aquatic ecosystems, but the effects of each type of disturbance may be synergistic among types and cumulative in space and/or time (Sidle, 1990). Although viewed as relatively local, they often have large-scale, far-reaching effects. Some large-scale stresses affecting aquatic ecosystems, whether natural or human induced, are rapid and dramatic. Examples include certain recent cases of massive deforestation, urbanization, development of crop- and pasturelands, forest fires, plant disease outbreaks, and insect infestations. Other disturbances occur over extended periods of time and, hence, often are not recognized as such until the situation becomes extremely difficult or impossible to reverse. These include acidification, some types of logging and mining, livestock grazing, fire suppression, irrigation, and potentially, global climate change (Minshall, 1992, 1993; Covich, 1993). Global climate change could profoundly alter riparian ecosystems through its effect on terrestrial vegetation, thermal and hydrologic regimes, nutrient cycles, and so on (Firth and Fisher, 1992). Fast or slow, disturbances of riparian ecosystems may result in changes in water temperature or runoff, channel straightening, scouring or sedimentation, loss of physical habitat, alteration of food base, and waterlogging or drying of riparian soils.
Although legislation calls for maintaining biological integrity, measuring the biological health of inland waters is extremely complex; this complexity results not only from the need to account for natural variations in time and space, but also from the need to consider individual species as well as interactions among organisms in a particular aquatic community.
There is no single correct scale for the study, assessment, or management of aquatic-riparian ecosystems (O'Neill et al., 1986; Levin, 1992; Johnson et al., 1993); rather, the appropriate scale depends on the scientific question or management problem being addressed. The importance of various environmental factors and the interpretation of measurements taken on aquatic ecosystems vary with scale (O'Neill et al., 1986; Minshall, 1988). Further, since ecosystem boundaries vary with scale, the spatial boundaries also must be correlated with the temporal framework appropriate for a particular disturbance (O'Neill et al., 1986).
Most ecosystems extend over comparatively large areas and persist for long periods of time. It is thus difficult to devise large-scale, single-value measurements of ecosystem integrity. However, the hierarchical structure of ecosystems results in a series of scaled interactions that can act as natural integrators of local processes. For example, measurement of community metabolism of a river segment or lake can serve to integrate the status of