Environmental Engineering Reference
In-Depth Information
topics and over a hundred guidance documents for perform-
ing ecological risk assessment, indicating the complexity of
such assessments (Bartell et al., 1992; Newman, 1998; Fisher
and Burton, 2004; Sorensen et al., 2004; Suter et al., 2005;
Barnthouse et al. 2007). This proliferation is due in part to
the inherent complexity of ecosystems, to the wide range of
species present, and to the diversity of forms (e.g. aquatic,
terrestrial), stages (e.g. egg, larvae, young, adult), phases
(e.g. moving from aquatic to terrestrial), and life spans (e.g.
hours to decades).
Because of ecologic complexity, ERA must be conducted
with a particular objective in mind. Merely examining the
risk from mercury in the environment is too broad, and
requires considerable refi nement (the problem- formulation
phase). Particular care should be devoted to the inclusion
of a wide range of stakeholders at this point to ensure that
the fi nal assessment addresses the concerns of interested
and affected parties, the general public, and the manage-
ment and regulatory needs of state, federal and tribal gov-
ernments. While risk assessment is basically scientifi c and
technical, determining the ecologic end point of concern is
mainly social and political (a case in point is that of Pacifi c
Salmon in the Northwest [Lackey, 1996; NRC, 1996]).
Formal risk assessment for mercury in ecosystems follows
the general methods described above, including problem
formulation, characterization of exposure and pathways,
characterization of effects, and risk characterization (refer
to Figure 12.2). For an ecologic evaluation of mercury, the
question to be addressed must be carefully defi ned. Further,
the difference between assessment end points and measure-
ment end points must be clarifi ed. Uncertainties exist for
ERA, particularly for gaps in available data on effects for dif-
ferent species, and in the magnitude of chemical exposures
(Washburn et al., 1998). Some of the uncertainty is reduced
by using probabilistic rather than deterministic risk assess-
ment for the effects of methylmercury, particularly for con-
sumption exposures in humans (Johnston and Snow, 2007).
As an example, probabilistic risk assessments for the
effects of mercury on different species in the Everglades of
South Florida relied on literature-derived life history param-
eters, combined with site-specifi c concentrations of mer-
cury in species (Duvall and Barron, 2000; Rumbold, 2005).
This assessment indicated that alligators had 100% exceed-
ances of long-term risk thresholds, and Great Egrets (
Egretta
alba
) had 99% exceedances. Mercury is one of the issues
of concern for the restoration of the Everglades (SFWMD,
2007, 2010). Aquatic ecosystems, such as the Everglades, are
particularly vulnerable to mercury and other contaminants
because of the potential for bio-accumulation and rapid
transport through aquatic systems (Burger, 1997b).
Similarly, probabilistic risk models have been used to
examine mercury contamination in the East Fork Poplar
Creek at the Department of Energy's Oak Ridge Site (Moore
et al., 1999). Over 50 years of operation at the site has resulted
in contamination of water, sediment, biota, and fl oodplain
soils. Monte Carlo simulations of total daily intake of mercury
by species was integrated with species-specifi c dose-response
curves to estimate risk. Methylmercury posed a risk to mink
(24% probability of at least 15% mortality) and to kingfi sh-
ers (50% probability of at least a 12-28% decline in fecundity
(Moore et al., 1999). These estimates allow for a comparison
of the relative risk between the two species, although it would
be more useful if the same end points had been used for both
species (e.g. either mortality or fecundity).
Exposure Assessment
Exposure evaluations for mercury follow directly from the
problem-formulation phase. In most cases, exposure assess-
ment involves determining the levels of mercury in whole
bodies (usually used for invertebrates and plants) or in tis-
sues, such as liver, kidney, muscle, or blood. Noninvasive
techniques for mercury evaluation include examining the
feathers for birds, hair for mammals (including humans),
and tail tips for reptiles. Levels of mercury in tissues are
usually called “biomarkers” of exposure, and optimally
they also indicate something about the possible effects of
mercury exposure. For example, levels of mercury in blood
are usually signifi cantly correlated with fi sh consumption
(Harnly et al., 1997). Once the concentration has been
determined, the opportunity for contact and the amount
of contaminated prey consumed must be estimated.
Benchmarks
As with human health risk evaluations, an important fi rst
step in evaluating a site is to screen for priority chemicals.
For aquatic systems, if the levels of a chemical are below
background concentrations, they can be ignored, and if
the chemical analysis yielded nondetectable levels and the
methods were deemed appropriate, the chemical may be
ignored. If the chemical concentrations are below concen-
trations that have been determined to constitute an ecotox-
icologic hazard, they may be ignored. Other chemicals also
need to be evaluated. The challenge is that there are very
few chemicals for which the ecotoxicologic adverse effect
level has been adequately demonstrated in any organism.
There are screening level benchmarks, especially for
chemicals in aquatic systems that are generally accepted. In
the United States, the U.S National Ambient Water Quality
Criteria for Protection of Aquatic Life (NAWQC) are used to
screen aqueous chemicals. However, there are few such cri-
teria involving very few species, making it useful to develop
other benchmarks with the use of the LOELs (lowest observed
adverse effect level) and NOELs; there is no NAWQC for mer-
cury. However, the lowest long-term effect values for meth-
ylmercury have been determined to be 0.52 µg/L for fi sh,
0.04 µg/L for Daphnids, and 0.8-4.0 µg/L for plants (Suter,
1996). Values are also given for inorganic mercury, which are
lower for fi sh, and higher for Daphnids. In general, bench-
marks are more highly developed for aquatic systems, mainly
because of the rapid transport of contaminants in aquatic
