and Clement 1992, Kelce et al. 1995). Even where evidence of effects exists, there are substantial difficulties in distinguishing these effects from natural variation due to the lack of baseline information. Given the rapid growth in synthesized chemicals being released into the environment, and the potentially profound impacts on ecological and human social systems if some of these materials were responsible for chronic effects that changed basic life history characteristics of the species involved, there is a substantial need for further research on these potential risks.
The effects of many other widespread air and water pollutants—such as nitrogen in water or sulfur dioxide in air—on ecological systems has received considerable study over the past two decades. Although site-specific predictions of the impacts of such pollutants are often uncertain, the general effects of those pollutants in different types of ecosystems can often be predicted with reasonable accuracy. As the experience with the Exxon Valdez oil spill in Prince William Sound in Alaska demonstrates, however, considerable uncertainty and inaccuracy can surround the evaluation of impacts of even widespread pollutants like oil. In this particular case, the widespread initial predictions of an ecological disaster proved far too pessimistic. While the spill may still have substantial long-term impacts on the ecological systems, it is unlikely that it will be possible to distinguish those impacts from natural fluctuations since the baseline of data in the region is not extensive.
One ecological generalization that has withstood the test of time is that the number of species found in a given habitat varies with the size of the habitat (see Figure 1). Consequently, when habitat is lost the number of species present will be reduced. While this species-area effect does allow a general assessment of the likely consequences of habitat change, it cannot provide detailed assessments of impacts on biodiversity. The actual impact of habitat conversion depends strongly on which areas are lost. Because species with restricted ranges are sometimes clustered in "hot spots" of diversity, protection of those hot spots could greatly reduce the effect of habitat loss on species diversity.
For a given species for which detailed demographic and life history information is available, we are able to make projections of the probability of survival of the species when its habitat is lost by use of Minimum Viable Population methodologies (Shaffer 1981). Insufficient demographic information is available for the vast majority of species, however, to allow such calculations.
The study of the impacts of fragmentation on ecological systems is still a relatively new field. Fragmentation can have significant effects both due to the increased ratio of edge habitat to core habitat (edge habitats have a different micro-climate and different species composition), and due to the access edge habitat may provide for certain species into core regions that would be otherwise inaccessible, as well as the barriers it may create for the dispersal and migration of