Methods for Deriving Pesticide Aquatic Life Criteria - Reviews of Environmental Contamination and Toxicology - page 66

model was used by the USEPA for deriving dissolved oxygen criteria for the Cape
Cod to Cape Hatteras region (USEPA 2000). The WCS notes that population
models are complex, and their application may be prohibitively resource intensive.
Nonetheless, such models constitute the best way of determining the significance
of effects on survival, growth, and reproduction that are typically measured in
laboratory toxicity tests. Ultimately, further literature searches for ecosystem
recovery studies may be the most practical way to determine appropriate excursion
frequencies.
In whatever format criteria are stated, monitoring programs must be designed to
include a compliance component. For criteria that are expressed as a single number
(ANZECC and ARMCANZ 2000; OECD 1995; CCME 1999; RIVM 2001;
Samsoe-Petersen and Pedersen 1995; Bro-Rasmussen 1994; Irmer et al. 1995;
Lepper 2002), the risk manager must determine how often and with what frequency
a criterion can be exceeded, and then design a monitoring program to assess
compliance. For criteria that include duration and frequency components (USEPA
1985; Roux et al. 1996; Zabel and Cole 1999), the risk manager has only to design
the monitoring program.
Exclusion of duration and frequency components from criteria statements leaves
those factors solely to policy-based decisions. It would be better if these compo-
nents could be science-based. Although the USEPA approach (1985) of expressing
criteria strengthens the science component, the duration and the frequency values,
used in acute and chronic criteria statements would benefit from a stronger scientific
basis. It is possible that a review of more recent literature could strengthen the duration
and the frequency values. The TTE models would give risk managers more science-
based information for determining the duration component; similarly, population
models and/or good ecosystem recovery studies would help in determining the
frequency component.
Aquatic life is exposed to contaminants by two routes: water and food. Water quality
criteria derived from single-species laboratory studies are based on water-only
exposures, which may considerably underestimate the actual environmental exposure
resulting from water and contaminated food sources (Benson et al. 2003). An extreme
example is demonstrated in a study of effects of selenium on fish (Lemly 1985). Loss
of diversity and reproductive failure occurred in fish communities exposed to sele-
nium at concentrations 10-35 times lower than concentrations causing adverse effects
in laboratory studies. Benson et al. (2003) noted that the extent of dose underestima-
tion caused by ignoring food exposure has not been well studied, because the
significance of the food pathway has only recently been recognized.
Studies of hydrophobic organic chemicals also show the importance of dietary
exposure. A model comparing food and water exposures of PCBs (polychlorinated
biphenyls) to lake trout in Lake Michigan, determined that 99% of body burdens
came from food exposure. Three-spine sticklebacks accumulated significantly more
hexachlorobenzene when feeding on contaminated Tubifex tubifex, compared to
water-only exposures (Egeler et al. 2001). Other studies have shown that the signifi-
cance of dietary uptake varies, but the underlying factors that determine food expo-
sure are still unclear. Relationships between log K ow values and dietary uptake of