo, where it forms the greater part of the canopy layer (Fig. 1). Even in apparently uniform and species-rich rainforest some at least of the commoner species show patchiness (Poore 1968; Ashton 1969). It has commonly been accepted that the composition of vegetation is determined by its habitat-'vegetation is the best measure of environment'. With the realization that pattern is ubiquitous and extends to the smallest scales, two questions arise. How far can habitat differences account for less intense and small-scale pattern? What other mechanisms may cause or reinforce spatial heterogeneity or patchiness? My object today is to examine these questions.
No takes yet. Share an insight, caveat, or question.
P. Greig-Smith (1979) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: