Methods for Deriving Pesticide Aquatic Life Criteria - Reviews of Environmental Contamination and Toxicology - page 75

For example, if no measured ACRs are available, then three assumed ACRs (of 18)
are used. If two measured values are available, then just one assumed value is used.
The geometric mean of the three values is then used as the SACR, which is used to
calculate the SCV, just as the ACR is used to calculate and FCV. For the AF
method, employed by the state of North Carolina, a factor of 3 is applied to the
lowest available LC 50 value. Factors ranging from 20 to 100 are used as default
ACRs. No justification for these factors is given (North Carolina Department of
Environment and Natural Resources 2003). Lillebo et al. (1988) use an additive
toxicity model to derive the suggested factor of 0.1; this factor is applied to the
geometric mean of the three lowest LOECs among acceptable studies. Because this
value was derived from metals effects data, the applicability to pesticides may or
may not be valid.
The factor of 10, used in deriving TVs from multispecies data in Australia/New
Zealand, is to account for variations in mesocosm types, and for the fact that more
sensitive species may not have been in the test systems (ANZECC and ARMCANZ
2000). No particular justification is given for factors used to derive moderate and
low reliability TVs; however, they are similar to those provided in the OECD guide-
lines (OECD 1995), on which much of the Australia/New Zealand methodology is
based. Acute-to-chronic conversions are accomplished, in the Australia/New
Zealand guidelines, in one of three ways: a chemical-specific ACR is applied; an
LC 0 is calculated according to Mayer et al. (1994) and Sun et al. (1995); or a default
ACR of 10 or more is applied. The chemical-specific ACR is the ratio of an acute
EC 50 to a chronic NOEC. If multiple ACRs are available, the geometric mean of all
ACRs for all species is used for derivation of criteria by the SSD method, while
the ACR for the most sensitive organism is used for the AF method.
The factors used in the EU guidance (Bro-Rasmussen et al. 1994) range from 10
(to account for experimental variability), to 100 (to account for lack of NOEC data),
to 1000 (to account for lack of NOEC and LC 50 data). Although not specifically
stated, the discussion in Bro-Rasmussen et al. (1994) suggests that the final EU
factor could be adjusted, if judged necessary because of bioaccumulative potential,
persistence, carcinogenicity, mutagenicity, or another concern. The French and the
Spanish methodologies utilize the EU factors, although Spain includes the possibility
of a factor as high as 100,000, if only acute data are available, if ecotoxicity data
for relevant species is lacking, a substance is persistent or bioaccumulative, and a
substance has genotoxic potential (Lepper 2002). The EU risk assessment TGD
(ECB 2003) uses AFs ranging from 1 to 1000; factor size depends to a large extent
on professional judgment. The factors are intended to account for variability of
laboratory toxicity data, variability within and between species, short to long-term
exposure extrapolation, and laboratory to field extrapolation (which includes
effects of mixtures). If more toxicity data are available for species of different
trophic levels, different taxonomic groups, and different lifestyles, smaller factors
are applied. If only one acute value is available from each of three trophic levels, a
factor of 1000 is applied. If only a single chronic NOEC is available from either
a fish or daphnid test, a factor of 100 is applied. If two long-term NOECs from two
different trophic levels exist, the factor is 50. A factor of 10 is used if chronic