Multiple Failings in EPA’s Pesticide Regulation Exposed by Analysis of Herbicides Risk to Salmon
A biological opinion released by the National Marine Fisheries service on the herbicides oryzalin, pendimethalin and trifluralin brings to light many areas in EPA’s current ecological risk assessment process that underestimate risk and fail to meet modern standards of analysis.
The biological opinion process began with EPA writing an “effects determination” or initial analysis of the risks these three herbicides – oryzalin, pendimethalin and trifluralin - pose to salmon. Those determinations are then turned over to NMFS to help them in completing the biological opinion.
EPA failed to provide any analysis of these chemical breakdown products or the other ingredients added to these chemicals before they are sold.
“No exposure estimates were provided for other identified stressors of the action including inert/other ingredients, other active ingredients with formulations, and for the toxic degradates of the active ingredients. These missing estimates introduce substantial uncertainty into the exposure analysis.” (5th BiOp p 392)
“The pendimethalin and trifluralin assessments [by EPA] are silent on the issue of degradate toxicity. … we [NMFS] question the assumption that a lack of evidence regarding other degradates is sufficient to not consider them. We find it reasonable to assume degradate chemicals which retain the characteristic dinitroaniline structure may also be toxic in the same fashion as the parent.” (p 386)
“In most cases, any specific degradate produced falls below the 10% of applied threshold EPA uses to define a “major” degradate, which are often included in their analyses.” (5th BiOp p 385)
NMFS's research also pointed out the failings of EPA's models to accurately evaluate risk.
"Recent formal consultation and reviews of EPA informal consultations by the Services found that concentrations measured in surface water sometimes exceed peak concentrations predicted with PRZM/EXAMS modeling (NMFS, 2007b, 2008c, 2009b; USFWS, 2008). These findings demonstrate that the EECs generated using PRZM-EXAMS can underestimate peak concentrations of active ingredients that occur in salmonid habitats. Consequently, underestimation of exposure and subsequent risk to species is likely. Below, we discuss the primary reasons why EPA’s exposure estimates do not represent worst case exposures to salmonid habitats." (p 397)
EPA did not evaluate all the ways salmon could be exposed to these herbicides
“EPA’s BEs mostly address the potential effects of exposure to the three dinitroanilines in the
dissolved phase. This is the compartment in which it is most bioavailable to aquatic organisms.
It is also the route of exposure evaluated by the guideline tests and the medium for which we
have measured and estimated exposure concentrations. However, there are other routes of
exposure not commonly addressed in the pesticide assessments developed by OPP.” (5th BiOp p 390)
EPA’s labels, legal documents explain how these herbicides can be used, lack significant information which could lead to very heavy use of these chemicals.
“EPA-authorized labels contain language that frequently does not provide clear
distinctions on product use (e.g., many labels do not specify the maximum number of
applications, application interval, or maximum annual application rate)…”
(5th BiOp p 424-25)