Methods for Deriving Pesticide Aquatic Life Criteria - Reviews of Environmental Contamination and Toxicology - page 65

The maximum acceptable concentration would be a concentration not to be
exceeded any time, and is intended to protect against episodic exposure events.
Within Canada, British Columbia (BC) has its own criteria derivation methodology
(Government of British Columbia 1995), which closely resembles that of the CCME
(1999), except that the BC guidelines recommend derivation of separate acute and
chronic criteria for substances that are known to be acutely toxic. Similarly, South
Africa utilizes a modified USEPA methodology (1985), in which final criteria are
stated as either acute effect values (AEV) or chronic effect values (CEV; Roux et al.
1996). Thus, both BC and South Africa criteria address the role of exposure duration
in toxicity, although they do not address frequency.
Analysis of ecotoxicity data by time to event (TTE) methods allows simultaneous
consideration of exposure magnitude and duration when making effects predictions
(Newman and Crane 2002). Among other things, the TTE models may be used for
(1) estimates of effects over any time period, rather than just at the end of an arbi-
trary test period; (2) extrapolation from acute-to-chronic exposures; (3) analysis of
time-varying exposure (e.g., pulse exposures); and (4) determination of changes in
relative risk over time (Crane et al. 2002).
Among current criteria derivation methodologies, only the Australia/New
Zealand guidelines (ANZECC and ARMCANZ 2000) allow the use of the TTE
methods of Mayer et al. (1994) and Sun et al. (1995) to calculate chronic toxicity
values from acute toxicity data. A computer program called ACE (acute-to-chronic
estimation; version 2.0) is available from USEPA to do such calculations (USEPA
2003c). Unfortunately, as noted in the Australia/New Zealand guidelines, it is
almost impossible to obtain the raw data required to use these models.
The TTE methods are under review in the US and the UK for possible revisions
to derivation methodologies. The Water-based Criteria Subcommittee (WCS) of the
USEPA is planning to propose that kinetic-based modeling be incorporated into
revised guidelines (USEPA 2005). Although the exact model has not been deter-
mined, the workgroup, who are considering a model (or models) that will describe
the time course of toxicity, will include a toxicant accumulation component. To
improve the UK methodology, Whitehouse et al. (2004) has recommend use of
survival time modeling, accelerated life testing, and theoretically derived functions
that may address the time dependence of toxicity (Dixon and Newman 1991;
Newman and Aplin 1992; Newman and McCloskey 1996; Sun et al. 1995). These
methods may be used to determine the risk of death within a given time interval,
depending on toxicant concentration. Whitehouse et al. (2004) determined that the
two-step linear regression method of Mayer et al. (2002) is a relatively easy way to
generate LC 0 values (i.e., chronic toxicity values derived from LC 50 data), which
may then be used to construct SSDs for determining hazardous concentrations.
Whitehouse et al. (2004) reported that data required for the TTE analysis (i.e., survival
at 0, 24, and 48 hr, etc.) is usually collected during standard ecotoxicity tests, but
is often not reported (and not obtainable). Thus, this type of analysis would not
require new test procedures, but would require new reporting procedures.
The USEPA WCS is considering the use of population models to provide means
for criteria to reflect population recovery, after toxic events (USEPA 2005). Such a