Four standardized aquatic microcosm (SAM) experiments were performed at two laboratories to test the reproducibility of controls and copper sulfate treatments. Each laboratory (University of Washington and Duluth-EPA) conducted two 63-day experiments consisting of six replicates each of 0 (control), 500, 1000, and 2000 μg L−1 nominal copper (24 microcosms total). In controls, nitrate was rapidly converted to algal biomass and subsequently to Daphnia. Increasing amounts of copper delayed this conversion. At 500 μg L−1 Cu, algal blooms consistently occurred during the absence of Daphnia. Slightly different amounts of initial copper were related to the duration of these delays. Recovery of the microcosm communities was made possible by detoxification of copper through precipitation, chelation, and sorption, as predicted by the MINEQL copper speciation model. Statistical evaluation of results within each experiment was displayed by an “interval of nonsignificant difference” around the control means. A procedure to statistically compare experiments for reproducibility is being developed, based upon discriminant analysis of distance measures between control and treatment microcosms.
Results from this statistical procedure are presented for the four experiments on three variables (nitrate, Daphnia, algal biovolume). Statistical analyses support the hypothesis that these moderately complex ecological experiments are more similar within laboratories than between laboratories. In spite of this, the conclusions reached from both sets of experiments are similar.