Procedure
121
als are long-time residents reflecting local conditions of selenium exposure in the HU. Because that may not be the case, faunal surveys
alone are not adequate to determine the presence or absence of selenium toxicity. Actual reproductive assessment based on tissue selenium residues and occurrence of terata or mortality in fish larvaelfry
and bird embryos (sensitive species) must be used to draw a conclusion of effect/no effect (see Chapters 2 and 5, and Skorupa and
Ohlendorf, 1991, for details and guidelines on reproductive assessment).
The absence of selenium-sensitive species in the HU is not a legitimate
justification for raising the criterion, because their absence could be
due to selenium poisoning. A comparison of species present in neighboring HUs, including supporting tissue selenium residue data, is necessary to evaluate whether sensitive species are likely absent due to
selenium contamination or some other environmental attribute (eg,
physical habitat quantitylquality, unique or additional competing species, other anthropogenic disturbances). A no-effect determination
should be made for at least one selenium-sensitive species in the HU
before raising a criterion is proposed. If this procedure is followed, and
a clean bill of health can be given to the entire HU, then the criterion
should be raised by 50%. In summary, criteria can be left unmodified,
adjusted upward by a fixed amount (50%), or adjusted downward by
one of 3 amounts (25%, 50%, or 75%).
Modified criterion values should be rounded to the nearest 0.5 !lg
SelL. A criterion that falls equidistant from 2 values is rounded down if
the criterion is being lowered, and up if the criterion is being raised.
For example, a criterion of 5 !lg SelL that is lowered by 25% (1.25 !lg
SelL) becomes 3.5 !lg SelL (5.0 - 1.25 = 3.75, rounded down to 3.5).
Lowering the same criterion by 50% yields a value of 2.5 !lg SelL, and
lowering it by 75% results in a final criterion of 1.0 !lg SelL (1.25
rounded down to 1.0). Criteria that are raised are modified by adding
50% of the criterion value and then rounding up, if necessary, to determine the final value. In theory, criteria could be set at smaller intervals, for example, 0.1 !lg SelL, but it would likely be impossible to
scientifically justify or reliably monitor and enforce such criteria. For
example, the range of bioaccumulation and effects displayed among
fish and aquatic birds (even among the most sensitive species) would
offset the perceived benefits of setting criteria at 2.2 and 2.6 !lg SelL
rather than 2.0 and 2.5 !lg SelL. Rounding to 0.5 !lg SelL intervals
should accomplish the goal of criterion revision-that is, to achieve
the desired reduction or prevention of selenium bioaccumulation/effects in aquatic ecosystems-yet also be practical for monitoring, compliance, and treatmentlremediation considerations. Diagnostic selenium
residues and examples of measured concentrationsleffects and appropriate modifications to criteria are given in Table 7.1.
121
als are long-time residents reflecting local conditions of selenium exposure in the HU. Because that may not be the case, faunal surveys
alone are not adequate to determine the presence or absence of selenium toxicity. Actual reproductive assessment based on tissue selenium residues and occurrence of terata or mortality in fish larvaelfry
and bird embryos (sensitive species) must be used to draw a conclusion of effect/no effect (see Chapters 2 and 5, and Skorupa and
Ohlendorf, 1991, for details and guidelines on reproductive assessment).
The absence of selenium-sensitive species in the HU is not a legitimate
justification for raising the criterion, because their absence could be
due to selenium poisoning. A comparison of species present in neighboring HUs, including supporting tissue selenium residue data, is necessary to evaluate whether sensitive species are likely absent due to
selenium contamination or some other environmental attribute (eg,
physical habitat quantitylquality, unique or additional competing species, other anthropogenic disturbances). A no-effect determination
should be made for at least one selenium-sensitive species in the HU
before raising a criterion is proposed. If this procedure is followed, and
a clean bill of health can be given to the entire HU, then the criterion
should be raised by 50%. In summary, criteria can be left unmodified,
adjusted upward by a fixed amount (50%), or adjusted downward by
one of 3 amounts (25%, 50%, or 75%).
Modified criterion values should be rounded to the nearest 0.5 !lg
SelL. A criterion that falls equidistant from 2 values is rounded down if
the criterion is being lowered, and up if the criterion is being raised.
For example, a criterion of 5 !lg SelL that is lowered by 25% (1.25 !lg
SelL) becomes 3.5 !lg SelL (5.0 - 1.25 = 3.75, rounded down to 3.5).
Lowering the same criterion by 50% yields a value of 2.5 !lg SelL, and
lowering it by 75% results in a final criterion of 1.0 !lg SelL (1.25
rounded down to 1.0). Criteria that are raised are modified by adding
50% of the criterion value and then rounding up, if necessary, to determine the final value. In theory, criteria could be set at smaller intervals, for example, 0.1 !lg SelL, but it would likely be impossible to
scientifically justify or reliably monitor and enforce such criteria. For
example, the range of bioaccumulation and effects displayed among
fish and aquatic birds (even among the most sensitive species) would
offset the perceived benefits of setting criteria at 2.2 and 2.6 !lg SelL
rather than 2.0 and 2.5 !lg SelL. Rounding to 0.5 !lg SelL intervals
should accomplish the goal of criterion revision-that is, to achieve
the desired reduction or prevention of selenium bioaccumulation/effects in aquatic ecosystems-yet also be practical for monitoring, compliance, and treatmentlremediation considerations. Diagnostic selenium
residues and examples of measured concentrationsleffects and appropriate modifications to criteria are given in Table 7.1.