Environmental Engineering Reference
In-Depth Information
The EU risk assessment TGD (ECB 2003) also provides instructions for how to
derive LC 50 /EC 50 or NOEC values from studies, in which those values are not reported:
- If raw data are available, the values can be directly calculated. The LC 50 /EC 50
should be calculated by probit analysis, or another regression technique; the
NOEC may be calculated using either the hypothesis test or regression approach
(the TGD does not claim a preference for one over the other, as a result of the
continuing controversy, previously discussed).
- Results presented as LC 10-49 /EC 10-49 can be used as LC 50 s, but if results are pre-
sented as LC > 50 /EC > 50 they cannot be used.
- An LOEC representing an effect more than 10% and less than 20% may be
converted to an NOEC: NOEC = LOEC/2.
- An LOEC representing an effect more than 20% is not used; an EC 10 is calcu-
lated from the data and is regarded as the NOEC.
- If the percent effect of a LOEC is unknown, then an NOEC cannot be estimated.
- An NOEC may be estimated as MATC/√2.
- An EC 10 from a long-term test is regarded to be an NOEC.
The EU TGD also describes data reduction procedures for cases in which there
are multiple toxicity values for one species (ECB 2003). Values are selected according
to which ones reflect realistic European environmental parameters. Also, the
database is evaluated to ensure that information will not be lost when averaging
procedures are used (e.g., for very sensitive endpoints). After these initial screening
steps, results for the most sensitive endpoints are selected. Multiple values for the
same endpoint for the same species are evaluated to determine why they are dif-
ferent. If values are determined to be equivalent, then the geometric mean is used.
If reasons exist for differences, then results may be grouped according to appropriate
factors (e.g., pH ranges). The effects of all possible data exclusions on the final
effects assessment must be explored and explained.
With the OECD methodology (1995), if several toxicity results are available
for the same species and endpoint, then the geometric mean of values is used.
If data are available for the same species, but for different endpoints (i.e., survival
vs growth vs reproduction), then only the lowest value is used. The OECD
guidelines (1995) require that only chronic NOEC or MATC values be used for
statistical extrapolation procedures. If only a chronic LOEC is reported, the
NOEC may be calculated as NOEC = LOEC/2, although this conversion is only
done if the LOEC corresponds to an effect less than 20%. If the measured effect
is greater than 20%, then further toxicity testing is required at lower concen-
trations. The factor of 2 is representative of the typical interval between test
concentrations; if the actual interval is known to be different, then it should be
used instead of 2.
For derivation of high reliability TVs, the Australia/New Zealand guidelines
(ANZECC and ARMCANZ 2000) require the following: if several NOECs are
available for different endpoints for the same species for a particular substance,
then the lowest NOEC (i.e., the most sensitive endpoint) is used for criteria deriva-
tion; if several NOECs are available for the same effect in the same species for a
Search WWH ::




Custom Search