Abstract

  Assessing the condition of an ecosystem to ascertain its 'health' presupposes tat we can diagnose pathological states in system measures. Frequently this means comparing current conditions to reference states, either historical or other sites, which also exhibit some natural range of variation. In the Fraser River, a 9th order river on the west coast of Canada, and one of the most productive salmon rivers in the world, we have studied assemblages of fish and benthos to assess ecosystem health. The biggest challenge to using species composition and abundance measures as indicators of system condition is the absence of appropriate reference conditions in many instances. There are few unperturbed rivers of large size in western North America, and indeed in much of the world, with which to compare the Fraser River or any other large river ecosystem. Multiple insults from point and non-point sources make it difficult to isolate factors from natural longitudinal changes in terms of their effects on river biota. Potential solutions include analysis of fragmentary historical data, making comparisons with other large rivers, and conducting extensive surveys within the basin to account for spatial gradients. An absolute diagnosis of ecosystem health of large rivers in natural science terms is unlikely, and otherwise will depend on relative changes through time assuming these can be isolated from natural variation and local effects. Definition of health for large, riverine ecosystems remains largely a case of expert opinion and weight of evidence rather than a testable hypothesis.

Keywords: biomonitoring, community structure, ecosystem health, fish, rivers