62
4. Protocol for Aquatic Hazard Assessment
should be given high priority in hazard assessment (Garrett and Inman
1984; Sorensen et al. 1984; Woock and Summers 1984; Lemly 1985a,
1985b; Baumann and Gillespie 1986; Gillespie and Baumann 1986; Heinz
et al. 1987). Utilizing this and other information, some investigators have
developed mathematical aquatic toxicity models (ATMs) to describe and
predict selenium cycling and toxicity (eg, Bowie and Grieb 1991; Porcella
et al. 1991; Peterson and Nebeker 1992; Bowie 1995). However, ATMs
have a major drawback. They are complex and require considerably more
information than is available from a typical contaminant monitoring program. For example, speciation of waterborne selenium, rates of selenium
movement to and from sediments, biological uptake and excretion rates
may be required. Thus, the potential for models to be routinely used in
hazard assessment of selenium is quite limited. Another approach utilizes
the United States national water quality criterion for selenium to derive a
Hazard Quotient (EPA 1992). but this method is flawed because it uses
geometric mean waterborne concentrations and a criterion value that is
too high (see section on Evaluation, this chapter).
Despite the need to evaluate selenium hazards at many locations across
the United States (Peterson and Nebeker 1992; Presser et al. 1994; Lemly
1999; Seiler et al. 1999), and elsewhere (see Chapter 1), there is no common assessment method in use. This has resulted in confusion and frustration for those involved in hazard assessment of both site-specific and
regional selenium contamination problems (Sylvester et al. 1991; Presser
et al. 1994). I developed the selenium Protocol described in this chapter to
provide a simple, yet Scientifically credible, technique for conducting
aquatic hazard evaluations of this trace element. It includes the key parameters that are implicit in basic toxicological risk assessment, such as
concentration and exposure, but it also integrates biotic and abiotic
cycling components that are essential for site-specific hazard evaluation
of selenium. The method generates numerical scores that can be compared between years and across sites and locations, thereby providing a
consistent approach for evaluating hazard. The assessment focuses on
food-chain bioaccumulation and associated reproductive impairment
in fish and aquatic birds, which is the most sensitive biological endpoint for determining ecosystem-level hazards of selenium (Lemly 1993a,
1997a).
Definition of Aquatic Hazard
In the context of the Protocol, aquatic hazard is an expression of the
toxic threat to fish and aquatic birds that use a specific habitat known
or suspected of being contaminated with selenium. Hazard is characterized from two types of information: (1) the degree of selenium con-
4. Protocol for Aquatic Hazard Assessment
should be given high priority in hazard assessment (Garrett and Inman
1984; Sorensen et al. 1984; Woock and Summers 1984; Lemly 1985a,
1985b; Baumann and Gillespie 1986; Gillespie and Baumann 1986; Heinz
et al. 1987). Utilizing this and other information, some investigators have
developed mathematical aquatic toxicity models (ATMs) to describe and
predict selenium cycling and toxicity (eg, Bowie and Grieb 1991; Porcella
et al. 1991; Peterson and Nebeker 1992; Bowie 1995). However, ATMs
have a major drawback. They are complex and require considerably more
information than is available from a typical contaminant monitoring program. For example, speciation of waterborne selenium, rates of selenium
movement to and from sediments, biological uptake and excretion rates
may be required. Thus, the potential for models to be routinely used in
hazard assessment of selenium is quite limited. Another approach utilizes
the United States national water quality criterion for selenium to derive a
Hazard Quotient (EPA 1992). but this method is flawed because it uses
geometric mean waterborne concentrations and a criterion value that is
too high (see section on Evaluation, this chapter).
Despite the need to evaluate selenium hazards at many locations across
the United States (Peterson and Nebeker 1992; Presser et al. 1994; Lemly
1999; Seiler et al. 1999), and elsewhere (see Chapter 1), there is no common assessment method in use. This has resulted in confusion and frustration for those involved in hazard assessment of both site-specific and
regional selenium contamination problems (Sylvester et al. 1991; Presser
et al. 1994). I developed the selenium Protocol described in this chapter to
provide a simple, yet Scientifically credible, technique for conducting
aquatic hazard evaluations of this trace element. It includes the key parameters that are implicit in basic toxicological risk assessment, such as
concentration and exposure, but it also integrates biotic and abiotic
cycling components that are essential for site-specific hazard evaluation
of selenium. The method generates numerical scores that can be compared between years and across sites and locations, thereby providing a
consistent approach for evaluating hazard. The assessment focuses on
food-chain bioaccumulation and associated reproductive impairment
in fish and aquatic birds, which is the most sensitive biological endpoint for determining ecosystem-level hazards of selenium (Lemly 1993a,
1997a).
Definition of Aquatic Hazard
In the context of the Protocol, aquatic hazard is an expression of the
toxic threat to fish and aquatic birds that use a specific habitat known
or suspected of being contaminated with selenium. Hazard is characterized from two types of information: (1) the degree of selenium con-