76
4. Protocol for Aquatic Hazard Assessment
A basic flaw with the HQ method is that it uses geometric mean
waterborne concentrations of selenium. A large part of the selenium
bioaccumulation and associated toxic impact to fish and wildlife in an
aquatic system can result from one-time or periodic waterborne inputs
that have high concentrations relative to the long-term mean (Garrett
and Inman 1984; Sorensen 1988; Presser et al. 1994). On a mass-balance basis, these spikes in concentration deliver a disproportionately
large amount of selenium that can load sediments and be cycled into
detrital food chains long after the waterborne spike is gone (Lemly
and Smith 1987; Lemly 1997b). Once in the aquatic system, this "pulse"
of selenium may be all that is necessary to rapidly escalate
bioaccumulation and cause a cascade of impacts. The toxicity threshold for selenium is very steep. An increase of only a few Ilg SelL,
which may represent the difference between mean and maximum
concentrations in HQ analysis, can cause a totally different, much
more serious impact on fish and aquatic birds (Lemly 1985b, 1993a;
Skorupa and Ohlendorf 1991). It is critical to capture this underlying
principle of selenium ecotoxicology when conducting hazard analyses
and risk assessments. Relying on mean waterborne concentrations diminishes and masks the importance of the peaks in terms of selenium
loading, bioaccumulation, and toxic threat. Results are biased, and
hazard is seriously underestimated. This could lead to risk management decisions that would not protect fish and wildlife from selenium
toxicity. In order to make the best decisions, resource managers need
to know what the maximum hazard is, not the mean hazard. Thus, it
is essential to estimate hazard using the highest concentrations of
selenium measured in the samples. The USEPA's HQ approach does
not do this.
Hazard profiles used in the Protocol method integrate updated information on selenium bioaccumulation and reproductive toxicity to fish
and wildlife. Hazard is characterized in terms of the potential for foodchain bioaccumulation and reproductive impairment in fish and aquatic
birds, which are the most sensitive biological responses for estimating
ecosystem-level impacts of selenium contamination (Lemly 1993a).
Hazard estimates generated from the Protocol are not driven by a single
concentration value or water criterion for effect/no effect evaluations,
as is the case for the HQ method. There is no implicit assumption or
evaluation of toxicity related to direct waterborne exposure. Rather,
water concentrations are used to estimate hazard associated with
bioaccumulation in the planktonic food chain. Each of the ecosystem
components used in the Protocol is given equal weight; thus, spatial or
temporal variation in the relative dominance of one selenium cycling
pathway will be accounted for in the assessment. This is a key point
that differentiates the two methods. The Protocol utilizes multiple parameters that determine selenium bioaccumulation and exposure on a
4. Protocol for Aquatic Hazard Assessment
A basic flaw with the HQ method is that it uses geometric mean
waterborne concentrations of selenium. A large part of the selenium
bioaccumulation and associated toxic impact to fish and wildlife in an
aquatic system can result from one-time or periodic waterborne inputs
that have high concentrations relative to the long-term mean (Garrett
and Inman 1984; Sorensen 1988; Presser et al. 1994). On a mass-balance basis, these spikes in concentration deliver a disproportionately
large amount of selenium that can load sediments and be cycled into
detrital food chains long after the waterborne spike is gone (Lemly
and Smith 1987; Lemly 1997b). Once in the aquatic system, this "pulse"
of selenium may be all that is necessary to rapidly escalate
bioaccumulation and cause a cascade of impacts. The toxicity threshold for selenium is very steep. An increase of only a few Ilg SelL,
which may represent the difference between mean and maximum
concentrations in HQ analysis, can cause a totally different, much
more serious impact on fish and aquatic birds (Lemly 1985b, 1993a;
Skorupa and Ohlendorf 1991). It is critical to capture this underlying
principle of selenium ecotoxicology when conducting hazard analyses
and risk assessments. Relying on mean waterborne concentrations diminishes and masks the importance of the peaks in terms of selenium
loading, bioaccumulation, and toxic threat. Results are biased, and
hazard is seriously underestimated. This could lead to risk management decisions that would not protect fish and wildlife from selenium
toxicity. In order to make the best decisions, resource managers need
to know what the maximum hazard is, not the mean hazard. Thus, it
is essential to estimate hazard using the highest concentrations of
selenium measured in the samples. The USEPA's HQ approach does
not do this.
Hazard profiles used in the Protocol method integrate updated information on selenium bioaccumulation and reproductive toxicity to fish
and wildlife. Hazard is characterized in terms of the potential for foodchain bioaccumulation and reproductive impairment in fish and aquatic
birds, which are the most sensitive biological responses for estimating
ecosystem-level impacts of selenium contamination (Lemly 1993a).
Hazard estimates generated from the Protocol are not driven by a single
concentration value or water criterion for effect/no effect evaluations,
as is the case for the HQ method. There is no implicit assumption or
evaluation of toxicity related to direct waterborne exposure. Rather,
water concentrations are used to estimate hazard associated with
bioaccumulation in the planktonic food chain. Each of the ecosystem
components used in the Protocol is given equal weight; thus, spatial or
temporal variation in the relative dominance of one selenium cycling
pathway will be accounted for in the assessment. This is a key point
that differentiates the two methods. The Protocol utilizes multiple parameters that determine selenium bioaccumulation and exposure on a