5.1 A bioconcentration test is conducted to obtain information concerning the ability of an aquatic species to accumulate a test material directly from water. This guide provides guidance for designing bioconcentration tests on the properties of the test material so that each material is tested in a cost-effective manner.

5.2 Because steady-state is usually approached from the low side and the definition of apparent steady-state is based on a statistical hypothesis test, the apparent steady-state BCF will usually be lower than the steady-state BCF. With the variation and sample sizes commonly used in bioconcentration tests, the actual steady-state BCF will usually be no more than twice the apparent BCF.

5.3 When both are determined in the same test, the projected steady-state BCF will usually be higher than the apparent steady-state BCF because the models used to calculate the projected BCF assume that the BCF steadily increases until infinite time.

5.4 The BCFs and rates and extents of uptake and depuration will depend on temperature, water quality, the species and its size, physiological condition, age, and other factors (1).³ Although organisms are fed during tests, uptake by means of sorption onto food is probably negligible during tests.

5.5 Results of bioconcentration tests are used to predict concentrations likely to occur in aquatic organisms in field situations as a result of exposure under comparable conditions, except that mobile organisms might avoid exposure when possible. Under the experimental conditions, particulate matter is deliberately minimized compared to natural water systems. Exposure conditions for the tests may therefore not be comparable for an organic chemical that has a high octanol-water partition coefficient or for an inorganic chemical that sorbs substantially onto particulate matter. The amount of the test substance in solution is thereby reduced in both cases, and therefore the material is less available to many organisms. However, sorption might increase bioaccumulation by aquatic species that ingest particulate matter (2), or food may be a more important source of residues in fish than water per se for stable neutral organic chemicals that have a Log K _ow between 4 and 6 (3) .

5.6 Results of bioconcentration tests can be used to compare the propensity of different materials to be accumulated. Nonionizable organic chemicals can also be ranked for bioconcentration using correlations that have been reported between steady-state BCFs and physical-chemical properties, such as the octanol-water partition coefficient and solubility in water (4). However, when such predictions are impossible, exceed the demonstrated limits of the correlation, or might be otherwise questionable (1, 5), a bioconcentration test may be necessary.

5.7 Results of bioconcentration tests can also be used to compare the abilities of different species to accumulate materials. At steady-state the concentration of a nonionizable organic chemical in individual organisms, and in various tissues within an organism, will probably be related to the concentration of lipids in the organisms and tissues (6).

5.8 Results of bioconcentration tests might be an important consideration when assessing hazard (see Guide E1023) or deriving water-quality criteria because consumer animals might be adversely affected by ingesting aquatic organisms that contain toxic materials. However, assessment of hazard to consumer organisms must take into account not only the quantity of material accumulated in tissues of aquatic organisms, but also the toxicity of the material to the consumer. Further, humans eat only certain portions of most aquatic organisms, whereas other predators often consume additional tissues.

5.9 Bioconcentration tests might be useful for studying structure-activity relationships between test materials, biological availability, metabolism of materials in aquatic organisms, and effects of various environmental factors on results of such tests.

5.10 Uptake and depuration rate constants might be useful for predicting environmental fate using compartmental models (7).