The case for causal inference analysis in environmental assessment: Making use of the literature

Rivers are often exposed to a range of stressors, both natural and human, which can impair the health and fitness of aquatic biota.  These stressors may act independently, antagonistically, or synergistically and biotic responses are the integrated outcomes.  Management decisions regarding stressor and ecological effects can be made confidently when the quality of the information on which they are based is understood.  A robust (e.g. BACI design) local study may answer relevant questions of cause and effect, however, such studies are often not possible because the putative causes have already occurred and treatments usually cannot be replicated.  Also, the results of local studies are unlikely to be generally applicable, would need to be undertaken at each new site and would not improve the confidence in reaching a similar conclusion at other locations.  If powerful studies are not possible and results from them not generally applicable, more types of evidence are needed.  Methods for testing causal relationships using the literature and considering the quality of individual studies have been widely accepted in epidemiology for more than 50 yrs.  This paper demonstrates the application of causal inference using published literature for an environmental problem of fine sediment addition to rivers.