Preliminary Results of Interlaboratory Testing of a Standardized Aquatic Microcosm

Abstract

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.

Factors Affecting the Growth of Pseudokirchneriella subcapitata in Single‐Species Tests: Lessons for the Experimental Design and the Reproducibility of a Multitrophic Laboratory Microcosm

“…Considerable effort was invested in standardizing and studying system behavior prior to conducting exposure experiments, but different timings and magnitudes of biological responses were measured among experimental controls in the host laboratory and 3 independent laboratories. Although the ecological experiments in the SAM were statistically closer within laboratories than between laboratories, similar enough patterns emerged to draw the same conclusions across laboratories (Taub et al 1986;Taub 1993). Whereas variations in system dynamics were probably caused by differing rearing histories of the grazers between laboratories in the SAM (Matthews et al 1996), in the TriCosm the lack of repeatability was medium related.…”

Problems associated with selecting the most sensitive species for toxicity testing

“…Reports indicate that many more complex tests are no more expensive than traditional tests, with costs ranging from $6,000 to $29,000 for microcosm tests and $7,500 to $29,000 for conventional tests (e.g., Van Voris et al, 1985;Perez & Morrison, 1985;Niederlehner et al, 1986;Taub et al, 1986;Sheehan et al, 1986). Micro-or mesocosm tests may be more efficient and less expensive ways of determining effects on some characteristics because they simultaneously expose many organisms (e.g., Niederlehner et al, 1986).…”

Predicting the Ecological Effects of Herbicides

“…Ecologists have recognized patterns in the responses of ecosystems to stress (Odum 1985, Schindler 1987, Schaeffer et al 1988. Many have suggested that ecological variables are amenable to controlled testing, and several well-tested procedures exist for assessing chemical effects on ecologically meaningful variables in microcosms (e.g., Giddings 1984, Shannon et al 1984, Taub et al 1984, Stay et al 1985. Some of these procedures are recommended protocols under TSCA, although microcosm testing has rarely been a factor in chemical assessments (Nabholz 1997).…”