forests as species refugia, and the role of natural disturbances in maintaining regional species richness. At a regional level, one also has to consider the importance of exotic species in the maintenance of species richness, particularly in ecosystems subjected to the impact of human activity. This approach seeks balance by considering factors that maintain species richness as well as those that decrease it. Considerable research is required to provide sound estimates based on this approach, because critical data concerning ecosystem function are not available in enough breadth to support enlightened management or policy making.

Most calculations of species extinction rates emphasize the negative aspects of the problem, and this can have beneficial effects in terms of public awareness of environmental issues. I call attention to the positive terms of this issue, using examples from the Caribbean. These examples must be used with caution, because natural conditions in the Caribbean (particularly the frequency of hurricanes) select for resilient ecosystem, and it could be argued that this selective force invalidates the examples given. However, human impacts have been so intense in the Caribbean that the region remains as a test case for theories that emphasize island fragility. And besides, the essence of my argument is that in the development of any prediction involving biotic phenomena (whether it is species extinction, global carbon cycle, or acid rain effects), it is necessary to include the plethora of checks and balances that typify ecosystem function. In the Caribbean example, ecosystems must cope with hurricanes and intensive human-induced disturbances, whereas elsewhere, periodic fire, earthquakes, frost, or landslides may play the natural role of ecosystem stressor.