a fact that supports on-site impressions of stream conditions in the two regions.
Functional integrity involves processes such as photosynthesis and community respiration, nutrient transfer, energy flow (secondary production), and decomposition. Abnormal rates of activity and accumulation or depletion of materials are indications of disruptions of an ecosystem's functional integrity. For example, during the course of a whole-lake acidification experiment, the entire nitrification process was halted at a pH of less than 5.7, possibly due to the loss of all nitrifying bacteria (Rudd et al., 1988). High rates of ecosystem metabolism commonly are associated with eutrophication; lowered or negative rates of metabolism may indicate active decomposition of excess supplies of organic matter trapped in the sediments. To distinguish abnormal conditions from normal ones, the natural range of variability must be known for each particular location (e.g., Frost et al., 1988).
Functional measures of ecosystem integrity are often considered less sensitive to environmental factors, including pollutants, than measures of structure. Also, system function may give a very different impression of the effects of environmental stress than structural responses, particularly species composition (Frost et al., 1995). However, these apparent anomalies simply may occur because functional measures operate at longer temporal or larger spatial scales than structural measures or because their responses are less well known. It often is assumed that ecosystems are functionally resilient to alteration of structure due to compensatory responses, but this assumption has not been tested adequately for aquatic ecosystems. A compensatory functional response occurs when rates and amounts of ecosystem processes remain unchanged in the face of changes in structural composition, such as alteration of species dominance or loss of species richness. Studies of compensation and complementarity in ecosystem function are few and limited largely to lakes (Schindler, 1987; Howarth, 1991; Frost et al., 1995). The loss of species from ecosystems produces inconsistent results. In some cases, such losses are accompanied by no apparent compensation (e.g., primary production remains unchanged); in others, considerable alteration of ecosystem processes occurs (e.g., Kitchell and Carpenter, 1993). Understanding the reasons for such apparent inconsistency is an important area for future research.
Measurement of ecosystem function has been avoided because methods dealing with it have been lacking or are thought to be more difficult and time-consuming to employ. Although it often is easier to evaluate aspects of system structure, this is not always the case. For example, it generally is easier to make a measurement of open-water community metabolism