Rationale
115
of what the criterion is, there will likely be a concerted effort by some to
raise or lower it. USEPA policy allows the national criterion to be modified by states on a site-specific basis when biological evidence warrants it
(Stephan et al. 1985; EPA 1985). Such modifications have been pursued;
local criteria-below 5 Ilg SelL in some instances and above 5 Ilg SelL in
others-are being considered or have been adopted in several western
states, including Arizona, California, Colorado, and New Mexico (Arizona 1992; CEPA 1992; CSWRCBI987; Guglielmone 1995; New Mexico
1995). However, there is no widely accepted or published method for sitespecific criterion revision, developed by the USEPA or from other sources,
that is targeted specifically at selenium. Consequently, criteria are being
developed with little pertinent direction or guidance. This has caused
some to use insufficient data and make assumptions about fish and wildlife exposureleffects ·that could introduce significant error into the derivation process (Hamilton and Lemly 1999). This type of error has important
implications. A derivation might indicate that changing a criterion is appropriate when, in fact. raising the criterion is not environmentally acceptable or lowering the criterion is not biologically necessary. A
scientifically credible method is needed that can be broadly and uniformly
applied yet is sensitive to selenium's site-specific variations in environmental cycling and biological effects. I developed the procedure described
in this chapter to provide that method. Some examples are given to illustrate cases that justify raising or lowering an existing criterion, leaving a
criterion unmodified, or examining a proposed criterion.
Rationale
There are two primary considerations in developing site-specific water
quality criteria for selenium: (1) local conditions must be evaluated,
and the resultant information used to decide if a modified criterion is
necessary; and (2) if changes are warranted, the existing or proposed
criterion must be modified by an appropriate amount. A simple twostep process can be used to derive criteria for chronic exposure.
The major principle underlying this process is that reproductive effects in fish and aquatic birds are the most sensitive biological indicators of aquatic ecosystem-level impacts of selenium (Ohlendorf 1989;
Lemly 1993b, 1997; Skorupa et al. 1996). Selenium is passed from parents to their offspring in eggs, where it can kill developing embryos
outright or induce a variety of lethal or sublethal teratogenic deformities (eg, Hoffman et al. 1988; Lemly 1993b). However, parents can
consume a selenium-laden diet and experience partial or complete
reproductive failure without exhibiting symptoms of selenium toxicosis themselves (Lemly 1985a, 1997; Ohlendorf 1989). Moreover, aquatic
organisms that are the food of wildlife strongly bioaccumulate sele-
115
of what the criterion is, there will likely be a concerted effort by some to
raise or lower it. USEPA policy allows the national criterion to be modified by states on a site-specific basis when biological evidence warrants it
(Stephan et al. 1985; EPA 1985). Such modifications have been pursued;
local criteria-below 5 Ilg SelL in some instances and above 5 Ilg SelL in
others-are being considered or have been adopted in several western
states, including Arizona, California, Colorado, and New Mexico (Arizona 1992; CEPA 1992; CSWRCBI987; Guglielmone 1995; New Mexico
1995). However, there is no widely accepted or published method for sitespecific criterion revision, developed by the USEPA or from other sources,
that is targeted specifically at selenium. Consequently, criteria are being
developed with little pertinent direction or guidance. This has caused
some to use insufficient data and make assumptions about fish and wildlife exposureleffects ·that could introduce significant error into the derivation process (Hamilton and Lemly 1999). This type of error has important
implications. A derivation might indicate that changing a criterion is appropriate when, in fact. raising the criterion is not environmentally acceptable or lowering the criterion is not biologically necessary. A
scientifically credible method is needed that can be broadly and uniformly
applied yet is sensitive to selenium's site-specific variations in environmental cycling and biological effects. I developed the procedure described
in this chapter to provide that method. Some examples are given to illustrate cases that justify raising or lowering an existing criterion, leaving a
criterion unmodified, or examining a proposed criterion.
Rationale
There are two primary considerations in developing site-specific water
quality criteria for selenium: (1) local conditions must be evaluated,
and the resultant information used to decide if a modified criterion is
necessary; and (2) if changes are warranted, the existing or proposed
criterion must be modified by an appropriate amount. A simple twostep process can be used to derive criteria for chronic exposure.
The major principle underlying this process is that reproductive effects in fish and aquatic birds are the most sensitive biological indicators of aquatic ecosystem-level impacts of selenium (Ohlendorf 1989;
Lemly 1993b, 1997; Skorupa et al. 1996). Selenium is passed from parents to their offspring in eggs, where it can kill developing embryos
outright or induce a variety of lethal or sublethal teratogenic deformities (eg, Hoffman et al. 1988; Lemly 1993b). However, parents can
consume a selenium-laden diet and experience partial or complete
reproductive failure without exhibiting symptoms of selenium toxicosis themselves (Lemly 1985a, 1997; Ohlendorf 1989). Moreover, aquatic
organisms that are the food of wildlife strongly bioaccumulate sele-