National Family Farm Coalition v. Usepa

FOR PUBLICATION

UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT

NATIONAL FAMILY FARM No. 17-70810
COALITION; FAMILY FARM
DEFENDERS; BEYOND PESTICIDES; EPA No.
CENTER FOR BIOLOGICAL DIVERSITY; EPA-HQ-OPP-
CENTER FOR FOOD SAFETY; 2016-0594
PESTICIDE ACTION NETWORK NORTH
AMERICA,
Petitioners,

v.

U.S. ENVIRONMENTAL PROTECTION
AGENCY; ANDREW R. WHEELER, in
his official capacity as
Administrator,
Respondents,

DOW AGROSCIENCES LLC,
Respondent-Intervenor.
2 NAT’L FAMILY FARM COALITION V. USEPA

NATURAL RESOURCES DEFENSE No. 17-70817
COUNCIL,
Petitioner, EPA No.
EPA-HQ-OPP-
v. 2016-0594

ANDREW R. WHEELER, in his official
capacity as Administrator of the OPINION
United States Environmental
Protection Agency; U.S.
ENVIRONMENTAL PROTECTION
AGENCY,
Respondents,

DOW AGROSCIENCES LLC,
Respondent-Intervenor.

On Petition for Review of an Order of the
Environmental Protection Agency

Argued and Submitted May 16, 2019
Submission Withdrawn May 30, 2019
Resubmitted July 22, 2020
Portland, Oregon

Filed July 22, 2020

Before: N. Randy Smith, Paul J. Watford, and
Ryan D. Nelson, Circuit Judges.

Opinion by Judge R. Nelson;
Concurrence by Judge R. Nelson;
Dissent by Judge Watford
 NAT’L FAMILY FARM COALITION V. USEPA 3

SUMMARY *

Environmental Protection Agency

The panel granted one petition for review, denied
another petition for review, and remanded without vacatur to
the Environmental Protection Agency (“EPA”) in actions
challenging the EPA’s decisions to register Enlist Duo – a
pesticide designed to kill weeds on corn, soybean, and cotton
fields – in 2014, 2015, and 2017.

Enlist Duo combines two chemicals – 2,4-
dichlorophenoxyacetic acid (“2,4-D”) choline salt and
glyphosate.

The panel held that the petitions for review were timely.
A petition for review challenging a pesticide registration
order in a court of appeal must be filed within 60 days after
entry of such order. Here, the 2017 Notice of Registration
was signed on January 12, 2017. The panel held that because
the “date of entry” was not “explicitly” provided in the
Notice of Registration, the “date of entry” was “two weeks
after … [the Notice of Registration was] signed” – January
26, 2017. 40 C.F.R. § 23.6. The petitions filed 54 days later
were therefore timely. 7 U.S.C. § 136n(b).

The panel next addressed petitioners’ Article III
standing. First, concerning the claims under the Federal
Insecticide, Fungicide, and Rodenticide Act (“FIFRA”),
alleging that EPA misapplied FIFRA’s procedural

*
This summary constitutes no part of the opinion of the court. It
has been prepared by court staff for the convenience of the reader.
4 NAT’L FAMILY FARM COALITION V. USEPA

requirements and lacked substantial evidence in support of
its decision that Enlist Duo’s registration complied with
those requirements, the panel held that petitioners National
Resource Defense Council (“NRDC”) and Center for Food
Safety (“CFS”), based on their members’ standing, both had
associational standing to bring FIFRA claims. Because one
petitioner from each petition had associational standing, the
panel did not need to decide whether the other National
Family Farm Coalition (“NFFC”) petitioners had
associational standing. Second, concerning the claims
under the Endangered Species Act (“ESA”), alleging that
EPA violated the ESA’s consultation procedures in
registering Enlist Duo, the panel held that because one of
CFS’s members had Article III standing, the organization
also had associational standing to bring the ESA claims. In
addition, the Article III standing of one NFFC petitioner
made the ESA claims asserted by NFFC petitioners
justiciable.

Turning to the merits, the panel considered petitioners’
FIFRA claims. FIFRA is a regulatory scheme aimed at
controlling the use, sale, and labeling of pesticides; and the
mechanism used to further this aim is a process called
“registration.” Registration can be unconditional or
conditional, and both types often involve “pesticide
products.”

The panel rejected NRDC’s claim that the EPA
incorrectly applied what NRDC believed to be the more
lenient “conditional” registration standard rather than the
more stringent “unconditional” standard when it registered
Enlist Duo in 2014. First, the panel held that NRDC waived
the argument. Second, even absent waiver, the panel held
that NRDC’s argument was not persuasive. The registration
 NAT’L FAMILY FARM COALITION V. USEPA 5

documents supported the conclusion that EPA was applying
the unconditional standard.

NFFC petitioners argued that EPA incorrectly applied
FIFRA’s “cause any unreasonable adverse effects”
unconditional registration standard in its 2017 registration
decision. EPA conceded that it cited the wrong standard, but
the panel held that any error was harmless because the
standard for unconditional registration was higher, not
lower, than the standard for conditional registration. The
panel held that the error did not show that EPA lacked
substantial evidence to support its conclusions.

Petitioners argued that EPA lacked substantial evidence
for its 2014, 2015, and 2017 registration decisions for four
reasons. First, the panel agreed with petitioners that EPA
failed to properly assess harm to monarch butterflies from
increased 2,4-D use on milkweed in target fields. The panel
held that given the record evidence suggesting monarch
butterflies may be adversely affected by 2,4-D on target
fields, EPA was required, under FIFRA, to determine
whether any effect was “adverse” before determining
whether any effect on the environment was, on the whole,
“unreasonable.” The panel concluded that EPA’s failure to
do so meant that its decision was lacking in substantial
evidence on the issue. Second, the panel rejected the
argument that EPA failed to consider that Enlist Duo would
increase the use of glyphosate over time. The panel held that
substantial evidence supported EPA’s conclusion that
neither the initial 2014 registration of Enlist Duo – nor the
subsequent approvals for new use – will increase the overall
use of glyphosate. Third, the panel rejected petitioners’
contention that EPA failed to properly consider 2,4-D’s
volatility – i.e., its tendency to evaporate into a gas and drift
to non-target plants. The panel held that EPA reasonably
6 NAT’L FAMILY FARM COALITION V. USEPA

relied on studies to support its conclusion that the volatility
of 2,4-D choline salt will not cause on unreasonable adverse
effects on the environment. Accordingly, substantial
evidence supported EPA’s findings. Fourth, the panel
rejected NFFS petitioners’ contention that EPA should have
accounted for the potential synergistic effect of mixing
Enlist Duo with a different chemical called glufosinate. The
panel held that this concern was speculative. In conclusion,
as to FIFRA, the panel granted NRDC’s petition for review
in part, and denied it in part.

The panel next addressed, and rejected, the petitioners’
ESA claims. The ESA and its implementing regulations
delineate a process – known as Section 7 consultation – for
determining the biological impacts of a proposed action.
The process starts with a determination whether the
proposed action will have “no effect” or if it “may effect”
listed species or critical habitat. If an action will have no
effect, no consultation with the expert agencies is needed.

First, the panel rejected NFFC petitioners’ challenge to
EPA’s “no effect” findings for plants and animals. The panel
held that the EPA did what the ESA required it to do: assess
risks to determine whether the exposure of protected species
and critical habitat to potentially harmful chemicals would
have any possible effect. The panel concluded that EPA’s
ultimate “no effect” findings, and adoption of mitigation
measures, were not arbitrary, capricious, or contrary to law.
Second, the panel rejected NFFC petitioners’ argument that
EPA’s rationale for limiting the “action area” to the treated
field was not sound. The panel accorded deference to the
EPA in the way it chose to define the action area. Third, the
panel rejected NFFC petitioners’ argument that EPA
violated its duty to insure no “adverse modification” of
“critical habitat” by relying on its 2016 risk assessment.
 NAT’L FAMILY FARM COALITION V. USEPA 7

Finally, the panel addressed the remedy for EPA’s error in
its registration decisions under FIFRA. The panel held that
remand without vacatur was warranted. EPA’s error in
failing to consider harm to monarch butterflies caused by
killing target milkweed was not “serious.” The panel
remanded so that EPA can address the evidence concerning
harm to monarch butterflies and whether the registration of
Enlist Duo will lead to an unreasonable adverse effect on the
environment.

Concurring, Judge R. Nelson wrote separately to address
how the interplay of FIFRA’s venue provision and standing
could make a difference in a future case. In this case, the
interplay between FIFRA’s venue provisions and Article III
standing did not make a difference because, for each
petition, one petitioner over which venue was proper also
demonstrated standing.

Dissenting, Judge Watford agreed with the majority that
there was jurisdiction to review the petitioners’ challenges
and that the EPA violated FIFRA by failing to assess the
impact that Enlist Duo’s use will have on the monarch
butterfly. However, in his view, EPA also violated the ESA
by failing to use the best scientific data to assess whether
Enlist Duo will adversely affect threatened or endangered
species. Accordingly, he would vacate the 2014 and 2017
registrations under review.
8 NAT’L FAMILY FARM COALITION V. USEPA

COUNSEL

George A. Kimbrell (argued), Amy Van Saun, and Sylvia
Shih-Yau Wu, Center for Food Safety, Portland, Oregon;
Paul H. Achitoff, Earthjustice, Honolulu, Hawaii; for
Petitioners National Family Farm Coalition, Family Farm
Defenders, Beyond Pesticides, Center for Biological
Diversity, Center for Food Safety, and Pesticide Action
Network North America.

Margaret T. Hsieh (argued), Mitchell S. Bernard, and Kaitlin
Morrison, Natural Resources Defense Council, New York,
New York; Peter J. DeMarco and Aaron Colangelo, Natural
Resources Defense Council, Washington, D.C.; for
Petitioner Natural Resources Defense Council.

J. Brett Grosko (argued), Senior Trial Attorney; Michele L.
Walter, Trial Attorney; Jeffrey H. Wood, Acting Assistant
Attorney General; Jonathan D. Brightbill, Principal Deputy
Assistant Attorney General; Environment and Natural
Resources Division, United States Department of Justice,
Washington, D.C.; Benjamin Wakefield and Michele Knorr,
Pesticides & Toxics Substances Law Office, Office of
General Counsel, United States Environmental Protection
Agency, Washington, D.C.; for Respondents.

Kathleen M. Sullivan (argued), Quinn Emanuel Urquhart &
Sullivan LLP, New York, New York; David B. Weinberg,
Wiley Rein LLP, Washington, D.C.; Stanley H. Abramson
and Donald C. McLean, Arent Fox LLP, Washington, D.C.;
Christopher Landau and Archith Ramkumar, Quinn
Emanuel Urquhart & Sullivan LLP, Washington, D.C.;
Andrew P. March, Quinn Emanuel Urquhart & Sullivan
LLP, Redwood Shores, California; for Respondent-
Intervenor.
 NAT’L FAMILY FARM COALITION V. USEPA 9

Kirsten L. Nathanson, David Y. Chung, and Amy B.
Symonds, Crowell & Moring LLP, Washington, D.C., for
Amicus Curiae CropLife America.

Matthew H. Lembke and Anna M. Manasco, Bradley Arant
Boult Cummings LLP, Birmingham, Alabama;
Bartholomew J. Kempf, Bradley Arant Bould Cummings
LLP, Nashville, Tennessee; for Amici Curiae American
Farm Bureau Federation, American Soybean Association,
National Cotton Council of America, and National Corn
Growers Association.

OPINION

R. NELSON, Circuit Judge:

Petitioners challenge EPA’s decisions to register Enlist
Duo—a pesticide designed to kill weeds on corn, soybean,
and cotton fields—in 2014, 2015, and 2017. According to
Petitioners, EPA’s decisions violate FIFRA and the ESA.
We grant one petition in part as to FIFRA, deny the other
petition as to both the ESA and FIFRA, and remand to the
agency without vacatur.

I

Corn, soybeans, and cotton are three of the most
important agricultural commodities in the United States.
Corn is the primary feed grain in the United States and
worldwide, soybeans are the world’s largest source of
protein feed for animals and the second largest source of
vegetable oil, and cotton is one of the most important textile
fibers in the world.
10 NAT’L FAMILY FARM COALITION V. USEPA

These crops provide vital value to the United States and
the world. Domestically, these three crops together have a
gross production value of approximately $103 billion per
year. Internationally, the United States is the world’s
leading corn and soybean producer and exporter. The United
States also provides, together with China and India, two-
thirds of the world’s cotton.

This important industry, however, is not immune from a
plight that threatens every gardener: weeds. Since the
1970s, glyphosate dimetyhlammonium salt (“glyphosate”)
has been used on corn, soybeans, and cotton crops to reduce
weeds. Over time, however, certain noxious weeds have
grown resistant to glyphosate. That resistance in turn
decreases crop yield, with severe economic consequences.

To help solve this problem, Dow Agrosciences LLC
(“Dow”) invented Enlist Duo. Enlist Duo combines two
chemicals—2,4-dichlorophenoxyacetic acid (“2,4-D”)
choline salt and glyphosate. Both 2,4-D and glyphosate have
been registered for certain uses as pesticides for decades.
When combined, however, they represent a significant
improvement over glyphosate and 2,4-D, used separately.
Combining the two chemicals delays the development of the
weeds’ resistance and allows pesticide use later in the
growing season, thereby improving yields.

EPA issued a final order registering Enlist Duo under the
Federal Insecticide, Fungicide, and Rodenticide Act
(“FIFRA”) in October 2014. In that registration decision,
EPA did not perform a risk assessment for Enlist Duo’s
glyphosate component. Instead, it found that a new
assessment was not needed because glyphosate was already
being used in the same way in other pesticides to treat weeds.
But the same was not true for 2,4-D. That component was
being approved for use later into the growing season and on
 NAT’L FAMILY FARM COALITION V. USEPA 11

taller-growing crops for the first time. So EPA performed a
full risk assessment for 2,4-D. That analysis assessed human
health risks; ecological risks; and risks to endangered
species, plants, and critical habitats posed by 2,4-D. It also
considered whether 2,4-D would volatilize—that is,
evaporate into a gas—and drift to non-target plants and
animals.

EPA found, based on multiple studies, that the type of
2,4-D in Enlist Duo—a choline salt variety—is less volatile
than other forms of 2,4-D. That meant there was no risk of
harm off the treated field, so long as the label
requirements—including the use of nozzles, buffers, and
avoiding application aerially—were followed to avoid the
risk of spray drift. This finding led EPA to limit the “action
area” to treated fields, thereby reducing the number of
species subject to an ESA analysis. EPA then concluded,
based on its FIFRA risk assessments and conservative ESA
analysis, that Enlist Duo’s registration would “not generally
cause unreasonable adverse effects on the environment”
under FIFRA and would comply with the ESA, subject to
certain use restrictions.

Based on this conclusion, EPA issued a registration of
Enlist Duo under FIFRA, which allowed Enlist Duo to be
used on corn and soybean crops in six states. EPA’s
decision, however, was ambiguous as to which FIFRA
registration standard it was applying. The pesticide license
approved an “unconditional” registration. So did the
Proposed Decision Document. But the final registration
document articulating EPA’s reasoning cited FIFRA’s
“conditional” registration provision instead. EPA also
referenced additional data requirements in the registration,
even though outstanding data requirements are typically
referenced in conditional registrations.
12 NAT’L FAMILY FARM COALITION V. USEPA

Petitioners National Family Farm Coalition, Family
Farm Defenders, Beyond Pesticides, Center for Biological
Diversity (“CBD”), Center for Food Safety (“CFS”), and
Pesticide Action Network North America (“PANNA”)
(collectively, the “NFFC Petitioners”) and Petitioner
National Resource Defense Council (“NRDC”) challenged
EPA’s 2014 registration decision in this Court. NRDC v.
EPA, No. 14-73353 (9th Cir. Oct. 30, 2014); Ctr. for Food
Safety v. EPA, No. 14-73359 (9th Cir. Oct. 30, 2014). While
that litigation was pending, EPA issued a final order
amending the 2014 registration to allow the use of Enlist
Duo on corn and soybean crops in an additional nine states.
That 2015 registration decision was supported by an
ecological risk assessment for the protected species in the
new states. The decision also relied on critical habitat
modification determinations for the new uses of 2,4-D.

Petitioners challenged the 2015 registration decision as
well, NRDC v. EPA, No. 15-71213 (9th Cir. Apr. 20, 2015);
Ctr. for Food Safety v. EPA, No. 15-71207 (9th Cir. Apr. 20,
2015), and the challenges to the 2014 and 2015 registration
decisions were consolidated in one proceeding. But briefing
was never completed. Instead, EPA moved to remand and
vacate the 2014 and 2015 registrations. EPA did so after
discovering that Dow had filed a patent application with the
U.S. Patent and Trademark office claiming “synergism”—
that is, two chemicals working together to produce a greater
combined effect than they would separately—between
glyphosate and 2,4-D. This Court granted the request to
remand the case, but denied the request for vacatur, leaving
the 2014 and 2015 registration decisions in place.

Shortly thereafter, on January 12, 2017, EPA issued
another registration decision regarding Enlist Duo. In that
decision, EPA relied on new data on synergy to conclude
 NAT’L FAMILY FARM COALITION V. USEPA 13

that no concern lay with synergy between the glyphosate and
2,4-D in Enlist Duo. EPA did not, however, address
evidence that destruction of milkweed on target fields would
harm the monarch butterfly population. The decision also
contained three main conclusions. First, it reaffirmed EPA’s
2014 and 2015 registration decisions. Second, it authorized
the use of Enlist Duo on corn and soybean crops in
19 additional states, bringing the total number of permitted-
use states to 34. Third, it authorized a new use of Enlist Duo
on cotton crops in all 34 states.

To support these decisions, EPA relied in part on its prior
analysis of glyphosate and 2,4-D. But EPA did perform
some new analysis. For example, EPA relied on an updated
ecological risk assessment for 531 ESA-listed species in the
34 states where Enlist Duo was approved. The updated risk
assessment, like the prior ones, used an iterative approach,
through which species were ruled out and given “no effect”
findings if their exposure to 2,4-D did not exceed a set “level
of concern” after screening-level and, in some cases,
species-specific assessments. Using this methodology, EPA
made “no effect” findings as to all plants and animals off the
treated field, after imposing similar mitigation measures as
it had in 2014. This same methodology supported EPA’s
“no effect” findings for 19 of 23 species on the treated field.
EPA therefore did not contact the consultation agencies as to
these species. As to the remaining four species, EPA
imposed location-based label restrictions to avoid harm to
three of them. EPA then consulted the Fish and Wildlife
Service (“FWS”) as to the Eskimo curlew—after which
FWS concurred with EPA’s conclusion that the Eskimo
curlew would not be adversely affected by Enlist Duo. The
2017 decision, relying on new critical habitat analysis, also
concluded that no critical habitats would be affected because
the eight species that occurred on corn, cotton, and soybean
14 NAT’L FAMILY FARM COALITION V. USEPA

fields did not have physical or biological features essential
to the species in agricultural fields.

Despite this new data and analysis, there were, for the
first time, data gaps relating to 2,4-D that were not present
during the prior registrations. These gaps—which related to
2,4-D generally—meant that EPA could not register Enlist
Duo unconditionally. Instead, EPA registered the entire
Enlist Duo product on a “conditional” basis under FIFRA.
In doing so, however, EPA cited FIFRA’s unconditional
“cause unreasonable adverse effects” standard rather than
FIFRA’s conditional “significantly increase the risk of
unreasonable adverse effects” standard.

Petitioners challenged EPA’s 2017 decision on March
21, 2017. In the resulting briefing, the parties disagreed
about whether EPA’s 2014 and 2015 registration decisions
could also be reviewed. We held, after oral argument, that
all three decisions were subject to review. We then ordered
the parties to submit supplemental briefing addressing any
challenges to the 2014 and 2015 registrations.

II

We first address whether this case is properly before us.
EPA does not raise any overarching challenge to
jurisdiction. 1 Dow, by contrast, argues that (1) the petitions

1
EPA does argue that NRDC lacks Article III standing to raise
arguments about glyphosate. According to EPA, any favorable decision
about glyphosate would not redress NRDC’s alleged injuries because
glyphosate will continue to be used in the same quantities. We address—
and reject—that argument below. See infra Section II.B.2.
 NAT’L FAMILY FARM COALITION V. USEPA 15

for review were untimely; and (2) Petitioners lack
associational standing. 2

A

We begin with Dow’s argument that the petitions for
review were untimely. A petition for review challenging a
pesticide registration order in a court of appeals must be filed
“within 60 days after the entry of such order.” 7 U.S.C.
§ 136n(b). 3 The “date of entry of an order” is governed by
regulation. 40 C.F.R. § 23.6. “Unless . . . [EPA’s]
Administrator otherwise explicitly provides in a particular
order, the time and date of entry of an order issued by the
Administrator” is “two weeks after it is signed.” Id.

Here, the 2017 Notice of Registration was signed on
January 12, 2017. In addition, the “Date of Issuance” on the
Notice of Registration is January 12, 2017. But Petitioners
did not file their petitions until March 21, 2017—68 days

2
Dow also argues that venue is improper as to three of the six NFFC
Petitioners (National Family Farm Coalition, Family Farm Defenders,
and Beyond Pesticides) because none of them “reside[]” or “ha[ve] a
place of business” in the Ninth Circuit. 7 U.S.C. § 136n(b). But we need
not address that argument. Venue is proper as to the other three NFFC
Petitioners (CFS, CBD, and PANNA) because they do “reside” or “have
a place of business” in the Ninth Circuit. So regardless whether venue
is improper as to three of the six NFFC Petitioners, we can address the
merits of the NFFC petition.
3
FIFRA also allows for judicial review of EPA’s decisions in
federal district courts when EPA refuses to “cancel or suspend a
registration or to change a classification” or for “other final actions of
the Administrator not committed to the discretion of the Administrator
by law.” 7 U.S.C. § 136n(a).
16 NAT’L FAMILY FARM COALITION V. USEPA

after the Notice of Registration was issued. According to
Dow, this means the petitions were eight days too late.

Dow’s argument rests on the date of issuance—January
12—being the “date of entry” of the order under 7 U.S.C.
§ 136n(b) and 40 C.F.R. § 23.6. But for this argument to
work, the date of issuance must say that it is the “date of
entry” “explicitly,” 40 C.F.R. § 23.6—that is, “without
ambiguity or vagueness,” Explicit, Black’s Law Dictionary
(10th ed. 2014). And here, considerable ambiguity exists.
The Notice of Registration does not “explicitly” include a
“date of entry.” Nor does “issue” mean the same thing as
“entry.” Compare Issue, Black’s Law Dictionary (10th ed.
2014) (“to send out or distribute officially”), with Entry,
Black’s Law Dictionary (10th ed. 2014) (“the placement of
something before the court or on the record”). Thus, the date
of issuance of the 2017 registration does not “explicitly”
indicate the “date of entry.”

Because the “date of entry” was not “explicitly”
provided in the Notice of Registration, the “date of entry”
was “two weeks after . . . [the Notice of Registration was]
signed”—January 26, 2017. 40 C.F.R. § 23.6. The petitions,
filed 54 days later, were therefore timely. 7 U.S.C.
§ 136n(b); 40 C.F.R. § 23.6. 4

B

Dow also argues Petitioners lack standing to bring their
petitions for review. To have associational standing, each
organization must show that “(a) its members would
otherwise have standing to sue in their own right; (b) the

4
This comports with EPA’s interpretation of the relevant statute and
regulation in this and other cases.
 NAT’L FAMILY FARM COALITION V. USEPA 17

interests it seeks to protect are germane to the organization’s
purposes; and (c) neither the claim asserted nor the relief
requested requires the participation of individual members
in the lawsuit.” Am. Diabetes Ass’n v. U.S. Dep’t of the
Army, 938 F.3d 1147, 1155 (9th Cir. 2019) (citation
omitted). Here, there is no dispute that at least one Petitioner
from each petition for review has satisfied the second and
third requirements. We therefore decide whether at least one
Petitioner from each petition has shown that at least one of
its members would have Article III standing to sue in his or
her own right. Mont. Shooting Sports Ass’n v. Holder,
727 F.3d 975, 981 (9th Cir. 2013) (“the presence in a suit of
even one party with standing suffices to make a claim
justiciable” (quoting Brown v. City of Los Angeles, 521 F.3d
1238, 1240 n.1 (9th Cir. 2008))).

Article III of the United States Constitution confines
federal courts to hearing only “[c]ases” and
“[c]ontroversies.” U.S. Const. art. III, § 2, cl. 1. “A suit
brought by a plaintiff without Article III standing is not a
‘case or controversy,’ and an Article III federal court
therefore lacks subject matter jurisdiction over the suit.”
City of Oakland v. Lynch, 798 F.3d 1159, 1163 (9th Cir.
2015) (citation omitted). To establish standing, a plaintiff
must demonstrate “(1) a concrete and particularized injury
that is ‘actual or imminent, not conjectural or hypothetical’;
(2) a causal connection between the injury and the
defendant’s challenged conduct; and (3) a likelihood that a
favorable decision will redress that injury.” Pyramid Lake
Paiute Tribe of Indians v. Nev., Dep’t of Wildlife, 724 F.3d
1181, 1187 (9th Cir. 2013) (quoting Lujan v. Defs. of
Wildlife, 504 U.S. 555, 560–61 (1992)).

To meet this standard, Petitioners must show a
“substantial probability” of standing, Nw. Requirements
18 NAT’L FAMILY FARM COALITION V. USEPA

Utils. v. FERC, 798 F.3d 796, 805 (9th Cir. 2015) (citation
omitted), which is the same burden “as that of a plaintiff
moving for summary judgment in the district court,” Sierra
Club v. EPA, 292 F.3d 895, 899 (D.C. Cir. 2002). Because
a “plaintiff must demonstrate standing for each claim he
seeks to press and for each form of relief that is sought,” we
analyze Petitioners’ Article III standing for the ESA and
FIFRA claims separately. Town of Chester v. Laroe Estates,
Inc., 137 S. Ct. 1645, 1650 (2017) (quoting Davis v. FEC,
554 U.S. 724, 734 (2008)).

1

First, the FIFRA claims. NFFC Petitioners and NRDC
both assert that EPA misapplied FIFRA’s procedural
requirements and lacked substantial evidence in support of
its decisions that Enlist Duo’s registration complied with
those requirements. These are procedural injuries. Indeed,
the registration standards at issue are the safeguards put in
place by Congress to ensure that approved pesticides do not
cause adverse effects on the environment. 7 U.S.C. § 136a.
And EPA’s alleged failure to follow those standards is what
leads to the alleged “substantive harm to the environment”
in this case. Citizens for Better Forestry v. U.S. Dep’t of
Agric., 341 F.3d 961, 971 (9th Cir. 2003) (internal quotation
marks and citation omitted). We therefore apply the rules of
Article III standing that apply to procedural injuries in
determining Petitioners’ standing to assert their FIFRA
claims. See Nat. Res. Def. Council v. Jewell, 749 F.3d 776,
783 (9th Cir. 2014). NRDC and one of NFFC Petitioners,
CFS, meet that standard here based on their assertion of
procedural violations of FIFRA.

Injury in Fact for NRDC. In the context of procedural
violations, the injury-in-fact requirement is met if “the
procedures in question are designed to protect some
 NAT’L FAMILY FARM COALITION V. USEPA 19

threatened concrete interest of [the petitioner] that is the
ultimate basis of his standing.” Salmon Spawning &
Recovery All. v. Gutierrez, 545 F.3d 1220, 1225 (9th Cir.
2008) (internal quotation marks and citation omitted).
NRDC meets that standard here. Members of NRDC have
submitted declarations stating that they enjoy watching the
monarch butterfly migration where they live, that Enlist Duo
is approved for use in their states, and that they are
concerned they will no longer be able to enjoy observing
monarch butterflies because of Enlist Duo’s effects on
milkweed.

These declarations show a concrete interest for two
reasons. First, a concrete interest can be “an aesthetic or
recreational interest in a particular place, or animal, or plant
species.” Ecological Rights Found. v. Pac. Lumber Co.,
230 F.3d 1141, 1147 (9th Cir. 2000). And second, there is a
“geographic nexus between the individual[s] asserting the
claim and the location suffering [the] environmental
impact.” Ashley Creek Phosphate Co. v. Norton, 420 F.3d
934, 938 (9th Cir. 2005) (internal quotation marks and
citation omitted); see id. (“[P]laintiffs who use the area
threatened by a proposed action or who own land near the
site of a proposed action have little difficulty establishing a
concrete interest.”).

Moreover, the registration provisions at issue are
designed to protect the environment. Salmon Spawning,
545 F.3d at 1226. Both the conditional and unconditional
registration provisions in FIFRA require EPA to consider
“unreasonable adverse effects on the environment.”
7 U.S.C. § 136a(c)(5)(C), (c)(7). These effects would
include any effect on monarch butterflies, which is what
NRDC’s members are concerned about.
20 NAT’L FAMILY FARM COALITION V. USEPA

Dow argues that NRDC cannot satisfy the injury-in-fact
requirement because it cannot prove that Enlist Duo has
caused a decline in the monarch butterfly population. But “a
credible threat of harm is sufficient to constitute actual injury
for standing purposes.” Cent. Delta Water Agency v. United
States, 306 F.3d 938, 950 (9th Cir. 2002). Thus, NRDC need
only show that the exercise of a procedural right “could
protect [its] concrete interests.” Cottonwood Envtl Law Ctr.
v. U.S. Forest Serv., 789 F.3d 1075, 1082 (9th Cir. 2015). It
has done so here.

Causation and Redressability for NRDC. We now turn
to the second and third requirements for Article III standing,
which are relaxed for NRDC because it has established
injury in fact. See Salmon Spawning, 545 F.3d at 1226. The
causation requirement is satisfied by showing a “reasonable
probability of the challenged action’s threat to [NRDC’s]
concrete interest.” Hall v. Norton, 266 F.3d 969, 977 (9th
Cir. 2001) (internal quotation marks and citation omitted).
To satisfy the redressability requirement, by contrast, NRDC
need only show “that the relief requested—that the agency
follow the correct procedures—may influence the agency’s
ultimate decision of whether to take or refrain from taking a
certain action.” Salmon Spawning, 545 F.3d at 1226–27
(emphasis added).

Here, the causation element is satisfied because there is
a “reasonable probability” that EPA may have further
minimized any alleged harm to monarch butterflies had it
adopted NRDC’s arguments. Hall, 266 F.3d at 977.
Moreover, adopting NRDC’s arguments “may influence the
agency’s ultimate decision of whether to take or refrain from
taking a certain action,” which satisfies the redressability
requirement. Salmon Spawning, 545 F.3d at 1226–27
(emphasis added).
 NAT’L FAMILY FARM COALITION V. USEPA 21

Both EPA and Dow argue that this conclusion cannot be
right. EPA contends, for example, that NRDC lacks
standing to raise any arguments about glyphosate.
According to EPA, vacatur of the registration of Enlist Duo
would not redress any alleged glyphosate-based harms
because, even if Enlist Duo was not registered, other
pesticides containing glyphosate would continue to be used
such that overall glyphosate use would not decrease. Dow,
for its part, advances a very similar argument, contending
that because the registration does not alter any existing uses
of glyphosate and 2,4-D, NRDC cannot show that a
favorable decision here would redress its supposed injuries.

These arguments misunderstand the redressability
inquiry for procedural injuries. “[T]he mere existence of
multiple causes of an injury does not defeat redressability,
particularly for a procedural injury.” WildEarth Guardians
v. U.S. Dep’t of Agric., 795 F.3d 1148, 1157 (9th Cir. 2015).
“So long as a defendant is at least partially causing the
alleged injury, a plaintiff may sue that defendant, even if the
defendant is just one of multiple causes of the plaintiff’s
injury.” Id. That is the case here.

Moreover, the redressability arguments Dow and EPA
advance ask the Court to perform its Article III standing
analysis on an argument-by-argument basis. But standing is
assessed based on the claims asserted, Town of Chester,
137 S. Ct. at 1650, and the type of injury alleged, Citizens
for Better Forestry, 341 F.3d at 971, not argument-by-
argument. NRDC therefore has Article III standing to seek
vacatur of the registration decisions under FIFRA.

Standing for NFFC Petitioners. CFS has likewise shown
“that the procedures in question are designed to protect some
threatened concrete interest . . . that is the ultimate basis of
[its] standing” for purposes of standing for NFFC
22 NAT’L FAMILY FARM COALITION V. USEPA

Petitioners. Salmon Spawning, 545 F.3d at 1225 (citation
omitted). Eric Pool, a member of CFS, submitted a
declaration stating that Enlist Duo is approved for use in his
home state of Illinois and that his crops are affected by the
use of the components of Enlist Duo on nearby fields. These
effects, according to his declaration, have caused economic
damage, including harming his grapevines and forcing him
to decrease the amount of acreage he plants on.

Mr. Pool’s declaration establishes a concrete interest that
is geographically linked to his home. Ecological Rights
Found., 230 F.3d at 1147; Ashley Creek, 420 F.3d at 938.
And FIFRA is designed to protect these interests. Indeed,
the statute requires EPA to determine whether any given
action will cause “unreasonable adverse effects on the
environment.” 7 U.S.C. § 136a(c)(5)(C), (c)(7). This
effects-based test considers not only “environmental costs,”
but also “economic” ones, which are the interests Mr. Pool
claims. 7 U.S.C. § 136(bb).

The Pool declaration also meets the relaxed threshold for
showing causation and redressability for procedural injuries.
The causation element is satisfied because there is a
“reasonable probability” that EPA would have assessed the
threat to farmers differently had it adopted CFS’s arguments.
Hall, 266 F.3d at 977. Moreover, vacating the registration
so that EPA reanalyzes the issues CFS raises “may influence
the agency’s ultimate decision of whether to take or refrain
from taking a certain action,” which satisfies the
redressability prong. Salmon Spawning, 545 F.3d at 1226–
27 (emphasis added).

In sum, NRDC and CFS, based on their members’
standing, both have associational standing to bring FIFRA
claims. Am. Diabetes, 938 F.3d at 1155. Because one
petitioner from each petition has associational standing, we
 NAT’L FAMILY FARM COALITION V. USEPA 23

need not decide whether the other NFFC Petitioners have
associational standing. Mont. Shooting Sports, 727 F.3d at
981. 5

2

Next, the ESA claims. NFFC Petitioners allege that EPA
violated the ESA’s consultation procedures in registering
Enlist Duo. These alleged violations are procedural in
nature, so the standing rules applicable to such violations
apply here, too. Jewell, 749 F.3d at 783.

Injury in Fact. CFS has shown injury in fact for NFFC
Petitioners. One of its members, Leslie Limberg, submitted
a declaration stating that she lives in a state where Enlist Duo
is approved for use and that she enjoys observing
endangered species where she lives, including the Indiana
bat. This declaration shows an aesthetic and recreational
interest that is geographically linked to the individual
asserting the claim, thereby satisfying the injury-in-fact
requirement. Ecological Rights Found., 230 F.3d at 1147;
Ashley Creek, 420 F.3d at 938.

5
Neither Dow nor EPA argue that any Petitioner lacks statutory
standing under the APA to challenge the EPA’s action under FIFRA.
But we conclude those requirements—that (1) “there has been final
agency action adversely affecting the plaintiff”; and (2) the plaintiff
“suffers legal wrong or that its injury falls within the zone of interests of
the statutory provision the plaintiff claims was violated,” Citizens for
Better Forestry, 341 F.3d at 976 (citation omitted)—are met as well.
These requirements are not relevant to the ESA claim because the ESA
provides for a private right of action outside of the APA. Wash. Toxics
Coal. v. EPA, 413 F.3d 1024, 1034 (9th Cir. 2005) (citation omitted),
abrogated on other grounds as recognized in Cottonwood, 789 F.3d
at 089.
24 NAT’L FAMILY FARM COALITION V. USEPA

Moreover, the ESA’s consultation procedures that CFS
claims have been violated—for example, the requirement
that EPA consult when a proposed action “may affect” any
listed species—are designed to protect these concrete
interests. Salmon Spawning, 545 F.3d at 1229; see id.
at 1225–26. “These procedures are designed to advance the
ESA’s overall goal of species preservation, and thus the
groups’ specific goals” as well as “ensur[e] agency
compliance with the ESA’s substantive provisions.” Id.
at 1226 (citing Bennett v. Spear, 520 U.S. 154, 176 (1997)).

Dow contends CFS cannot establish injury in fact as to
the ESA claims because its members’ concerns are
speculative and far from imminent. But CFS need only
provide evidence of “an increased risk [of harm] based on a
violation of a statute.” Ocean Advocates v. U.S. Army Corps
of Eng’rs, 402 F.3d 846, 860 (9th Cir. 2005) (alteration
adopted and citation omitted). That standard is met here
because a reconsideration of ESA’s consultation standards
could lead to a different result—that is, it “could protect
[Petitioners’ members’] concrete interests.” Cottonwood,
789 F.3d at 1082.

Causation and Redressability. We now turn to the
second and third requirements for Article III standing.
Redressability is satisfied here because the consultation CFS
argues is required under the ESA may have modified EPA’s
decision. See Hall, 266 F.3d at 977 (holding redressability
prong met where relevant decision “could be influenced” by
environmental studies plaintiff requested).

Dow argues that causation is not met because CFS
acknowledges that other factors besides Enlist Duo may
endanger the species at issue and because other similar
pesticides may cause the same harms even if Enlist Duo is
never used. For purposes of standing, however, “the causal
 NAT’L FAMILY FARM COALITION V. USEPA 25

connection . . . need not be so airtight . . . to demonstrate that
the plaintiffs would succeed on the merits.” Ecological
Rights Found., 230 F.3d at 1152. Instead, the standing
inquiry focuses on whether the petitioner’s injury is “fairly
traceable to the challenged conduct,” WildEarth, 795 F.3d at
1154—that is, whether the claim of injury relies merely on
“the behavior of other parties” or “an attenuated chain of
conjecture” as to what could happen in the future, Hall,
266 F.3d at 977 (citation omitted). And here, CFS’s ESA
claim does not rely on other parties to take action or an
attenuated chain of conjecture. As a result, a “reasonable
probability” exists that EPA’s failure to consult threatens
CFS’s concrete interests and the causation prong is satisfied.
Hall, 266 F.3d at 977 (citation omitted).

Because one of CFS’s members has Article III standing,
the organization also has associational standing to bring its
ESA claims. Am. Diabetes, 938 F.3d at 1155. The Article
III standing of one NFFC Petitioner makes the ESA claims
asserted by NFFC Petitioners justiciable. Mont. Shooting
Sports, 727 F.3d at 981.

III

FIFRA “is a comprehensive regulatory scheme aimed at
controlling the use, sale, and labeling of pesticides.” Nathan
Kimmel, Inc. v. DowElanco, 275 F.3d 1199, 1204 (9th Cir.
2002). Under the statute, the mechanism used to further this
aim is a process called “registration.” 7 U.S.C. § 136a(a).
Before any pesticide can be sold or used in the United States,
EPA must register the pesticide—that is, provide a license
that establishes the terms and conditions under which a
pesticide may be lawfully sold, distributed, and used within
the United States. Id. § 136a(c). The terms and conditions
on the license include exactly what product can be sold, the
specific packaging it must be sold in, and labeling that
26 NAT’L FAMILY FARM COALITION V. USEPA

contains instructions on proper use. Id. § 136(p); 40 C.F.R.
§§ 152.115, 156.10.

Registration occurs in a variety of ways. The principal
type of registration is called unconditional registration.
7 U.S.C. § 136a(c)(5). For unconditional registration, EPA
must “review[] all relevant data in [its] possession” and
“determine[] that no additional data are necessary” to its
decision. 40 C.F.R. § 152.112(b), (c). EPA can
unconditionally register the pesticide only if it will “not
generally cause unreasonable adverse effects on the
environment” “when used in accordance with widespread
and commonly recognized practice.” Id. § 152.112(e).
“[U]nreasonable adverse effects on the environment” is
defined, in relevant part, as “any unreasonable risk to man or
the environment, taking into account the economic, social,
and environmental costs and benefits of the use of any
pesticide.” 7 U.S.C. § 136(bb).

FIFRA also allows EPA to “conditionally register or
amend the registration” of a pesticide for use in certain
“special circumstances.” 7 U.S.C. § 136a(c)(7) (emphasis
added). Conditional registration allows for an existing
registration to be amended with less data than is required for
an unconditional registration. See id. § 136a(c)(7)(B). But
it still imposes a burden on EPA. Id. To conditionally
register a pesticide or amend a pesticide registration, EPA
must determine that “the applicant has submitted satisfactory
data pertaining to the proposed additional use”; and that
“amending the registration in the manner proposed . . .
would not significantly increase the risk of any unreasonable
adverse effect on the environment.” Id.

Both types of registration often involve “pesticide
product[s].” 40 C.F.R. § 152.15. A “pesticide product” is a
“pesticide in the particular form (including composition,
 NAT’L FAMILY FARM COALITION V. USEPA 27

packaging, and labeling) in which the pesticide is, or is
intended to be, distributed or sold.” Id. § 152.3. As such, a
“pesticide product” may include one or more active or inert
chemical ingredients. See id. (“Pesticide means any
substance or mixture of substances intended for preventing,
destroying, repelling, or mitigating any pest, or intended for
use as a plant regulator, defoliant, or desiccant . . . .”).
“Active ingredient[s]” are those ingredients in the pesticide
product that do the work to “prevent, destroy, repel or
mitigate any pest.” Id. Enlist Duo, for example, is a
“pesticide product” composed of two “active ingredients”
that do the work: glyphosate and 2,4-D. See 7 U.S.C.
§ 136a(c)(7)(C) (contemplating a “pesticide” containing
more than one “active ingredient” and some “active
ingredient[s]” being registered previously).

Under FIFRA’s implementing regulations, EPA takes a
specific approach in cases involving “active ingredient[s]”
that have already been registered as part of other “pesticide
product[s].” 40 C.F.R. § 152.111. In such cases, “the
Agency will not commence a complete review of the
existing data base on a given chemical in response to receipt
of an application for registration.” Id. “Instead, the Agency
will review the application using the criteria for conditional
registration” under 7 U.S.C. § 136a(c)(7)(A) and (B). Id.
Under those provisions, EPA may “conditionally register or
amend the registration of a pesticide” if “the pesticide and
proposed use are identical or substantially similar to any
currently registered pesticide and use thereof, or differ” only
slightly and the registration would not “significantly
increase the risk of any unreasonable adverse effect on the
environment.” 7 U.S.C. § 136a(c)(7)(A); see also 40 C.F.R.
§ 152.113 (allowing conditional registration for pesticide
products “that do not contain a new active ingredient”).
These registrations are often called “me-too” registrations.
28 NAT’L FAMILY FARM COALITION V. USEPA

FIFRA also allows EPA to cancel or change the
classification of a registration if it determines that the
pesticide “generally causes unreasonable adverse effects on
the environment.” 7 U.S.C. § 136d(b). FIFRA also contains
a provision requiring that “[t]he registrations of
pesticides”—including “active ingredients”—“be
periodically reviewed,” a process called “registration
review” that takes place every 15 years. 7 U.S.C.
§ 136a(g)(1)(A)(i), (iii). During registration review, EPA
“evaluate[s] elements of FIFRA 3(c)(5) including the
composition, labeling and other required material (including
studies and other data), risks and benefits of a pesticide, and
incident data or other information relating to its use.”
Pesticides; Procedural Regulations for Registration Review,
65 Fed. Reg. 24,586, 24,587 (Apr. 26, 2000). Based on its
evaluation, which includes public input, 7 U.S.C.
§ 136a(g)(2); 40 C.F.R. §§ 155.25, 155.30, 155.42, EPA can
cancel an existing registration, 7 U.S.C. § 136a(g)(1)(A)(v).

We review EPA’s compliance with these requirements
for “substantial evidence when considered on the record as a
whole.” 7 U.S.C. § 136n(b); Nat. Res. Def. Council v. EPA
(Nanosilver II), 857 F.3d 1030, 1035 (9th Cir. 2017). For
there to be substantial evidence, the administrative record
must show “such relevant evidence as a reasonable mind
might accept as adequate to support a conclusion even if it is
possible to draw two inconsistent conclusions from the
evidence.” Nanosilver II, 857 F.3d at 1036 (internal
quotation marks and citation omitted). This review is
“relatively deferential to the agency factfinder,” but must
still be “searching and careful, subjecting the agency’s
decision to close judicial scrutiny.” Containerfreight Corp.
v. United States, 752 F.2d 419, 422 (9th Cir. 1985) (internal
quotation marks and citation omitted).
 NAT’L FAMILY FARM COALITION V. USEPA 29

A

NRDC first claims EPA incorrectly applied what NRDC
believes is the more lenient “conditional” registration
standard rather than the more stringent “unconditional”
standard when it registered Enlist Duo in 2014. We disagree.

As a preliminary matter, NRDC waived any argument
that EPA applied the incorrect standard when it registered
Enlist Duo in 2014. NRDC did not raise this argument
during the administrative process, or during its first
challenge to the 2014 registration. In fact, NRDC
affirmatively acknowledged that EPA’s 2014 registration of
Enlist Duo was an unconditional registration. NRDC has
therefore waived this particular challenge to the 2014
registration. See N. Plains Res. Council, Inc. v. Surface
Transp. Bd., 668 F.3d 1067, 1081 (9th Cir. 2011) (“Parties
must alert an agency to their position and contentions.”).

Even absent waiver, however, NRDC’s argument that
EPA applied the wrong standard is not persuasive. NRDC’s
primary support for this argument is one line in the 2014
Final Registration Decision citing FIFRA’s conditional
registration provision. But this stray line appears to be a
typographical error. The EPA’s 2014 registration was
plainly unconditional—not conditional. The Notice of
Registration—which is the actual license—states that Enlist
Duo “is unconditionally registered in accordance with
FIFRA section 3(c)(5)” and the “Term of Issuance” is
“Unconditional.” Similarly, the Proposed Decision
Document cites subsection 3(c)(5) and “concludes that . . .
approving this application as set forth below will not cause
any unreasonable adverse effect on the environment,”
language tracking the unconditional registration standard.
30 NAT’L FAMILY FARM COALITION V. USEPA

Moreover, the analysis in the registration documents
does not suggest EPA was applying the more limited
“significantly increase the risk of unreasonable adverse
effect[s]” standard. To the contrary, the documents indicate
EPA applied the broader “cause any unreasonable adverse
effects” standard for unconditional registrations. For
example, EPA did not just analyze how the use of 2,4-D in
Enlist Duo differed from how 2,4-D was being used in
previous pesticides and then address the risks associated
with those new uses. Instead, it analyzed the human and
ecological risks of using 2,4-D. This context supports that
EPA was applying the unconditional registration standard, a
conclusion NFFC Petitioners agreed with in their briefing.

That EPA mentioned outstanding data Dow should
provide to the agency in unconditionally approving the
product does not change that conclusion. True, it is FIFRA’s
conditional registration provision that allows EPA to
approve a product without all relevant data. 7 U.S.C.
§ 136a(c)(7)(A). But that does not mean EPA cannot request
additional data when unconditionally approving a product.
EPA can. When it does so, however, any such data must not
be “necessary to make the determinations” under subsection
(c)(5), 40 C.F.R. § 152.112(c)—that is, not needed to
determine whether the product will “generally cause
unreasonable adverse effects on the environment,” 7 U.S.C.
§ 136a(c)(5)(D). EPA’s decision to unconditionally register
Enlist Duo therefore did not prevent it from requesting
additional data; it merely prevented it from requesting
additional data relevant to whether the unconditional
registration standard was met. Here, NRDC does not argue
that the additional data EPA requested was necessary to that
inquiry.
 NAT’L FAMILY FARM COALITION V. USEPA 31

Even assuming the requested data were necessary to
assessing the risks associated with Enlist Duo, however,
EPA’s unconditional registration would not be transformed
into a conditional one. As discussed above, EPA’s analysis
and the relevant registration documents make clear that EPA
was applying the broader unconditional registration
standard. Thus, to the extent EPA was requesting data
relevant to assessing risk under FIFRA, any such request
only suggests EPA erred in applying the unconditional
registration standard. And no Petitioner argues there was
any such error here. 6

B

NFFC Petitioners argue that EPA incorrectly applied
FIFRA’s “cause any unreasonable adverse effects”
unconditional registration standard in its 2017 registration
decision rather than the “significantly increase the risk of
any unreasonable adverse effect” conditional registration
standard. EPA concedes that it cited the wrong standard but
argues any error is harmless because the standard for
unconditional registration is higher, not lower, than the
standard for conditional registration.

6
We reach a similar conclusion as to NRDC’s argument that EPA
unlawfully imposed “conditions” on its 2014 registration decision. We
are not persuaded that EPA was imposing “conditions” on registration
merely because it used the phrase “provided that” before listing a few
additional requirements for Dow. Nor do we decide whether EPA can
impose “conditions” on unconditional registrations under FIFRA. Cf.
40 C.F.R. § 152.115(c) (authorizing EPA to establish “other conditions
applicable to registration[]” for conditional registrations). Even if EPA
were imposing “conditions” on its unconditional registration, that would
only suggest error in applying that standard—not transform EPA’s
unconditional registration into a conditional one.
32 NAT’L FAMILY FARM COALITION V. USEPA

We agree with EPA. At first blush, the conditional
registration standard appears to impose a higher standard
than the unconditional one. After all, “cause,” as used in the
unconditional registration standard, entails the pesticide
“produc[ing]” harm, Cause, Black’s Law Dictionary (10th
ed. 2014), while “risk,” as used in the conditional standard,
connotes the “possibility of harm,” Risk, Black’s Law
Dictionary (10th ed. 2014). But those words are not the only
clues as to what this statute means. We must also look to the
statute as a whole to find meaningful context. See Exxon
Mobil Corp. v. EPA, 217 F.3d 1246, 1249 (9th Cir. 2000).

That context provides significant guidance here. The
conditional registration standard, with its “risk” language,
only applies when a registrant is proposing to use an already-
registered pesticide or active ingredient in a new way.
7 U.S.C. § 136a(c)(7)(A), (B). That means conditional
registration is an option only when a pesticide or active
ingredient has already been registered using the
unconditional registration, causation-based standard. The
conditional registration standard is therefore best understood
as limiting the scope of new evidence EPA must consider in
making its registration decision, a conclusion Petitioner
NRDC agreed with in its briefing. EPA need only consider
evidence that bears on whether the new or additional use
changes EPA’s original conclusion that the pesticide or
active ingredient will “not generally cause unreasonable
adverse effects.” Id. § 136a(c)(7). Any error by EPA in
citing the more burdensome unconditional registration
standard therefore does not show that EPA lacked substantial
evidence to support its conclusions.

C

Petitioners argue that EPA lacked substantial evidence
for its 2014, 2015, and 2017 registration decisions because
 NAT’L FAMILY FARM COALITION V. USEPA 33

EPA failed to: (1) properly assess harm to monarch
butterflies from increased 2,4-D use on milkweed in target
fields; (2) consider that Enlist Duo would increase the use of
glyphosate over time; (3) correctly consider the volatility of
Enlist Duo’s 2,4-D component; and (4) consider the
synergistic effects of mixing Enlist Duo with glufosinate.
We address each argument in turn.

1

With the approval of Enlist Duo, the use of Enlist Duo’s
2,4-D ingredient will increase. NRDC argues that EPA
failed to consider the harm of expanded 2,4-D use to certain
monarch butterfly habitats and human health.

We decline to address NRDC’s challenge to EPA’s
analysis of human health risks because NRDC failed to
address that argument in its opening brief and therefore
waived it. NRDC’s opening brief states that “Enlist Duo
may pose serious risks to human health” and discusses those
risks in the Statement of the Case but does not meaningfully
address them in any of its briefing. We have regularly held
that “an issue referred to in the appellant’s statement of the
case but not discussed in the body of the opening brief is
deemed waived.” Martinez-Serrano v. INS, 94 F.3d 1256,
1259 (9th Cir. 1996). This case is no exception.

As to the impact on the monarch butterfly population,
EPA did assess some of these risks as part of its registration
decisions. Before the 2017 decision, for example, EPA
performed a risk assessment that considered the “toxic
effects to non-target plants (a grouping that includes plants
important to monarchs).” EPA found “no concerns for
terrestrial invertebrates (including monarchs)” because
Enlist Duo would only affect treated fields—not non-target
plants—as long as it was used under the “conditions
34 NAT’L FAMILY FARM COALITION V. USEPA

prescribed by the label.” These conditions include requiring
a “30 foot downwind buffer (in the direction in which the
wind is blowing)” and a specific, low drift nozzle. These
mitigation measures, among others, will avoid spray drift of
2,4-D to non-target fields, thereby ensuring that no non-
target milkweed is affected. EPA reached similar
conclusions, using similar reasoning, in its 2014 registration
decision.

So far, so good. But NRDC also argues EPA should
have considered how the destruction of milkweed on target
fields would affect monarch butterflies. This argument
carries some force because EPA acknowledged in its
briefing that it did not assess those risks. According to EPA,
it was not required to do so because “farmers will control the
same amount of milkweed on their crop fields through the
use of herbicides or other means and at the same crop growth
stages, with or without Enlist Duo.”

Despite the intuitive appeal of EPA’s argument, we must
reject it. EPA did not assert this rationale as a reason for
declining to assess the destruction of milkweed on target
fields, so neither can we. See Motor Vehicle Mfrs. Ass’n of
U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43
(1983) (“We may not supply a reasoned basis for the
agency’s action that the agency itself has not given.”
(quoting SEC v. Chenery Corp., 332 U.S. 194, 196 (1947))).
After all, “judicial review of agency action is limited to ‘the
grounds that the agency invoked when it took the action.’”
Dep’t of Homeland Sec. v. Regents of the Univ. of Cal.,
140 S. Ct. 1891, 1907 (2020) (quoting Michigan v. EPA,
576 U.S. 743, 758 (2015)). Moreover, even had EPA
asserted such a rationale, it would likely be premised on
legal error. That milkweed would likely be targeted in the
same ways even absent Enlist Duo’s registration suggests
 NAT’L FAMILY FARM COALITION V. USEPA 35

that registering Enlist Duo may not be “unreasonable” under
FIFRA. 7 U.S.C. § 136a(c)(5), (c)(7). But it says nothing
about whether an effect would be “adverse.” Id. Given the
record evidence suggesting monarch butterflies may be
adversely affected by 2,4-D on target fields, EPA was
required, under FIFRA, to determine whether any effect was
“adverse” before determining whether any effect on the
environment was, on the whole, “unreasonable.” EPA’s
failure to do so means that its decision was lacking in
substantial evidence on this issue.

2

NRDC also challenges EPA’s conclusion that Enlist
Duo’s glyphosate ingredient “would not cause unreasonable
adverse effects on the environment” because glyphosate was
already being used in the same locations and doses and on
the same crops. According to EPA, Enlist Duo’s registration
“would only impact which glyphosate product was used”—
not how much glyphosate was used.

NRDC contends that FIFRA does not allow EPA to take
this ingredient-by-ingredient approach. But as discussed
above, FIFRA allows an ingredient-by-ingredient approach.
40 C.F.R. § 152.111. For cases involving “active
ingredient[s]” that have already been registered—like
glyphosate—EPA need not perform a complete review of the
data. Id. “Instead, the Agency will review the application
using the criteria for conditional registration” under 7 U.S.C.
§ 136a(c)(7)(A) and (B), id., to determine whether the
proposed use is “identical or substantially similar” to a prior
use and whether the proposed use will “significantly increase
the risk of any unreasonable adverse effect on the
environment,” 7 U.S.C. § 136a(c)(7)(A).
36 NAT’L FAMILY FARM COALITION V. USEPA

Here, there is no such increase in the risk of unreasonable
adverse effects because substantial evidence supports EPA’s
conclusion that neither the initial 2014 registration of Enlist
Duo—nor the subsequent approvals for new uses—will
increase the overall use of glyphosate. That is because corn,
cotton, and soybean crops have long been genetically
engineered to be glyphosate resistant, meaning that the use
of glyphosate on these crops was nearly ubiquitous before
Enlist Duo was registered in 2014. Indeed, “during the 2010
to 2014 period,” 80 to 85 percent of corn was “treated one or
more times with glyphosate.” The same is true of soybean
and cotton. During the same period, “approximately
95 percent” of soybean was treated one or more times and
between “75 and 90 percent” of cotton was sprayed with
glyphosate at least once. Even absent Enlist Duo’s
registration, therefore, farmers would continue to use
glyphosate on these same crops. Thus, there was no
increased risk of unreasonable adverse effects caused by
glyphosate in approving Enlist Duo. 7

This does not mean, of course, that new data about
glyphosate will go unconsidered. If a proposed new use in a
future registration will, unlike in this case, “significantly
increase the risk of any unreasonable adverse effect on the
environment,” EPA must consider those risks. 7 U.S.C.
§ 136a(c)(7)(A). Or, in the alternative, EPA can cancel a
registration if it determines the pesticide “generally causes
unreasonable adverse effects on the environment.” 7 U.S.C.

7
NRDC argues that, at a minimum, the 2015 and 2017 amendments
to the 2014 registration increase the overall use of glyphosate because
they allow for use in new states and on new crops. But this argument
fails for the same reason discussed above. The 2015 and 2017
amendments may increase the overall use of Enlist Duo relative to the
2014 registration. But that does not mean these amendments will
increase the overall use of glyphosate.
 NAT’L FAMILY FARM COALITION V. USEPA 37

§ 136d(b). And if it does not do so, the “registration review”
process serves as a backstop to ensure that pesticides do not
remain registered once new data has shown them to be
harmful to humans or the environment. 7 U.S.C. § 136a(g). 8

3

Third, NFFC Petitioners contend that EPA failed to
properly consider 2,4-D’s volatility—that is, its tendency to
evaporate into a gas and drift to non-target plants. Here, too,
we disagree.

In its 2014 and 2017 registration decisions, EPA
concluded that the 2,4-D in Enlist Duo—a choline salt
variety distinct from the 2,4-D used in many other pesticide
products—exhibited lower volatility and off-site vapor drift
than other registered forms of 2,4-D. EPA began with the
“Ouse” laboratory study. The Ouse study was designed to
examine the degree of visual damage or injury—like
cupping of leaves or twisting of foliage—caused by
exposure to vapors of 2,4-D. Soybean, tomato, grape, and
cotton plants were exposed to varying doses of 2,4-D vapors
for various lengths of time. Researchers then observed the
level of visible injury each plant exhibited based on the dose
administered. Using the data gathered from these

8
NRDC also argues that EPA’s decision is unsupported by
substantial evidence because once EPA determined in 2017 that there
was “outstanding data” about 2,4-D that would preclude an
unconditional registration, it was “barred from reissuing its earlier
[unconditional] approvals for Enlist Duo.” But the 2017 license—which
is the operative license approving Enlist Duo for use—approves Enlist
Duo “conditionally” because of the “outstanding data.” So although
EPA’s 2017 registration decision reissued its 2014 and 2015 decisions,
it did not reissue an unconditional license for the uses of Enlist Duo
approved in 2014 and 2015.
38 NAT’L FAMILY FARM COALITION V. USEPA

observations, the study concluded that grapes were the most
sensitive to 2,4-D vapor, followed by cotton, tomatoes, and
soybeans. Further, a “dose-response curve” indicated the
level of visual plant injury that would occur at a particular
dose.

Recognizing that this study did not meet the regulatory
measure for assessing plant damage—plant growth or
survival—EPA relied on six publicly available studies that
assessed the relationship between visual damage to plants
and effects on plant growth and survival. According to these
studies, for example, 20% visual damage to grape plants
resulted in decreases in plant growth in grapes. Cotton and
soybean had a much higher threshold. For those plants,
visual damage between 35 and 66% resulted in decreases in
plant growth and yield.

EPA also relied on another vapor flux study called the
“Havens” study. Plants were placed directly on fields and
also 5 and 15 meters away from treated fields. Those fields
were treated with a higher dose of 2,4-D than the label would
allow, and researchers measured outward signs of damage to
the plants. Only plants on treated fields showed growth or
survival damage; even plants located within 5 meters of the
treated fields did not.

EPA used the most conservative estimates from this data
to perform computer modeling. Specifically, EPA used the
data’s finding that growth or survival occurred at statistically
significant levels when there was a minimum of 20% visual
plant damage and the dose-response curve from the Ouse
study to find the dose level of 2,4-D that would produce 20%
visual plant damage. That dose level was 1.9 ug/m3/hour—
which measures the mass of pesticide in a cubic meter of air
space to which an organism is exposed over a one-hour time
period.
 NAT’L FAMILY FARM COALITION V. USEPA 39

EPA then used this dose level to predict the air
concentration of 2,4-D that would be expected at the edge of
a field and various distances beyond. The results showed
that the air concentrations of 2,4-D at the edge of a treated
field were below 1.9 ug/m3/hour, the threshold for what
might cause 20% of visual plant damage thereby affecting
plant growth or survival. The model therefore predicted no
adverse damages to plants off-field.

EPA also relied on two additional data points. First, EPA
performed AERSCREEN modeling, which assesses drift of
wet and dry depositions of the pesticide. That modeling
showed negligible risk. EPA also relied on atmospheric
monitoring data showing that negligible amounts of 2,4-D
were detected in rainwater samples and air samples. These
data points provided further evidence to support EPA’s
conclusion that the choline salt form of 2,4-D in Enlist Duo
is less prone to volatilization than other forms of 2,4-D.

NFFC Petitioners claim that EPA’s conclusion is based
on the flawed underlying Ouse study, which EPA conceded
was limited due to its methodology and “was not well-
aligned with the 850.4150 protocol.” But using a limited
study does not make EPA’s actions lacking in substantial
evidence as long as EPA “acknowledge[s] the limitations”
and does not “rely solely upon [the study’s] conclusions.”
Cent. Ariz. Water Conservation Dist. v. EPA, 990 F.2d 1531,
1543 (9th Cir. 1993) (holding that EPA’s reliance on a
“seriously flawed study” did not undermine EPA’s overall
conclusions); see also Ctr. for Biological Diversity v. Esper,
958 F.3d 895, 910–11 (9th Cir. 2020) (finding substantial
evidence in support of a Department of Defense conclusion
relying on studies the Department itself “criticized” as not
“rigorous,” “extremely poorly-done,” and “not withstanding
scientific scrutiny”).
40 NAT’L FAMILY FARM COALITION V. USEPA

NFFC Petitioners also argue that EPA requested a new
study to replace the Ouse study but failed to wait for that new
study before rendering its decision. Neither premise is
correct. EPA did not request a “replacement” study. It
recommended an additional study—that is, a “vapor-phase
study with vegetative vigor endpoints” that would “further
characterize the risk to plants from” exposure to 2,4-D
vapor. EPA then relied on that study—the Havens study—
in arriving at its conclusion about 2,4-D. 9

NFFC Petitioners also criticize the studies EPA relied on
in several other ways. We address each criticism in turn.

First, NFFC Petitioners claim that the Ouse study is too
different from the other six studies because the grape plants
used in the six studies were of a different age and species
than those used in the Ouse study. This difference matters,
according to NFFC Petitioners, because the threshold level
of harm for a plant is dependent on its species and growth
stage. But Petitioners have not pointed to any record
evidence showing that the species and age of the grape plants
affected the validity of the six studies. Without such
evidence, we cannot conclude that EPA lacked substantial
evidence to support its conclusion.

Second, NFFC Petitioners argue that EPA’s conclusion
about a 20% visual damage threshold was contradicted by
EPA scientists, who concluded that the damage threshold
was actually 5%. But EPA did not contradict itself. EPA
summarized the Ouse study’s conclusion that the harm
9
NFFC Petitioners claim the Havens study, which was a non-
Guideline 850.4150-compliant field study, was not the study EPA
requested. But EPA never requested that the recommended study
comply with particular test guidelines or be performed in a laboratory.
The Havens study was the requested study.
 NAT’L FAMILY FARM COALITION V. USEPA 41

threshold was 5%. But then EPA analyzed additional data
and determined that growth or survival occurred at
statistically significant levels when 20% visual damage
could be observed.

Third, NFFC Petitioners posit that the six studies relied
on by EPA do not satisfy the regulatory guidelines in the
vegetative vigor test—which look to survival, height, and
biomass—because the studies examine yield and growth,
which are not endpoints listed in the guidelines. But EPA is
not required to follow the regulatory guideline NFFC
Petitioners cite. See OCSPP 850.4150, at i (Jan. 2012),
(“[T]hese guidelines are not binding on either EPA or any
outside parties, and the EPA may depart from the guidelines
. . . .”). Moreover, the studies closely tracked the specific
endpoints in the guidelines. Examining visual growth and
yield directly addresses whether plants are growing (height
and biomass) and flourishing (survival). EPA’s decision to
rely on studies that do not precisely track a regulatory
guideline therefore does not undermine its decision. EPA
“may apply [its] expertise to draw conclusions from . . .
probative preliminary data not yet certifiable as fact.” Cent.
Ariz. Water, 990 F.2d at 1543 (internal quotation marks and
citation omitted).

Fourth, NFFC Petitioners argue EPA’s computer
modeling considered field sizes much smaller than the
average corn, cotton, or soybean field even though vapor
drift increases with the size of a sprayed field. But
Petitioners provide no reason why EPA could not extrapolate
from its modeling. “The Administrator may apply his
expertise to draw conclusions from . . . theoretical
projections from imperfect data.” Cent. Ariz. Water,
990 F.2d at 1543 (internal quotation marks and citation
omitted). Moreover, NFFC Petitioners’ argument ignores
42 NAT’L FAMILY FARM COALITION V. USEPA

the AERSCREEN modeling EPA performed, which
provides further evidence supporting EPA’s volatilization
conclusions.

Ultimately, EPA’s evaluation of 2,4-D volatility
probably could have been better. But it is not our role to
second-guess EPA’s conclusion. Moreover, there is no
evidence in the record that its conclusion was wrong.
Petitioners do not suggest that, in the five-plus years since
Enlist Duo was originally approved, their fears surrounding
2,4-D volatility have materialized in the real world. 10 We
therefore hold that a “reasonable mind might accept” the
studies on which EPA relied “as adequate to support a
conclusion” that the volatility of 2,4-D choline salt will not
cause unreasonable adverse effects on the environment.
Nanosilver II, 857 F.3d at 1036 (citation omitted).
Accordingly, substantial evidence supports EPA’s findings.

4

Finally, NFFC Petitioners contend EPA should have
accounted for the potential synergistic effect of mixing
Enlist Duo with a different chemical called glufosinate.
According to them, Dow intends to mix glufosinate with
Enlist Duo.

NFFC Petitioners’ concern about mixing Enlist Duo with
glufosinate is speculative. Nothing in the record suggest that

10
NFFC Petitioners point to generic data on 2,4-D drift from the
1970s through early 2000s. This data mainly addresses spray drift—not
drift from volatilization—and does not address 2,4-D choline salt. Even
assuming the data is relevant to volatility here, EPA was aware of the
general risk of 2,4-D volatility. EPA nonetheless concluded, relying on
the above studies, that the choline salt variety of 2,4-D was less prone to
volatilization than other forms of 2,4-D.
 NAT’L FAMILY FARM COALITION V. USEPA 43

such mixing has occurred in the five-plus years since Enlist
Duo was first registered. Indeed, EPA has stated that Enlist
Duo cannot be tank-mixed with any product that has not
been tested, approved, and listed on the website
EnlistTankMix.com. And no product containing glufosinate
is listed on that website. It is therefore currently unlawful to
mix Enlist Duo with glufosinate.

NFFC Petitioners point to Dow’s now-abandoned patent
application, which claims synergism between 2,4-D and
glufosinate, and notes that crops genetically engineered to
withstand Enlist Duo are also designed to withstand
glufosinate. But FIFRA only requires EPA to consider the
uses of the pesticide contemplated by the label. See 7 U.S.C.
§ 136a(c)(5). That includes currently planned tank mixing.
It does not include theoretical tank mixing—that is, tank
mixing that might occur at a future date. NFFC Petitioners
may separately challenge any future EPA final action
approving this potential tank mixing.

NFFC Petitioners argue that the current registration is its
only opportunity to challenge the potential mixing of Enlist
Duo and glufosinate as having an unreasonable adverse
effect on the environment because the Enlist Duo label
allows tank mixing to be approved as long as mixing is
shown “not to adversely affect” “spray drift properties” and
does not require testing on synergy. EPA’s decision not to
require testing for potential synergy may appear to provide a
loophole for ingredients to be mixed long after initial review.
But this decision reflects EPA’s broader stance on synergy.
Following the recommendation of the National Research
Council, EPA “views synergism to be a rare event” and
assumes that the components of pesticide products will not
have “synergistic effects.” EPA’s assumption is not
undermined by the limited evidence NFFC Petitioners cite
44 NAT’L FAMILY FARM COALITION V. USEPA

supposedly showing synergism between glufosinate and
Enlist Duo. And nothing prevents a Petitioner from
approaching the EPA with concerns about synergy in the
future.

Moreover, FIFRA does not prohibit EPA from
undertaking review of a pesticide product whenever it
becomes necessary. 7 U.S.C. § 136a(g). According to EPA,
this “means that the Agency must continue to respond to
emerging risk concerns and not defer action until a
pesticide’s regularly scheduled registration review.”
Pesticides; Procedural Regulations for Registration Review,
70 Fed. Reg. 40,251-01, 40,270 (July 13, 2005). What’s
more, EPA can cancel a registration, change a classification,
or amend a label at any time if it determines that the
registration, as constituted, “generally causes unreasonable
adverse effects on the environment.” 7 U.S.C. § 136d(b).
Finally, there are opportunities for NFFC Petitioners to
provide input during the registration review process,
40 C.F.R. §§ 155.25, 155.30, 155.42, which is currently
ongoing for both 2,4-D and glyphosate. Petitioners’ claim
that they have no remedy for potential future violations of
FIFRA involving glufosinate therefore rings hollow.

* * *

Almost all of EPA’s registration decisions are supported
by substantial evidence because the record evidence was of
the type that “a reasonable mind might accept as adequate to
support a conclusion even if it is possible to draw two
inconsistent conclusions from the evidence.” Nanosilver II,
857 F.3d at 1036 (internal quotation marks and citation
omitted). However, as discussed above, EPA failed to
consider risks to monarch butterflies caused by the treatment
of milkweed on target fields. As to FIFRA, therefore, we
 NAT’L FAMILY FARM COALITION V. USEPA 45

grant NRDC’s petition for review in part and deny it in part.
We discuss the remedy for this partial grant below.

IV

To protect endangered or threatened species, the ESA
sets forth legal requirements with which federal agencies
must comply. See 16 U.S.C. § 1531. “Each Federal agency
shall, in consultation with and with the assistance of the
Secretary, insure that any action authorized, funded, or
carried out by such agency” “is not likely to jeopardize the
continued existence of any endangered species . . . or result
in the destruction or adverse modification of habitat of such
species.” 16 U.S.C. § 1536(a)(2).

The ESA and its implementing regulations delineate a
process—known as Section 7 consultation—for determining
the biological impacts of a proposed action. 16 U.S.C.
§ 1536. The process starts with two possible roads, which
turn on whether the proposed action will have “no effect” or
if it “may affect” listed species or critical habitat. If a listed
species is outside the proposed “[a]ction area”—that is, it
will not be “affected directly or indirectly by the Federal
action,” 50 C.F.R. § 402.02—it will, by definition, not be
affected by the proposed action and consultation is not
required. Similarly, if the action agency finds “that its action
will have no effect on listed species or critical habitat” even
within the “action area,” it need not consult with the expert
agencies, FWS or the National Marine Fisheries Service
(“NMFS”). California ex rel. Lockyer v. U.S. Dep’t of
Agric., 575 F.3d 999, 1019 (9th Cir. 2009). If, however, the
action agency’s proposed action “may affect”—that is,
might have “[a]ny possible effect, whether beneficial,
benign, adverse or of an undetermined character,” Karuk
Tribe of Cal. v. U.S. Forest Serv., 681 F.3d 1006, 1027 (9th
Cir. 2012) (internal quotation marks and citation omitted)—
46 NAT’L FAMILY FARM COALITION V. USEPA

on a “listed species or critical habitat,” consultation is
required, 50 C.F.R. § 402.14(a). In determining whether to
consult, the action agency must use “the best scientific and
commercial data available.” 16 U.S.C. § 1536(a)(2).

Consultation also arises in the context of “critical
habitat” determinations. Under the ESA, “critical habitat” is
“the specific areas within the geographical area occupied by
the species . . . on which are found those physical or
biological features (I) essential to the conservation of the
species and (II) which may require special management
considerations or protection.” Id. § 1532(5)(A)(i).
“[C]ritical habitat” may also include “specific areas outside
the geographical area occupied by the species,” but only
“upon a determination by [the Services] that such areas are
essential for the conservation of the species.” Id.
§ 1532(5)(A)(ii). In either case, the critical habitat must
contain physical or biological features “essential” to the
species, id. § 1532(5)(A)(i)–(ii)—features known as
“primary constituent elements or PCEs,” Alaska Oil & Gas
Ass’n v. Jewell, 815 F.3d 544, 555 (9th Cir. 2016) (internal
quotation marks omitted) (citing 50 C.F.R. § 424.12(b)(5)).
The consultation agencies define, by regulation, critical
habitat and the corresponding PCEs for endangered species.
16 U.S.C. § 1533(a)(3)(A); 50 C.F.R. § 17.95. The action
agency’s role is to determine whether a proposed action
modifies or affects these critical habitats. 16 U.S.C.
§ 1536(a)(2).

Mitigation measures are frequently adopted as part of
ESA compliance. E.g., Defs. of Wildlife v. Zinke, 856 F.3d
1248, 1258 (9th Cir. 2017). Such mitigation measures are
permissible as long as they are the result of “specific and
binding plans.” Id. (citation omitted). Mitigation measures
must also be “reasonably certain to occur.” Nat’l Wildlife
 NAT’L FAMILY FARM COALITION V. USEPA 47

Fed’n v. Nat’l Marine Fisheries Serv., 524 F.3d 917, 936
n.17 (9th Cir. 2008); see also Sierra Club v. Marsh, 816 F.2d
1376, 1388 (9th Cir. 1987) (consultation should occur if
“mitigation efforts” “have been delayed,” “may not take
place at all,” or are otherwise ineffective), abrogated on
other grounds as recognized in Cottonwood, 789 F.3d
at 1088–91.

Because the ESA does not specify a standard of review,
we review EPA’s compliance under the APA and uphold
agency action unless it is arbitrary, capricious, an abuse of
discretion, or contrary to law. Or. Nat. Res. Council v. Allen,
476 F.3d 1031, 1036 (9th Cir. 2007); 5 U.S.C. § 706(2)(A).
An agency decision is arbitrary or capricious “only if the
agency relied on factors Congress did not intend it to
consider, entirely failed to consider an important aspect of
the problem, or offered an explanation that runs counter to
the evidence before the agency or is so implausible that it
could not be ascribed to a difference in view or the product
of agency expertise.” Lands Council v. McNair, 629 F.3d
1070, 1074 (9th Cir. 2010) (internal quotation marks and
citation omitted). Here, EPA’s “no effect” findings, decision
about the scope of the “action area,” and “critical habitat”
determinations survive this deferential review.

A

NFFC Petitioners primarily challenge EPA’s “no effect”
findings for plants and animals as legally erroneous. We
disagree.

In making its “no effect” findings, EPA used “risk
quotients” and interpretative “levels of concern” developed
as part of compliance with FIFRA, but applied much more
conservative assumptions. The “risk quotients” are
calculated by estimating the amount of exposure to the
48 NAT’L FAMILY FARM COALITION V. USEPA

pesticide. That estimate is then divided by established acute
and chronic ecotoxicity levels for specific classes of plants
and animals—for example, aquatic animals and terrestrial
mammals and birds. The calculated risk quotient is then
compared to EPA’s “levels of concern.” If the risk quotient
does not exceed the “levels of concern,” EPA determines
there will be “no effect” on the listed species. If, however,
the risk quotient exceeds the “level of concern” for acute or
chronic exposure, EPA conducts a refined, species-specific
assessment before making a “no effect” finding.

This methodology applies the correct legal standard.
The “may affect” standard is only met—triggering
consultation—if there is “[a]ny possible effect, whether
beneficial, benign, adverse or of an undetermined character.”
Karuk Tribe, 681 F.3d at 1027 (quoting Lockyer, 575 F.3d at
1018 (quoting Interagency Cooperation―Endangered
Species Act of 1973, as Amended; Final Rule, 51 Fed. Reg.
19,926, 19,949 (June 3, 1986))). And EPA’s risk
quotient/level of concern methodology found that there
would be no such effects. True, EPA concluded that
protected species and critical habitats would be exposed to
potentially harmful chemicals. But it concluded that any
such exposure would have “no effect” on listed species and
habitats. EPA’s recognition of exposure is not a recognition
that Enlist Duo “may affect” protected species and critical
habitats. It is a recognition that EPA did what the ESA
requires it to do: assess risks to determine whether the
exposure would have “any possible effect.” Id.

Our decision in Friends of Santa Clara River v. United
States Army Corps of Engineers supports this conclusion.
887 F.3d 906, 915, 925–26 (9th Cir. 2018). In that case,
steelhead salmon would be exposed to copper concentrations
as a result of the discharge of dredged or fill material in the
 NAT’L FAMILY FARM COALITION V. USEPA 49

Santa Clara River. Id. at 910, 914–15. But that did not mean
the relevant agency was required to consult with NMFS. Id.
at 915, 924–26. Instead, the agency could reach its own “no
effect” conclusion—that the amount of copper to which the
salmon would be exposed was within a normal range—
without consultation. Id. at 925–26. EPA’s conclusions
here—which recognize potential exposure but nonetheless
conclude there will be “no effect”—are no different.

This conclusion is not altered by the fact that EPA’s
preliminary risk assessments—which relied on conservative
assumptions—found a chance that Enlist Duo “may affect”
hundreds of protected species. Those assessments were just
that—preliminary. EPA did a more refined, species-specific
assessment for the species initially believed to be at risk.
After that assessment, EPA concluded that none of the
species believed to be at risk after the initial assessment were
in fact at risk. EPA therefore made “no effect” findings for
those species. Nothing about this iterative process suggests
EPA’s ultimate “no effect” findings were arbitrary,
capricious, or contrary to law.

Nor did EPA’s adoption of mitigation measures,
including a 30-foot downwind buffer and certain label
restrictions, to reach a “no effect” finding as to plants and
animals off the treated field render EPA’s conclusions
arbitrary, capricious, or contrary to law. Mitigation
measures are frequently adopted to avoid effects on listed
species or habitats. See Ctr. for Biological Diversity v. U.S.
Dep’t of Interior, 563 F.3d 466, 482 (D.C. Cir. 2009) (noting
that “satisfaction of the ESA mandate that no endangered life
be jeopardized must be measured in view of the full
contingent of . . . checks and balances and all mitigating
measures adopted in pursuance thereof” (internal quotation
marks and citation omitted)).
50 NAT’L FAMILY FARM COALITION V. USEPA

To be sure, mitigation measures that merely “reduce,”
but cannot scientifically “eliminate” an “effect” probably
compel a “may affect” finding. Karuk Tribe, 681 F.3d
at 1028. In Karuk Tribe, gold miners argued that mitigation
measures taken by the action agency showed that there
would be “no effect” on threatened species although the
agency never made a “no effect” finding. Id. We held that
the mitigation measures “cut[] against” a “no effect” finding
because they merely “reduce[d]” but could “not eliminate”
the impact to threatened species. Id.

Here, by contrast, EPA was able to rule out any effect on
plants and species off the treated field in partial reliance on
mitigation measures. EPA applied the correct legal standard
and supported its conclusions, as discussed in greater detail
below. Under these circumstances, EPA’s use of mitigation
measures is not evidence of a required “may affect” finding
as in Karuk Tribe. Instead, the mitigation measures are
reasonable under the ESA when they are, like the label
restrictions here, “specific and binding plans.” E.g., Defs. of
Wildlife, 856 F.3d at 1258; see also Selkirk Conservation All.
v. Forsgren, 336 F.3d 944, 954–56 (9th Cir. 2003).

Nor did EPA’s use of the risk quotient/level of concern
methodology for its ESA analysis violate its statutory duty
to “use the best scientific and commercial data available.”
16 U.S.C. § 1536(a)(2). The purpose of this requirement “is
to ensure that the ESA not be implemented haphazardly, on
the basis of speculation or surmise.” Bennett, 520 U.S. at
176. Under this standard, the agency must not “disregard
available scientific evidence that is in some way better than
the evidence it relies on.” San Luis & Delta-Mendota Water
Auth. v. Locke, 776 F.3d 971, 995 (9th Cir. 2014) (alterations
adopted and citation omitted). “On the other hand, where
the information is not readily available, we cannot insist on
 NAT’L FAMILY FARM COALITION V. USEPA 51

perfection.” San Luis & Delta-Mendota Water Auth. v.
Jewell, 747 F.3d 581, 602 (9th Cir. 2014). As a result, the
standard does not “require an agency to conduct new tests or
make decisions on data that does not yet exist.” Locke,
776 F.3d at 995. Nor does it allow us to second guess the
agency’s decisions using our own judgment. Because “what
constitutes the best scientific and commercial data available
is itself a scientific determination,” id., it “belongs to the
agency’s special expertise and warrants substantial
deference,” Santa Clara River, 887 F.3d at 924 (internal
quotation marks and citation omitted). A court should
therefore “be especially wary of overturning such a
determination on review.” Locke, 776 F.3d at 995 (citation
omitted).

We cannot overturn EPA’s scientific determination here.
True, a 2013 National Academy of Sciences (“NAS”) report
called the risk quotient/levels of concern methodology relied
on by EPA “not scientifically defensible” if the goal “is to
base a decision on the probabilities of various possible
outcomes.” NAS also recommended that EPA adopt a
“probabilistic approach” to assessing risk to endangered
species. But that same report recognized that the data
needed to adopt the recommended approach was not readily
available. Instead, EPA would be required to generate new
data, a process that would require the “integration of the
uncertainties (from sampling, natural variability, lack of
knowledge, and measurement and model error) into the
exposure and effects analyses by using probability
distributions.” Those distributions would then be
“integrated mathematically to calculate the risk as a
probability.” In short, this new process and generation of
data would require a “transition” that could use some
available scholarship as a model, but would nonetheless
involve significant changes to EPA’s risk quotient/level of
52 NAT’L FAMILY FARM COALITION V. USEPA

concern approach and the data on which EPA relied. That is
why NAS expressly recognized that EPA would not be able
to begin implementation of the new approach on many
pesticide registrations immediately.

EPA therefore did not reject better data that was “readily
available” in registering Enlist Duo using the risk
quotient/level of concern approach. Jewell, 747 F.3d at 602.
Instead, it elected to continue applying that approach while
it put a system in place to use NAS’s proposed approach, as
set forth in the Interim Report to Congress EPA and the
consultation agencies sent to Congress in November 2014. 11
In that Report, EPA and the consultation agencies agreed
that they would implement NAS’s proposed approach in
stages. EPA and the consultation agencies agreed that
NAS’s proposal would not be applied to all pesticide
registration decisions, including specifically the registration
of Enlist Duo. Instead, the agencies elected to focus
application of NAS’s new methodology on pesticides that
were at the time subject to nationwide litigation. The
agencies also highlighted their efforts to implement the new
approach up until the Report was submitted.

EPA and the consultation agencies therefore specifically
agreed that the risk quotient/levels of concern approach
could be used for the registration process for Enlist Duo
while EPA began implementing NAS’s new approach. EPA
and the consultation agencies noted that the risk
quotient/level of concern approach was “highly

11
Although the Interim Report is not in the administrative record,
we can consider it “for the limited purpose[] of reviewing [Petitioners’]
ESA claim.” W. Watersheds Project v. Kraayenbrink, 632 F.3d 472, 497
(9th Cir. 2011).
 NAT’L FAMILY FARM COALITION V. USEPA 53

conservative” and “will be protective of non-target species,
including endangered species.”

This interagency agreement does not absolve EPA of its
duty to “use the best scientific and commercial data
available.” 16 U.S.C. § 1536(a)(2). But it does show that
EPA’s actions in implementing NAS’s proposed approach
have been reasonable and protective of endangered species.
After all, the consultation agencies EPA would have been
required to consult had there been a “may affect” finding
have recognized as much. Ultimately, however, we ground
our decision in the fact that the best-scientific-data-available
requirement “does not require the agency to conduct new
tests or make decisions on data that does not yet exist.” Ctr.
for Biological Diversity v. U.S. Fish & Wildlife Serv.,
807 F.3d 1031, 1047 (9th Cir. 2015) (internal quotation
marks and citation omitted). EPA’s determination that this
well-established rule applied here is ultimately a scientific
judgment that we will not overturn. See Locke, 776 F.3d
at 995. 12 It is also one we do not expect to reoccur given
EPA’s commitment to gather the data necessary to
implement NAS’s new methodology going forward.

Finally, we are not persuaded that EPA’s “no effect”
conclusions are rendered arbitrary, capricious, or contrary to
law because EPA relied in part on the 1993 Wildlife
Exposure Factors Handbook. EPA relied on the 1993
Handbook for its ESA level assessment to measure how 2,4-

12
NFFC Petitioners and the dissent focus in large part on the NAS
approach being a superior methodology. But the statutory requirement
EPA must comply with is about using the “best scientific and
commercial data available.” 16 U.S.C. § 1536(a)(2) (emphasis added).
That is why we have focused on whether the data EPA needed was
available—not whether the NAS methodology is superior to the one EPA
used.
54 NAT’L FAMILY FARM COALITION V. USEPA

D within Enlist Duo might affect protected species’ food and
water consumption within the action area. According to
NFFC Petitioners, that 1993 Handbook is not the “best
scientific and commercial data available,” 16 U.S.C.
§ 1536(a)(2), because it was intended only for screening
assessments—not ESA level risk assessments—and never
mentions several of the protected species at issue.

But again, “[t]he determination of what constitutes the
best scientific data available belongs to the agency’s special
expertise and warrants substantial deference.” Santa Clara
River, 887 F.3d at 924 (internal quotation marks and citation
omitted). Such deference is warranted for two reasons.
First, the Handbook appears to contemplate EPA’s
approach. To be sure, some portions of the Handbook
suggest it is intended for “screening-level risk assessments.”
But other portions acknowledge it should be used in support
of “assessments for species of concern in a risk assessment,”
like the one here, for “endangered and threatened species.”
Second, although the Handbook does not include
consumption rates for some of the species at issue, it allows
for the use of equations to calculate the consumption rates of
species for which no measurement exists. And EPA used
these equations here. EPA therefore did not act arbitrarily in
partially relying on the 1993 Handbook to make “no effect”
findings. 13

13
NFFC Petitioners argue for the first time in their supplemental
briefs that EPA was required to perform an ESA-level analysis and make
“no effect” findings for glyphosate as well. We decline to address this
argument because it was not raised in Petitioners’ initial opening brief
even though it was directly relevant to the 2017 registration decision.
See Martinez-Serrano, 94 F.3d at 1259–60.
 NAT’L FAMILY FARM COALITION V. USEPA 55

B

NFFC Petitioners also argue that EPA’s rationale for
limiting the “action area” to the treated field was not sound.
We accord deference to EPA in the way it chose to define
the action area. See Friends of the Wild Swan v. Weber,
767 F.3d 936, 950 (9th Cir. 2014) (“The choice of
appropriate action areas requires application of scientific
methodology and, as such, is within the agency’s
discretion.” (internal quotation marks and citation omitted)).

NFFC Petitioners first argue that because 2,4-D is known
to drift beyond treated fields, the “action area”—that is, the
area “to be affected directly or indirectly by the Federal
action,” 50 C.F.R. § 402.02—cannot be limited to the treated
fields. But EPA accounted for this risk as to spray drift by
including the mitigation measure of a 30-foot buffer zone
and other label restrictions, including a prohibition on aerial
application and specific nozzle, temperature, and wind speed
requirements. And its decision to impose these measures
was not based on a mere “assumption that spray drift will
stop at field boundaries” as long as those measures were in
place. EPA performed its own evaluation of the risk of spray
drift using “empirical data” that measured spray drift
deposition rates when Enlist Duo was used according to its
label restrictions. EPA then compared this data to its effect
thresholds for plants and animals off the treated field and
concluded that any exposure to Enlist Duo off the treated
field would “be below effects thresholds.” 14

14
To the extent NFFC Petitioners argue that EPA only made “no
adverse damage”—not “no effect”—findings with respect to volatility,
they are wrong. EPA specifically found, based on “spray drift mitigation
label requirements and analyses of volatility and runoff loadings” that
56 NAT’L FAMILY FARM COALITION V. USEPA

EPA did not violate the “no effect” standard in
performing this analysis. As discussed above, a recognition
of “exposure” is not the same as a recognition of an “effect.”
Nor is the use of mitigation measures like label restrictions
legally erroneous, as long as the measures themselves are not
arbitrary or capricious. See Marsh, 816 F.2d at 1388
(consultation should occur if “mitigation efforts” “have been
delayed,” “may not take place at all,” or are otherwise
ineffective). No such error is evident here. EPA had good—
and science-based—reasons for limiting the action area to
the treated field. And NFFC Petitioners have not pointed to
any record evidence—such as data undermining EPA’s
scientific conclusion or showing that mitigation measures
are not working—suggesting that the mitigation measures
EPA selected are not “specific and binding” and “reasonably
certain to occur.” Nat’l Wildlife Fed’n, 524 F.3d at 936 n.17;
see also Managed Pharmacy Care v. Sebelius, 716 F.3d
1235, 1244 (9th Cir. 2013) (placing the burden on the party
challenging agency action to show that the action was
arbitrary, capricious, an abuse of discretion, or contrary to
law). 15 We therefore cannot conclude there is not “a rational
connection between the facts [EPA] found and the choices

“any risks of effects to non-target organisms will be confined to the
treated field.”
15
Petitioners’ argument that EPA failed to consider whether ESA-
protected plants and animals off the treated field would be indirectly
affected by the on-field treatment is not persuasive for a similar reason.
EPA considered indirect effects and Petitioners have not cited any
evidence contradicting EPA’s conclusions.
 NAT’L FAMILY FARM COALITION V. USEPA 57

[it] made.” Arrington v. Daniels, 516 F.3d 1106, 1112 (9th
Cir. 2008) (internal quotation marks and citation omitted).16

C

Finally, NFFC Petitioners argue EPA violated its duty to
insure no “adverse modification” of “critical habitat” by
relying on its 2016 risk assessment. EPA used an iterative
approach to make this determination. First, EPA determined
that critical habitats had been designated for 184 of the 531
species in the states proposed for Enlist Duo registration.
EPA then determined that because 176 of those species were
not found on corn, cotton, or soybean fields, Enlist Duo’s
registration would not modify their critical habitats or any
PCEs. That left eight species with critical habitat
designations that did use corn, cotton, or soybean fields.
EPA then considered whether Enlist Duo “may affect” those
eight species or their PCEs. Because none of those eight
species’ critical habitats contained PCEs “related to
agriculture,” EPA concluded that there would be no
“modification” to their critical habitats.

This methodology did not misapply the “may affect”
standard as to critical habitats. EPA explicitly considered
whether there would be any effect on “one or more of the
16
Petitioners’ claims of error as to the whooping crane and Indiana
bat—which rely on outdated species-specific analysis and selective
quotation of the record—are based largely on arguments we have already
rejected. EPA’s recognition that the whooping crane and Indiana bat
would be exposed to Enlist Duo did not necessitate a “may affect”
finding because exposure is not the same thing as effect. And EPA’s
statements about the whooping crane and Indiana bat in its preliminary
assessment are irrelevant because EPA subsequently performed a
species-specific assessment to make its “no effect” finding. EPA’s
analysis about these two species was technically sound. It was neither
arbitrary, capricious, nor contrary to law.
58 NAT’L FAMILY FARM COALITION V. USEPA

designated PCEs,” as required by the statute. 16 U.S.C.
§ 1532(5)(A). True, EPA at times used the word “modify”
instead of “effect,” but EPA was using “modify” as a
synonym for “effect” because EPA used both words
interchangeably in its analysis.

We likewise reject NFFC Petitioner’s argument that
EPA erred by only considering species who use corn, cotton,
or soybean fields. As discussed above, EPA reasonably
concluded there would be no effect outside of the treated
field and limited the action area accordingly. By extension,
it was not unreasonable for EPA to consider only species
who use corn, cotton, or soybean in assessing effect on
critical habitat. 16 U.S.C. § 1532(5)(A)(i), (ii) (“critical
habitat” includes the “geographical area occupied by the
species” unless additional areas are designated by the
consultation agencies).

Nor are we persuaded that EPA’s PCE conclusions are
contradicted by the record as to the Virginia big-eared bat
and the whooping crane. According to NFFC Petitioners,
EPA’s conclusion that there were no PCEs “related to
agriculture” for those two species cannot be right because
EPA explicitly recognized that there are PCEs “related to
agriculture” in a table attached to its 2016 risk assessment.
But regardless of EPA’s notations, no PCEs have been
designated by the consultation agencies for the Virginia big-
eared bat or the whooping crane. 50 C.F.R. § 17.95-a-
Mammals (listing critical habitat and PCEs for protected
species without including any PCEs for the Virginia big-
eared bat); 50 C.F.R. § 17.95-b-Birds-Part 1 (same for the
whooping crane). Any contradiction by EPA as to whether
these species have PCEs related to agriculture is thus legally
irrelevant.

* * *
 NAT’L FAMILY FARM COALITION V. USEPA 59

In its ESA analysis, EPA did not “rel[y] on factors
Congress did not intend it to consider, entirely fail[] to
consider an important aspect of the problem, or offer[] an
explanation that runs counter to the evidence before the
agency or is so implausible that it could not be ascribed to a
difference in view or the product of agency expertise.”
League of Wilderness Defs. Blue Mountains Biodiversity
Project v. Allen, 615 F.3d 1122, 1130 (9th Cir. 2010)
(internal quotation marks and citation omitted). EPA
therefore did not violate the ESA in registering Enlist Duo.

V

Having found error in EPA’s registration decisions under
FIFRA, we now analyze the remedy. “Whether agency
action should be vacated depends on how serious the
agency’s errors are and the disruptive consequences of an
interim change that may itself be changed.” Cal. Cmtys.
Against Toxics v. EPA, 688 F.3d 989, 992 (9th Cir. 2012)
(internal quotation marks and citation omitted). We also
look to “whether the agency would likely be able to offer
better reasoning or whether by complying with procedural
rules, it could adopt the same rule on remand, or whether
such fundamental flaws in the agency’s decision make it
unlikely that the same rule would be adopted on remand.”
Pollinator Stewardship Council v. EPA, 806 F.3d 520, 532
(9th Cir. 2015). Finally, we “consider whether vacating a
faulty rule could result in possible environmental harm.” Id.

Here, remand without vacatur is warranted. EPA’s
error—failing to consider harm to monarch butterflies
caused by killing target milkweed—is not “serious,” Cal.
Cmtys., 688 F.3d at 992, especially in light of EPA’s full
compliance with the ESA and substantial compliance with
FIFRA. Moreover, given the technical nature of EPA’s
error, EPA will “likely be able to offer better reasoning” and
60 NAT’L FAMILY FARM COALITION V. USEPA

“adopt the same rule on remand.” Pollinator, 806 F.3d
at 532. Thus, regardless of how “disruptive” the
consequences of vacatur would be, Cal. Cmtys., 688 F.3d
at 992 (citation omitted)—and there is evidence of
potentially serious disruption if a pesticide that has been
registered for over five years can no longer be used—vacatur
would not be warranted. We therefore remand without
vacatur so EPA can address the evidence that monarch
butterflies may be harmed by the destruction of milkweed on
target fields in determining whether the registration of Enlist
Duo will lead to any “unreasonable adverse effect” on the
environment. We “expect and urge EPA to move promptly
on remand.” EME Homer City Generation, L.P. v. EPA,
795 F.3d 118, 132 (D.C. Cir. 2015); cf. In re Core
Commc’ns, Inc., 531 F.3d 849, 862 (D.C. Cir. 2008)
(Griffith, J., concurring) (“Remand without vacatur is
common . . . [b]ut experience suggests that this remedy
sometimes invites agency indifference.” (citation omitted)).

VI

NFFC Petitioners’ petition for review is DENIED.
Petitioner NRDC’s petition for review is GRANTED in
part and DENIED in part. The case is REMANDED
WITHOUT VACATUR.

R. NELSON, Circuit Judge, concurring:

We have addressed venue and standing to ensure that we
have jurisdiction over one petitioner for each petition. Here,
the interplay between FIFRA’s venue provision and Article
III standing does not make a difference because, for each
petition, one petitioner over which venue is proper has also
demonstrated standing. I write separately, however, to
 NAT’L FAMILY FARM COALITION V. USEPA 61

address how the interplay of FIFRA’s venue provision and
standing could make a difference in a future case.

Under FIFRA’s venue provision, “any person who will
be adversely affected by [a pesticide registration] order and
who had been a party to the proceedings may obtain judicial
review by filing in the United States court of appeals for the
circuit wherein such person resides or has a place of
business.” 7 U.S.C. § 136n(b). “Person,” as used in FIFRA,
“means any individual, partnership, association, corporation,
or any organized group of persons whether incorporated or
not.” Id. § 136(s). 1 Here, three of the NFFC Petitioners
(National Family Farm Coalition, Family Farm Defenders,
and Beyond Pesticides) do not “reside” or “have a place of
business” in the Ninth Circuit and three do (Center for Food
Safety (“CFS”), Center for Biological Diversity (“CBD”),
and Pesticide Action Network North America (“PANNA”).

As I read the statute, this means venue is not proper as to
the first three petitioners. Nothing in FIFRA’s venue
provisions suggests Congress intended venue to be analyzed
petition-by-petition. To the contrary, the provision’s use of
the singular noun “person” and definition of that word in a
way that requires individual analysis suggests venue should
be analyzed on a petitioner-by-petitioner basis. I therefore
read the statute as requiring us to analyze venue on an
individual basis, even if multiple petitioners join one
petition. Id. §§ 136n(b), (s).

1
Venue should be addressed before standing because venue is, like
forum non conveniens, a nonmerits issue that “den[ies] audience to a case
on the merits” without assuming “substantive law-declaring power.” See
Sinochem Int’l Co. v. Malay. Int’l Shipping Corp., 549 U.S. 422, 432–
33 (2007) (internal quotation marks omitted).
62 NAT’L FAMILY FARM COALITION V. USEPA

This reading of the statute is consistent with 28 U.S.C.
§ 2112(a)(1)–(5), cited in 7 U.S.C. § 136n, which recognizes
that petitions for review may be filed in multiple courts of
appeal. Id. If that happens, one of two things occurs. If
petitions for review are filed in “at least two courts of
appeals” “within ten days after issuance of the” relevant
order, a lottery before the judicial panel on multidistrict
litigation is triggered, through which one court of appeals is
designated to adjudicate all of the petitions. Id. § 2112(a)(1),
(a)(3). If, however, a petition for review is filed in only one
court of appeals within ten days, or, if multiple petitions for
review are filed more than ten days after the order, the
administrative record is filed in the court where the first
petition for review was filed. Id. § 2112(a)(1). Any
subsequent petitions “shall [be] transfer[red] . . . to the court
in which the record [was] filed.” Id. § 2112(a)(5). The court
of appeals where the administrative record is filed then
becomes the court with “exclusive jurisdiction to affirm or
set aside the order complained of in whole or in part.”
7 U.S.C. § 136n(b); see also Remington Lodging & Hosp.,
LLC v. NLRB, 747 F.3d 903, 904 (D.C. Cir. 2014)
(describing the lottery provision).

This process would be circumvented if all petitioners
could join a single petition in the same circuit, regardless of
whether each petitioner had proper venue. The Tenth Circuit
reached the same conclusion in a case involving a similar
venue statute under the Natural Gas Act. See Amerada
Petroleum Corp. v. Fed. Power Comm’n, 338 F.2d 808, 810
(10th Cir. 1964). In that case, six corporations and four
individuals filed a “joint petition” for review of a Federal
Power Commission order in the Tenth Circuit even though
only one of those petitioners did business or resided in the
Tenth Circuit. Id. at 809. Relying on 28 U.S.C. § 2112, the
Tenth Circuit held that the proper procedural step was for
 NAT’L FAMILY FARM COALITION V. USEPA 63

each petitioner to file in the proper venue, so that
proceedings could be consolidated in one circuit. Id. at 810;
see also Fed. Power Comm’n v. Texaco, Inc., 377 U.S. 33,
39 (1964) (dismissing one of two petitions for review
because one petitioner did not have its principal place of
business within the circuit where the petition was filed);
Dalton Trucking, Inc. v. EPA, 808 F.3d 875, 878–82 (D.C.
Cir. 2015) (dismissing petition for review filed in D.C.
Circuit for failure to satisfy Clean Air Act’s venue
provision).

Whether FIFRA requires an individualized venue
analysis becomes important when there are joint petitioners,
only some of whom can show proper venue and Article III
standing. In this case, one NFFC Petitioner—CFS—
“resides or has a place of business” in the Ninth Circuit,
7 U.S.C. § 136n(b), and submitted sufficient proof to satisfy
the requirements of Article III, as discussed in the majority
opinion, Majority Op. at 15, 16–17. But that will not always
happen. By way of example, one of the other two NFFC
Petitioners with proper venue here—CBD—would not have
satisfied Article III standing with respect to NFFC
Petitioners’ FIFRA claims. That is so because the interests
asserted by CBD’s members relate exclusively to the ESA—
not FIFRA. So CBD did not show that FIFRA is the statute
“designed to protect” the “threatened concrete interest[s]”
that CBD asserts. Salmon Spawning & Recovery All. v.
Gutierrez, 545 F.3d 1220, 1225 (9th Cir. 2008) (internal
quotation marks omitted).

And PANNA would not have satisfied Article III’s
standing requirements with respect to either claim advanced
in the NFFC Petition. PANNA’s lone declaration does not
include any statements “establishing that at least one
identified member ha[s] suffered or would suffer harm.”
64 NAT’L FAMILY FARM COALITION V. USEPA

Summers v. Earth Island Inst., 555 U.S. 488, 498 (2009).
Instead, it generally asserts that PANNA’s members will be
adversely affected by the registration of Enlist Duo. This
generalized harm is not enough to establish injury in fact for
purposes of associational standing. Associated Gen.
Contractors of Am., Inc. v. Cal. Dep’t of Transp., 713 F.3d
1187, 1194–95 (9th Cir. 2013) (dismissing appeal for lack of
standing because plaintiff did not “submit[] declarations by
any of its members attesting to harm they have suffered or
will suffer” under the challenged program).

What happens, then, if venue is proper as to some
petitioners, but only a petitioner without proper venue
satisfies the requirements for Article III standing? In my
view, the petition for review should be dismissed. I therefore
believe that future panels should closely scrutinize both
venue and Article III standing in FIFRA cases to ensure that
both requirements are met.

WATFORD, Circuit Judge, dissenting:

I agree with my colleagues that we have jurisdiction to
review the petitioners’ challenges and that the
Environmental Protection Agency (EPA) violated the
Federal Insecticide, Fungicide, and Rodenticide Act by
failing to assess the impact that Enlist Duo’s use will have
on monarch butterflies. But in my view, EPA also violated
the Endangered Species Act by failing to use the best
scientific data available to assess whether Enlist Duo will
adversely affect threatened or endangered species. For that
reason, I would vacate the 2014 and 2017 registrations under
review.
 NAT’L FAMILY FARM COALITION V. USEPA 65

The Endangered Species Act and its implementing
regulations require EPA to determine whether registering a
pesticide for use “may affect” any species listed as
threatened or endangered. 16 U.S.C. § 1536(a)(2); 50 C.F.R.
§ 402.14(a). If registering the pesticide for use may affect
one or more listed species, EPA must then consult with
expert wildlife agencies to ensure that its action is not likely
to jeopardize the continued existence of any such species.
16 U.S.C. § 1536(a)(2). In making its threshold “may
affect” determination, EPA must, in the words of the statute,
“use the best scientific and commercial data available.” Id.
EPA did not comply with that statutory mandate here
because the method it used to assess Enlist Duo’s effects on
listed species is scientifically unsound.

To evaluate the risks that Enlist Duo poses to listed
species, EPA applied the “risk quotient” method, an
approach that integrates exposure and toxicity data using
single-point estimates. When EPA calculates risk quotients,
it divides a single-point estimate of the maximum amount of
a pesticide to which a species might be exposed by a single-
point estimate of the minimum amount of that pesticide
expected to adversely affect the species. In theory, the
higher the resulting number, the higher the purported risk.

But as the National Academy of Sciences explained in a
2013 report—issued in response to EPA’s own request for
advice on the subject—the risk quotient method does not
“estimate risk” at all. National Research Council of the
National Academies, Assessing Risks to Endangered and
Threatened Species From Pesticides 149 (2013). 1 It

1
The report is available at: https://www.nap.edu/catalog/18344/
assessing-risks-to-endangered-and-threatened-species-from-pesticides
[https://perma.cc/M7V3-AYEL].
66 NAT’L FAMILY FARM COALITION V. USEPA

“provides no information about the probability of an adverse
effect” because single-point estimates do not account for the
full range of possible exposure scenarios. Id. at 149–50.
What’s more, risk quotients may not even reflect the worst-
case scenario, despite EPA’s attempt to maximize the
numerator and minimize the denominator in the equation.
That is because the underlying data on which EPA relies to
calculate the numerator and denominator includes single-
point estimates of uncertainties. Id. If EPA gets those
single-point estimates wrong, the ultimate risk quotient
could underestimate risk and EPA would never know it. Id.
Thus, even if EPA uses conservative thresholds to assess
whether a particular risk quotient is high enough to warrant
consulting with the expert wildlife agencies, that
precautionary measure will not offset the method’s
fundamental flaws. No matter how conservative its
thresholds, EPA will still be interpreting an unreliable metric
of risk. See id.

In light of these observations, the Academy concluded
that risk quotients “are not scientifically defensible for
assessing the risks to listed species posed by pesticides or
indeed for any application in which the desire is to base a
decision on the probabilities of various possible outcomes.”
Id. at 15. The Academy further concluded that adoption of
an alternative methodology entirely—one that actually
measures probabilities—is the only way to achieve
“realistic, objective estimates of risk.” Id. And though the
Academy recognized that EPA would not be able to
implement probabilistic methods overnight, it reiterated that
EPA’s current approach to risk assessments is “not
appropriate.” Id. at 150, 152.

EPA does not dispute the Academy’s scientific
conclusions. Nor has EPA made any attempt to justify, on a
 NAT’L FAMILY FARM COALITION V. USEPA 67

scientific basis, its continued reliance on the risk quotient
method. Instead, EPA simply highlights its use of
conservative assumptions—without addressing the
Academy’s criticisms of such assumptions—and points to
practical reasons for adhering to its risk quotient approach.
Specifically, EPA emphasizes the administrative burdens of
applying the probabilistic method to all pesticide registration
decisions, and explains that it will use risk quotients to assess
Enlist Duo’s effects until it allocates the necessary resources
to switch methodologies. That approach is permissible, EPA
contends, because the wildlife consultation agencies have
agreed that EPA can implement the Academy’s
recommendations in stages. But it should go without saying
that neither pragmatic concerns nor an interagency
agreement can absolve EPA of its statutory obligation to use
the best scientific data available. Rather, EPA must apply
its scientific expertise to identify and use data that meets the
statutory standard. See Conservation Congress v. Finley,
774 F.3d 611, 620 (9th Cir. 2014). EPA failed to do so here.

The majority nonetheless upholds EPA’s use of the risk
quotient method for making “may affect” determinations. In
the majority’s view, an agency does not violate its statutory
obligation to use the best scientific data available unless it
rejects better existing data. Because probabilistic data for
Enlist Duo was not available at the time EPA conducted its
analysis, the majority concludes that EPA fulfilled its
statutory duty by using the data it already possessed,
however unreliable that data may have been.

In reaching this outcome, the majority has created a new
rule with serious implications. Following today’s decision,
an agency may rely on data produced by a scientifically
indefensible methodology so long as better data, produced
by a methodology that is scientifically defensible, has not yet
68 NAT’L FAMILY FARM COALITION V. USEPA

been generated. Because courts in our circuit must now
accept that fundamentally flawed data as the “best” scientific
data available, the agency will have no incentive to
implement the scientific methods necessary to obtain
reliable data. That is not what Congress intended when it
required EPA and other federal agencies to use the best
scientific data available, and it is certainly not the outcome
that our cases demand.

The purpose of the “best scientific data available”
requirement is to “ensure that the [Endangered Species Act]
not be implemented haphazardly, on the basis of speculation
or surmise.” Bennett v. Spear, 520 U.S. 154, 176 (1997).
This standard does not require agencies to use “the best
scientific data possible,” San Luis & Delta-Mendota Water
Authority v. Jewell, 747 F.3d 581, 602 (9th Cir. 2014)
(emphasis added), but it does require them to “support their
conclusions with accurate and reliable data,” Conservation
Congress, 774 F.3d at 620. Accordingly, while we have
permitted agencies to rely on “imperfect” or “weak” data, we
have never suggested that agencies may rest their decisions
on data that is scientifically unsound. See id. Yet that is
precisely what the majority has allowed EPA to do here.

Of course, the statutory standard does not compel
agencies to “conduct new tests,” and we cannot direct EPA
to obtain better data using the probabilistic approach
recommended by the National Academy of Sciences. See
San Luis & Delta-Mendota Water Authority v. Locke,
776 F.3d 971, 995 (9th Cir. 2014). But we can—and indeed,
must—ensure that EPA uses the best scientific data available
to assess Enlist Duo’s effects on listed species. By relying
on a scientifically indefensible method that generated
speculative and unreliable estimates, EPA failed to meet its
burden. While the unavailability of better data can excuse
 NAT’L FAMILY FARM COALITION V. USEPA 69

an agency’s reliance on flawed or weak data, that rule has no
application when, as here, an agency’s data does not qualify
as “scientific data” in the first place.

EPA’s use of the risk quotient method violated the
Endangered Species Act, and the 2014 and 2017
registrations of Enlist Duo should be vacated as a result.