An overview of river health assessment: philosophies, practice, problems and prognosis

1. Philosophically, the term 'river health' is useful because it is readily interpreted by the general public and evokes societal concern about human impacts on rivers. The common goal of achieving healthy rivers unites ecolo gists and the general public because the value of the ecologists' contribut ions is clear (and, hence, funded). The difficulty arises in the choice of relevant symptoms because there is a wide variety that can be measured with varying accuracy at a broad range of spatial scales. These indicators may respond to impacts at different time scales, and no single indicator is a ' silver bullet' that reveals river health unequivocally. 2. In practice, choice of indicator often shows personal bias, technical co nsiderations, and constraints of knowledge. Selection of appropriate spatia l and temporal scales for these measures is crucial. Although most measurem ents are spot samples (e.g. concentration, abundance, species richness), as sessment of river health based on changes in ecological processes such as p ost-disturbance recovery rate or nutrient spiralling lengths may be more su itable in some cases. 3. Problems include validation of the indicator, its response time at a ran ge of scales, and the reliability of its measurement.,Assessment of river h ealth should be accurate, timely (warning of deterioration instead of waiti ng until the patient is terminal), rapid (so that the response is swift), a nd inexpensive. The connectedness of running waters with their floodplains and catchments must be explicitly recognized. Hydrological and geomorpholog ical modifications of rivers usually affect their health by severing or imp airing the linkages, and the 'cure' may lie in addressing these causes. Oft en, we need landscape-level data for management because this is the scale w here cumulative effects of impacts are evident. 4. The prognosis is uncertain. We need to explore further the use of integr ative measures of river health, and focus on establishing a link between th e measure and impaired ecological integrity. Ecosystem-level variables (e.g . estimates of production or respiration) show promise and recent technolog ical advances make these more accessible. Data analytical approaches (e.g. multimetric vs. predictive models) need further debate but must not overloo k the importance of high quality and relevant input data. Appropriate choic e of indicators, rigorous sampling and analysis, and careful data interpret ation must be matched with effective communication to policy-makers and the public. When this occurs, the concept of 'river health' becomes more than just a rhetorical tool.'