User Guide to Thresholds and Classification - Environmental ...

317User Guide for Thresholds and ClassificationsWhere measured data are available for the start and end of test (as is normal for the acute Daphnia andalgal tests), the L(E)C50, for classification purposes, may be calculated based on the geometric mean ofthe start and end of test concentrations. Where the end of test concentrations are below the analyticaldetection limit, such concentrations shall be considered to be half that detection limit.where measured data are available at the start and end of media renewal periods (as may be availablefor the semi-static tests), the geometric mean for each renewal period should be calculated, and themean exposure over the whole exposure period calculated from these data.Where the toxicity can be attributed to a degradation breakdown product, and the concentrations of thisare known, the L(E)C50 for classification purposes, may be calculated based on the geometric mean ofthe degradation product concentration, back calculated to the parent substance.Similar principles may be applied to measured data in chronic toxicity testing.Poorly soluble substancesThese substances, usually taken to be those with a solubility in water of < 1 mg/L, are frequently difficult todissolve in the test media, and the dissolved concentrations will often prove difficult to measure at the lowconcentrations anticipated. For many substances, the true solubility in the test media will be unknown, andwill often be recorded as less than the detection limit in purified water. Nevertheless, such substances canshow toxicity, and where no toxicity is found, judgement must be applied to whether the result can beconsidered valid for classification. Judgement should err on the side of caution and should not underestimatethe hazard.Ideally, tests using appropriate dissolution techniques and with accurately measured concentrations withinthe range of water solubility should be used. Where such test data are available, they should be used inpreference to other data. It is normal, however, particularly when considering older data, to find suchsubstances with toxicity levels recorded in excess of the water solubility, or where the dissolved levels arebelow the detection limit of the analytical method. Thus, in both circumstances, it is not possible to verify theactual exposure concentrations using measured data. Where these are the only data available on which toclassify, some practical rules can be considered by way of general guidance.Where the acute toxicity is recorded at levels in excess of the water solubility, the L(E)C 50 forclassification purposes, may be considered to be equal to or below the measured water solubility. Insuch circumstances it is likely that a 9.1A classification should be applied. In making this decision, dueattention should be paid to the possibility that the excess undissolved substance may have given rise tophysical effects on the test organisms. Where this is considered the likely cause of the effects observed,the test should be considered as invalid for classification purposes.Where no acute toxicity is recorded at levels in excess of the water solubility, the L(E)C 50 forclassification purposes may be considered to be greater than the measured water solubility. In suchcircumstances, consideration should be given to whether the 9.1D classification should apply. In makinga decision that the substance shows no acute toxicity, due account should be taken of the techniquesused to achieve the maximum dissolved concentrations. Where these are not considered as adequate,the test should be considered as invalid for classification purposes.January 2012 EPA0109