Incorporation of 28-d Leptocheirus plumulosus toxicity data in a sediment weight-of-evidence framework

A series of side-by-side trials were conducted to evaluate the variability of 28-d Leptocheirus plumulosus amphipod toxicity test data using existing and modified test protocols. One modification included examination of the influence of press-sieving on the sediment chemistry and the toxicity data. Press-sieving sediment did not reduce the variability in the toxicity data and also contributed uncertainty to the chemistry data. The second modification involved determining the sex of surviving adult amphipods so that the reproduction data could be measured as offspring/surviving female instead of only as offspring/surviving amphipod. Normalizing reproductive output to the number of adult females was ineffective in reducing the variability. The data from sediment toxicity tests are often interpreted in the context of 20% reductions and/or statistically significant reductions relative to negative controls. High inter-replicate variability makes default application of these decision criteria to the 28-d L. plumulosus toxicity test inappropriate regardless of whether or not samples are press-sieved or the sex of surviving amphipods is determined. This is not to say that the 28-d L. plumulosus toxicity test has no value for a sediment WOE: it provides long-term chronic data that may not otherwise be available. However, test-specific decision criteria must be established as part of the problem formulation based on the overall management goal, the availability of other lines of evidence (other toxicity tests as well as other types of data) and the desired level of certainty with respect to decision-making.