Proposed Changes to Interim Methodology for Endangered Species Risk Assessments for Pesticides

Susan L. Stephens and Steven Kahn

The Environmental Protection Agency (EPA) is currently seeking comments on proposed revisions to the interim method used to evaluate risks pesticides pose to endangered species. EPA uses its risk assessment method to make biological evaluations (BEs) for pesticides being evaluated for registration under the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) for endangered and threatened species and their critical habitats. Those BEs result in one of two determinations: a pesticide’s use will have “no effect” on listed species or their habitats or “may affect” them. A “may affect” determination triggers the involvement of the U.S. Fish and Wildlife Service and/or the National Oceanic and Atmospheric Administration’s National Marine Fisheries Service (hereinafter, the Services) pursuant to section 7 of the Endangered Species Act (ESA). These proposed revisions are the most recent step in what the EPA expects to be an extended iterative process.

I. Background

This series of guidance documents is intended to streamline the regulatory framework created by FIFRA and ESA. Under section 7(a)(2) of the ESA, federal agencies are required to ensure that their actions are consistent with ESA. If a federal action “may affect” a threatened or endangered species or its federally designated critical habitat, consultation with the Services is required to determine whether the agency action is likely to jeopardize a listed species or destroy or adversely modify any designated critical habitat of such species.EPA registration of pesticides under FIFRA is an agency action that must comply with ESA.

The EPA and the Services have similar policy objectives in this area—ensuring that pesticides do not harm the environment, endangered or threatened species, or their designated critical habitats—but their statutory mandates have historically made it difficult for them to settle on a risk assessment framework. Under FIFRA, the EPA may only register a pesticide if it will “perform its intended function without unreasonable adverse effects on the environment.” This occurs when there is not an “unreasonable risk to man or the environment taking into account the economic, social, and environmental costs and benefits of the use of any pesticide.”In contrast, under the ESA, any federal action that “may affect” an endangered species creates a duty to consult with one or both of the Services to determine whether the action if “likely to adversely affect” listed species or critical habitat and, if so, to ensure that it “is not likely to jeopardize the continued existence of any endangered species or threatened species or result in the destruction or adverse modification of [critical habitat] of such species.” This encompasses any action that “reasonably would be expected, directly or indirectly, to reduce appreciably the likelihood of both survival and recovery of a listed species in the wild . . . .”

The main difference between the two statutes lies in how they contemplate risk. FIFRA requires the EPA to consider economic costs in determining unreasonable effect, whereas the ESA requires a risk mitigation–based approach irrespective of its economic consequences. In addition, FIFRA involves a national registration process, whereas the ESA requires far more localized data. As a consequence, the EPA and the Services each developed their own risk assessment methods, which at times has made it difficult for them to agree.

To simplify things, in 2010, the agencies requested that the National Research Council (NRC) form a Committee on Ecological Risk Assessment, which eventually resulted in a report called Assessing Risks to Endangered and Threatened Species from Pesticides. The agencies then incorporated many of these recommendations into a new risk assessment framework for pesticide-related impacts on endangered species.

II. The Interim Method

In 2013, the agencies released a draft interim method largely based on the 2010 NRC report with the goal of using the 2014 pesticide registration review program to field test their methodology and then make iterative improvements to it. The interim method currently in use involves a three-step interagency consultation between the EPA and the Services. Step one requires the EPA to determine if a proposed federal action “may affect” a species or critical habitat. If the proposed action is reasonably likely to cause individual endangered species or their habitats to be exposed to a pesticide, then the process moves on to step two.However, if exposure is not reasonably likely to occur, the EPA instead issues a “no effect” determination and consultation ends. At step two, the EPA determines if exposure to a pesticide will either be “not likely to adversely affect” (NLAA) or “likely to adversely affect” (LAA) a species or habitat. Together, these two steps make up the BE. If the BE results in a NLAA determination, then the EPA registration process only requires an informal consultation with the Services to confirm the determination. On the other hand, if the BE results in a LAA determination (or if the Services disagree with the NLAA determination), then the process continues to step three, formal consultation, where the Services issue a biological opinion discussing the likelihood that the proposed federal action “jeopardize[s] the continued existence of the species.” Because the BE is a gatekeeper to the Services’ ability to issue a jeopardy finding that invokes the substantive requirements of the ESA, it is important that it be based on transparent and reliable scientific methodologies.

III. Problems with the Interim Method

Actual application of the interim method led to results that were far more restrictive than expected, much to the chagrin of stakeholders in pesticide-using industries, such as agribusinesses. In 2016, the EPA posted three draft BEs for chlorpyrifos, diazinon, and malathion using its 2013 interim method. Using that method, a significant majority of species considered in those BEs were found to be “likely to [be] adversely affect[ed]” by the pesticides; each would thus require a formal consultation and Biological Opinion before the pesticide could be registered.For chlorpyrifos in particular, the BE found that it “may affect” 99 percent of all listed species and that it was “likely to adversely affect” 97 percent of all listed species. Although the threshold for requiring consultation is intended to be a relatively low one, any screening process that allows nearly 100 percent of a dataset to be flagged as likely adversely affected is likely an ineffective one. This extraordinarily high rate of “may affect” and LAA determinations has been attributed to several extremely conservative assumptions in the draft interim method, such as an assumption that all pesticides would be applied to all possible crops at maximum application rates simultaneously, rather than collecting actual usage data corresponding to realistic rates of application.

In addition, the 2016 BEs are also extremely complex. Together, the BEs for chlorpyrifos, diazinon, and malathion are more than 12,000 pages long and are largely composed of dense technical material. Review of so much material creates a high barrier to entry for non-industry aligned commenters and results in a less deliberative process overall. In fact, even many large agribusiness corporations like Dow and Bayer found it difficult to meaningfully comment on the three draft BEs within the 60-day timeframe due to the length, complexity, and the resources involved in reproducing the EPA’s methodologies.

IV. Revisions to the Interim Method

The EPA recognized after publishing the 2016 draft BEs and collecting comments that the 2013 interim method was overly inclusive and did not produce meaningful results. In 2019, EPA announced a public meeting to discuss possible changes to the method with the goal of making the risk assessment process less conservative and more efficient. EPA has proposed four changes to the method that should ultimately decrease the total amount of LAA determinations and simplify the resulting BEs:

  1. collect actual usage data for pesticides instead of relying on conservative assumptions;
  2. reinterpret a less than one percent overlap between a species’ habitat and a pesticide use site as being outside acceptable margins of error; 
  3. replace the deterministic methods used to determine the likelihood of a species being affected by a pesticide with probabilistic methods, which should better account for uncertainties and lead to more accurate results; and
  4. apply a weight of the evidence framework to better enable administrative expertise to differentiate between data that should lead to LAA or NLAA determinations.

The adjustment made to the acceptable margin of error for overlap between a species’ habitat and pesticide use sites is intended to address the possibility that no endangered species lives within the areas where a pesticide will be used.The EPA utilizes several geospatial data services in its overlap analysis to record where endangered species habitats are located and where pesticides will be used, but these services are limited in the granularity and resolution of the information that they provide. When the mapping data shows a less than one percent overlap, it means that there is an actual overlap of “only a few pixels” when the two maps are compared, which the EPA no longer considers to be reliable. The replacement of deterministic methods for probabilistic methods allows the EPA to make judgment calls on the likelihood or severity of an exposure to individual species based on their unique dietary and lifestyle patterns. The EPA did not propose an exhaustive list of these criteria in its proposal, but it has listed several examples of what it intends to consider in this portion of its analysis. These example considerations will include analysis of species dormancy, migration patterns, or seasonal dietary preferences in conjunction with seasonal pesticide uses and EPA evaluation of the use and toxicity data that it is using to reach these conclusions.

These changes should lead to a net improvement in the interagency consultation process. Insofar as it is justified by the science, cutting down on the quantity of LAA determinations is a reasonable goal, as it will enable the agencies to focus on actual risks rather than getting bogged down analyzing harms that are not meaningfully likely to occur. Furthermore, these changes should simplify the results of the BE process, allowing for more stakeholder and citizen input and a more deliberative registration process.

Comments are being collected until July 1, and the public meeting is set for July 10.

Susan L. Stephens is a shareholder at Hopping Green & Sams in Tallahassee, Florida. Steven Kahn is a law clerk (Florida State University College of Law) at Hopping Green & Sams in Tallahassee, Florida.