Numbers for species (rows) represent abundance in quadrats (columns); a dot indicates that the species was absent from the quadrat. Zeros and ones on the right-hand side and the bottom of the table indicate the dichotomies. Source: Modified with permission from Jongman et al. (1987)
Minimum-variance clustering and TWINSPAN are computationally complex and time consuming compared with clustering algorithms developed earlier. However, the widespread availability of powerful and inexpensive computers during the last 15 years has allowed them to become the most widely used clustering algorithms in ecological studies.
There are three primary process-oriented models of plant community structure (Austin 1986; Keddy and MacLellan 1990): the plant strategy model (Grime 1979), the gap dynamics/regeneration model (Grubb 1977; Pickett and White 1985), and the resource-ratio hypothesis (Tilman 1985, 1988). These models have not been applied to the management of plant communities, at least partially because they are not yet adequately integrated with descriptive approaches (Keddy and MacLellan 1990). Thus, they consider the effects of resource availability and disturbance on the interactions between plants, but they do not predict species composition in complex communities.
An additional model of community structure, the centrifugal organization model (Rosenzweig and Abramsky 1986; Keddy 1989, 1990), incorporates underlying mechanisms into descriptions of community structure along environmental gradients. Thus, this model integrates pattern with process. The centrifugal organization model represents an extension of the competitive hierarchy model described in the previous chapter. The model assumes that all species share a central habitat in which they exhibit maximum performance (e.g., growth, survival, reproductive output), but that each species has another (peripheral) habitat in which it is the best competitor. This model represents a variant of inclusive niche structure, in which species have overlapping fundamental niches along only one axis. Entire environmental gradients ("niche axes") may radiate outward from the central preferred habitat in the centrifugal organization model. Near the center (i.e., optimal habitat), species may have entirely inclusive fundamental niches; at or near the periphery, species' fundamental niches may include only a few adjacent neighboring species in the direction of the central habitat. In the latter case, negative interactions would be completely asymmetrical (i.e., interference), and removal experiments should show a species to increase nearer the central habitat, but not toward the periphery. Wetlands appear to be organized in this manner: a central habitat characterized by low disturbance and high fertility is dominated by large leafy species capable of forming dense canopies (e.g., Typha spp.), constraints such as
disturbance and fertility create radiating axes along which different groups of species and vegetation types are arrayed, and rare species occur only in peripheral habitats with low biomass (Moore and Keddy 1989; Moore et al. 1989; Keddy 1990) (Figure 3.16). Communities of forest trees (Keddy and MacLellan 1990) and desert rodents (Rosenzweig and Abramsky 1986) also exhibit distribution patterns consistent with the centrifugal organization model.
Was this article helpful?