The model presented above suggests how we can increase our ability to predict community change from a proposed or suspected human alteration, to a coastal system. As suggested, one primary source of uncertainty is ignorance of the relative importance of different functional groups and individual species within a group for a given system.
In a system in which diversity is changed through intentional exploitation (e.g. a fishery), uncertainty can be reduced relatively easily by using pilot harvests to gauge the community consequenccs of those actions. For instance, the effects of the proposed harvest of a predatory fish could be explored by allowing harvest at a few closely monitored sites. That is, uncertainty could be reduced by using the pilot programs as experiments (Holling 1978; Walters 1986). Such work must necessarily include controls and consideration of factors that may obscure important effects (such as immigration from unexploited areas). It would also be informative to perform the reverse experiment: small-scale refuges from harvest for a particular species to judge the current community effects of a fishery. Such work has yielded valuable insights into a system when the experiment is closely monitored (e.g. C'astilla and Duran 1985; Castilla and Bustamante 1989; see Rowley 1992 for a review of reserve effects).
For threats in which specific species are not intentionally manipulated, such as pollution, habitat destruction or global climate change, reducing uncertainty is more complicated. In such cases, it will be necessary to determine both the interaction category of an affected group of species (strong vs. weak, keystone vs. diffuse) and the relative susceptibility of species to the threat. The latter is necessary because in systems dominated by a single influential species the impact of human activities will be greatly modified by effects on the keystone species. As suggest earlier, approximate community dynamics can be determined by a combination of short-term experimentation, pattern quantification and natural history investigation. Determining the relative susceptibility of individual species to the threat can be approached through laboratory studies, transplant experiments to areas affected differently by a threat, and field tests simulating different levels of the threat.
Extrapolation from local pilot studies to potential large-scale manipulations must be approached cautiously, however, because an understanding at a local scale docs not necessarily translate directly to regional scales (Ricklefs 1987; Quinn et al. 1993; Carr and Reed 1993). Nevertheless, insights gained from such investigations will be invaluable for managers, policy makers and researchers in further attempts to understand and reduce the consequences of diversity loss.
Was this article helpful?