Environmental Engineering Reference
In-Depth Information
concentrations and a control, and include two replicates per concentration. In addition,
the concentration of the compound should be measured periodically throughout the
study and physical-chemical parameters should be monitored [pH, temperature,
hardness, TSS (total suspended solids), etc.]. Test endpoints should include bio-
mass and population density, as well as species diversity and species richness.
Although field and semi-field data are not used in criteria derivation, the OECD
guidelines (1995) offer criteria for assessing the acceptability of ecosystem studies;
such studies are useful for assessing effects of chemicals under field conditions. To
be acceptable, test results must include an NOEC for key components of the
ecosystem and must show a concentration-response relationship. To avoid overpre-
diction of toxicity, the test should include an ecosystem recovery component.
Test systems should include a range of taxonomic groups, preferably including
fish, and must have properly simulated ecosystem properties such as nutrient cycling
and trophic structure. Physical and chemical parameters (pH, dissolved oxygen,
hardness, and temperature) must be monitored throughout the test. Biological
response measurements should include individual level parameters (survival,
growth, reproduction, and bioaccumulation) as well as population (age/size structure,
production, and recover rates) and community level (species composition and rela-
tive abundance) measurements. Tests must be conducted at time and space scales
that are appropriate to the physical-chemical characteristics of the toxicant and life
history of organisms. Ecosystem studies must include a control and —two to three
test concentrations and should be duplicated.
Because of the challenges in conducting and interpreting multispecies toxicity tests,
and the relative cost-effectiveness, reproducibility and reliability of single-species
tests, most methodologies do not utilize multispecies data for criteria derivation.
However, Australia/New Zealand, Germany, the UK, and the EU (in risk assessment
TGD) do have provisions for using field or microcosm data to derive criteria, providing
it meets acceptability criteria (ANZECC and ARMCANZ 2000; Irmer et al. 1995;
Zabel and Cole 1999; ECB 2003). In practice, very few criteria are derived from field
data. Methodologies that do not use field or semi-field data directly do use them as a
comparison to criteria derived from single-species data (RIVM 2001; OECD 1995). In
some cases, a final criterion may be adjusted if strong multispecies evidence indicates
that the single-species criterion is over- or under-protective (USEPA 1985, 2003a;
Zabel and Cole 1999; RIVM 2001).
Survival, growth, and reproduction are traditional measurement endpoints in eco-
toxicity tests. Because these effects can be readily linked to population-level effects,
they are favored for use in deriving water quality criteria that are to be protective of
ecosystems. Nontraditional endpoints, such as endocrine disruption, enzyme induction,
enzyme inhibition, behavioral effects, histological effects, stress protein induc-
tion, changes in RNA or DNA levels, mutagenicity, and carcinogenicity, are often
more sensitive than traditional endpoints, but have had very few links established
between effects seen at the individual versus population, community, or ecosystem
levels. For this reason, they are rarely used for derivation of water quality criteria.
Using the USEPA methodology (1985), nontraditional endpoints fall into the
category of “other data,” and are rarely used in criteria derivation. The recent “Ambient
Search WWH ::




Custom Search