Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 6 of 29

INTEREST OF AMICUS CURIAE 1
David A. Mortensen, Ph.D., is a Distinguished Professor of Weed and
Applied Plant Ecology in the Plant Sciences Department at The Pennsylvania State
University. Dr. Mortensen applies his background in applied plant ecology and
ecologically-based pest management to improve the sustainability of land resource
management. His work explores the landscape-level interplay between the ecology
of agricultural fields, field edges, and forest fragments. An example of this work
involves assessing approaches to integrating weed management with the goal of
reducing reliance on herbicide use. Dr. Mortensen has a long-standing interest in
making weedy plant management more sustainable through understanding how
management tactics interact. Dr. Mortensen balances his research interests with
teaching such courses as: Principles of Weed Management, Flora of the Central
Appalachian Region, and Ecology of Agricultural Systems. He received the
Outstanding Research Award and Fellow from the Weed Science Society of
America, and has chaired the USDA national competitive grants program in the
Weed Science Field. He received his Ph.D. in Crop Science from North Carolina

1

Pursuant to Fed. R. App. P. 29(a)(2), all parties have consented to the filing of
this brief. Pursuant to Fed. R. App. P. 29(a)(4)(E), the Amicus states that no
party or counsel in this case and no person except counsel of record for Amicus
authored or contributed money to fund the preparation of this brief.
1

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 7 of 29

State University, his M.S. in Botany from Duke University, and his B.S. in Botany
from Drew University.
A decade ago, Dr. Mortensen was the first Weed Scientist to call attention to
a number of concerns regarding EPA’s proposed registration of dicamba
formulations for use on dicamba-resistant crops. Since that time, and as explained
in his brief, he has published and spoken on the subject numerous times. He also
participated in the public comment portions of the EPA’s decision-making process
that led to this case. As an academic stakeholder whose professional interests will
be adversely affected by the EPA’s approval of the dicamba product XtendiMax,
Dr. Mortensen has a strong interest in presenting his concerns to the Court.
ARGUMENT
A. Introduction.
I was trained as a weed scientist at North Carolina State University, where I
completed my doctoral degree in 1986. At the time, North Carolina State
University’s Weed Science program was considered one of the top two or three
programs in the country. I rose to the rank of Full Professor at the University of
Nebraska, then moved from the University of Nebraska in 2001 to take a similar
position at The Pennsylvania State University, where I continue to work today. In
the past month, I was honored with the title Distinguished Professor. These two
2

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 8 of 29

university work assignments, coupled with my graduate work in North Carolina
and an extended collaboration at the Jamie Whitten Research Center in Stoneville,
Mississippi, have made it possible for me to conduct applied weed science work on
some 300 practicing commodity crop farms in North Carolina, Nebraska, Iowa,
Pennsylvania, New York and Mississippi during the course of my career.
I was the first Weed Scientist to call attention to a number of concerns
regarding EPA’s approval of dicamba formulations for use on dicamba-resistant
crops. From the beginning, a principal concern was non-target damage to crops and
non-crop plant communities – so-called non-target injury – resulting from dicamba
vapor drift. I was invited to speak on the subject at the 2009 International
Integrated Pest Management conference, where I presented a paper entitled
Unintended consequences of stacking herbicide tolerance traits in soybean.2 Given
the 4.3 million acres of dicamba-injured soybeans reported for the 2017 field
season, and given that I am fairly certain a similar problem will repeat itself in the
2018 field season, my words in that 2009 abstract are particularly forward looking:
Widespread adoption of glyphosate tolerant soybeans has increased
the number of weedy species resistant to this herbicide. To address

2

David A. Mortensen et al., Unintended consequences of stacking herbicide
tolerance traits in soybean, Paper presented at the 6th International IPM
Symposium, Portland OR (March 2009).
3

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 9 of 29

this problem, the industry is commercializing soybeans that are
resistant to glyphosate and to dicamba. Dicamba, a broadleaf
herbicide, is highly volatile and extremely active on many broadleaf
crop and field edge plants. The high risk of injuring soybean not
carrying the dicamba trait will drive adoption of these new cultivars.
This practice has a high potential of injuring other broadleaf crops and
of significantly reducing floristic biodiversity in field edges and
nearby non-crop habitat and the ecosystem services they provide. 3
I made that statement in a published abstract in 2009, and that’s exactly what
occurred in the 2017 field season.
B. Background work on the issue.
I have strongly opposed widespread adoption of crops with dicamba
resistance traits in soybean and cotton because of the increased use of dicamba
they entail since first learning about the trait in the mid 1990’s when then Dr.
Donald Weeks was interviewing for the directorship of the newly constructed
Biotechnology Center at the University of Nebraska. Dr. Weeks was coming from
a private research lab and during his interview seminar spoke excitedly about the
dicamba resistance gene he and his colleagues had isolated. I remember exiting his
seminar with a long-time farmer and extension specialist friend, Dr. Alex Martin,
who remarked that a dicamba resistance trait in soybean “is the most

3

Id.; J. Franklin Egan, Eric Bohnenblust, Sarah Goslee, David A. Mortensen
& John Tooker, Herbicide drift can affect plant and arthropod communities,
185 Agriculture, Ecosystems, and Environment 77-87 (2014).
4

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 10 of 29

knuckleheaded idea I’ve ever heard of.” Dr. Martin was a knowledgeable field
weed scientist and farmer in eastern Nebraska. I valued his views highly. What Dr.
Martin was referring to was the incredibly high potential to damage adjacent, nontransformed plants (plants without the dicamba-resistance trait) with this, one of
the most drift (particularly vapor drift) prone herbicides. Dr. Martin’s concerns are
reflected in the pesticide drift enforcement surveys conducted by the Association
of American Pesticide Control Officials (AAPCO) that have found dicamba to be
among the top three herbicides implicated in pesticide drift episodes, despite its
limited use. 4
In my view, USDA should not have deregulated dicamba-resistant crops nor
should EPA have approved post-emergence application of dicamba to them. I
strongly argued in public comments to both agencies and in numerous public
presentations and peer-reviewed papers 5 that this transformed crop and herbicide
package wouldn’t address the herbicide resistance problem in a sustainable way
and that there were significant gaps in our understanding of the non-target risk of

4

5

Association of American Pesticide Control Officials (AAPCO), 1999 Pesticide
Drift Enforcement Survey: 1996 to 1998, AAPCO (1999); Association of
American Pesticide Control Officials (AAPCO), 2005 Pesticide Drift
Enforcement Survey: 2002 to 2004, AAPCO (2005).
David A. Mortensen et al., Navigating a Critical Juncture for Sustainable Weed
Management, 62 BioScience 75-84 (2012).
5

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 11 of 29

widespread, landscape scale use of dicamba (I address this point more fully later in
the brief).
My colleagues and I visited the US EPA to share our concerns about
dicamba drift in 2010. During that visit, our seminar and the ensuing discussion
was transmitted live to US EPA laboratories across the country. In 2012, I
submitted comments to US EPA on Monsanto’s first application to register
dicamba for use on dicamba-resistant soybeans. 6 In my 2016 public comment to
the US EPA on the new uses of dicamba on herbicide-tolerant cotton and
soybeans, I argued that we had critical data gaps having to do with the frequency
and impact of off-site movement of dicamba herbicide.7 I went on to state that
approval of these new dicamba uses “poses PROFOUND risks to broadleaf crop
growers as the risk of physical and vapor drift of dicamba is large (we’ve spent six
years studying the problem). Recent multi-year assessments of how farmers co-

6

See https://www.regulations.gov/document?D=EPA-HQ-OPP-2012-0545-0025.

7

David A. Mortensen, Public participation for dicamba: New use on herbicideresistant cotton and soybean, Comments to EPA, Docket No. EPA-HQ-OPP2016-0187-0838 (2016), available at:
https://www.regulations.gov/document?D=EPA-HQ-OPP-2016-0187-0838
6

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 12 of 29

exist when divergent practices are used on their farms concluded that pesticide
drift should be reduced wherever possible.”8
In those public comments, I argued that deregulation of dicamba-resistant
soybean and EPA approval of dicamba use on that transformed crop increases the
risk of chemical trespass through herbicide drift by virtue of the fact that spray and
vapor drift are extremely common with dicamba, and increasing its use 5-7 fold 9 at
times of the year when drift is highest and when adjacent plants are most
susceptible 10 would dramatically increase injury of adjacent crops and field-edge
plants. I also argued that the drift injury would dramatically decrease plant
diversity and the important provisioning of natural enemies and pollinators.11
I just returned (January 30, 2018) from the Weed Science Society of
America meeting in Arlington, Virginia, where dicamba drift and the resulting

8

Id. at 1.

9

Mortensen et al. (2012).
J. Franklin Egan, Kathryn B. Barlow & David A. Mortensen, A meta-analysis on
the effects of 2,4-D and dicamba on soybean and cotton, 62 Weed Science 193206 (2014).
Eric W. Bohnenblust, Anthony D. Vaudo, J. Franklin Egan, David A. Mortensen
& John F. Tooker, Effects of the herbicide dicamba on non-target plants and
pollinator visitation, 35 Environmental Toxicology and Chemistry, Journal of
Pest Science 144-151 (2016); Melanie A. Kammerer, David J. Biddinger, Edwin
G. Rajotte & David A. Mortensen, Local plant diversity across multiple habitats
supports a diverse apple pollinator community, 45 Environmental Entomology
32-48 (2016).

10

11

7

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 13 of 29

injury were discussed for much of the meeting. Based on those presentations and
on estimates I’ve read about in the farm press, approximately 25 million acres of
dicamba-resistant soybean and cotton were planted in the US in 2017, 12 and most
of those acres received one or more applications of XtendiMax or another dicamba
formulation approved for use on dicamba-resistant crops. As a result of that
practice, 4.3 million acres of non-transformed soybeans were injured by dicamba
drift. The extent of crop injury in response to this crop and herbicide use practice is
unprecedented. It is also a significant under- representation of the plant injury that
occurred. Importantly, the 4.3 million acres only addresses injury to soybean and
doesn’t include other broadleaf crops nor does it include non-crop broadleaf
flowering plants.
C. EPA should have foreseen the substantial dicamba drift injury.
This unprecedented extent of damage begs the question: Should the EPA
have foreseen substantial dicamba drift injury resulting from the use of
XtendiMax? I strongly believe that this sort of injury should have been anticipated.
During the EPA’s review, several important factors were missed, and/or

12

Greg D. Horstmeier, Dicamba’s PTFE Problem, DTN The Progressive Farmer
(8/29/2017), https://www.dtnpf.com/agriculture/web/ag/perspectives/
blogs/editors-notebook/blog-post/2017/08/29/dicambas-ptfe-problem.
8

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 14 of 29

misunderstood, and surprisingly naïve assumptions were made or overlooked
entirely when extrapolating from laboratory tests and small field plot studies to
field and farmstead scales. First, dicamba is highly phytotoxic to broadleaf plants,
with some plant families being hypersensitive to the herbicide. Plants in the
legume family (e.g. peas, soybean) are extremely sensitive to dicamba. Because
dicamba is a growth regulator herbicide, it is most active on plants when they are
actively growing. For most summer grown arable crop plants, that period of rapid
growth occurs at the very same time XtendiMax herbicide is applied for
postemergence weed control in the dicamba-resistant crop. This is the very time
when many summer arable crop plants and field edge flowers enter their
exponential growth stage, or their stage of most rapid growth, and therefore when
they would be most susceptible to injury. This point was borne out in a peerreviewed meta-analysis we published in a prominent weed science journal in 2014,
and for which we received the Outstanding Paper Award from the Weed Science
Society of America. 13
Second, in addition to the fact that the neighboring plants are at a
particularly sensitive stage of development, these new, widespread postemergence

13

Egan et al. (2014).
9

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 15 of 29

applications of XtendiMax are made during the hottest months of the summer.
While it is true that other formulations of dicamba have been used before dicambaresistant soybean was deregulated in 2015, that use was on a small fraction of the
acres and importantly, in the spring when summer crops haven’t emerged and
when field and soil temperatures are 30-65 degrees F. cooler. Why does
temperature matter so much? Dicamba is a volatile herbicide. This means it is
prone to volatilize (vaporize) from soil and plant surfaces, which can occur hours
to days after dicamba is sprayed. To volatilize means to undergo a phase change
from a liquid to a gas or a solid to a gas. The amount of dicamba (or any substance)
that volatilizes increases with temperature. It should be clear, we’re not talking
about subtle differences in temperature here. The springtime soil surface
temperature varies from 40-65 degrees while a mid-summer application is typically
made when soils are 75 to as much as 180 degrees at the soil surface. 14 This
difference in temperature is both large in absolute terms and very important to the
problems that resulted with XtendiMax use. EPA’s registration of XtendiMax for
use as a postemergence herbicide – which would therefore by applied in the
summer months – guaranteed that XtendiMax would sprayed on fields when the

14

Thomas J. Monaco et al., Weed Science: Principles and Practices 148-160 (4th
ed. 2002).
10

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 16 of 29

soil and air temperatures would be much warmer and the risk of volatilization
much higher.
Finally, it is well understood that as the area treated (land area) and amount
of a chemical applied (mass) increases, the risk of finding that chemical in
unwanted places increases. A careful review of the peer-reviewed environmental
chemistry literature indicates that amount of herbicide used is positively correlated
with the appearance of herbicides in surface and ground water. 15 Auxinic
herbicides like 2,4-D and dicamba have historically been linked to a high
frequency of drift injury events.16 By many practitioners in the field they are
referred to as “the bad actors” – not a little more drift-prone, but much more driftprone than most pesticides. All of this was well understood and well known well
before USDA moved to deregulate dicamba-resistant crops and EPA approved
XtendiMax herbicide for use on them. Taken together, these spillover effects
would be small in the absence of the transformed crop and associated agronomic
practices. In addition to concerns about compromised environmental quality,
herbicide spillover of the kind that occurred with XtendiMax use should have been

15

16

Jack E. Barbash et al., Major herbicides in ground water: Results from
the National Water Quality Assessment, 30 Journal of Environmental Quality
831-845 (2001).
AAPCO 1999; AAPCO 2005; Egan et al. (2014).
11

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 17 of 29

foreseen. Again, I raised these issues in my public comments and apparently, they
went unheeded.
D. My academic studies on the issue.
I began this brief by stating I was the first weed scientist to call attention to
the potential for significant and widespread drift problems with auxinic herbicides
like dicamba if we proceeded to develop and use dicamba-resistant crops. I put my
credentials on the line when I called this “new” direction of weed science into
question. While my credentials in this field of study attest to my deep
understanding of this problem – I’ve received the Outstanding Research Award
and Fellow from the Weed Science Society of America, as recently as 2016
chaired the USDA national competitive grants program in the Weed Science field,
and just learned that I’ve received the title of Distinguished Professor – many
weed scientists were reluctant to join me (early on) in questioning the risks of
dicamba-resistant crops and the resulting widespread use of dicamba. Weed
scientists finally caught on to the risks of adjacent crop injury when in 2014 many
Extension Weed Scientists flatly stated dicamba-resistant soybean was a nonstarter, simply too dangerous from a vapor and from a spray drift perspective.
My concerns about the potential for drift injury led my lab to initiate a series
of field studies exploring the drift dynamics of dicamba herbicide. I will begin by
12

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 18 of 29

stating, studies of this kind are difficult to conduct and almost any approach taken
has limitations associated with the methods used. Our approach was to conduct
plot scale experiments with the goal of estimating the drift potential of two
commercially available formulations of dicamba (diglycolamine and
dimethylamine) with the goal of informing our understanding of how far dicamba
could move under a simulated field application. The study was thoughtfully
designed and carefully executed and concluded that the dimethylamine formulation
was more drift-prone than the newer diglycolamine formulation.17 However, we
observed a significant amount of drift with both formulations. Our study was
carefully designed to focus on vapor drift as the “target” bioindicator plants were
placed in the field (they were potted and growing in the greenhouse) after the
herbicide had been applied, thus minimizing the likelihood of spray drift injury.
The most striking result from the study was the degree to which vapor drift
was correlated with temperature. During my career I’ve worked in soybean fields
in Pennsylvania, Nebraska, Iowa, Mississippi and North Carolina. The
summertime temperatures in Pennsylvania are far cooler than any other place I’ve

17

J. Franklin Egan & David A. Mortensen, Quantifying vapor drift of
dicamba herbicides applied to soybean, 31 Environmental Toxicology and
Chemistry 1-9 (2012).
13

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 19 of 29

worked, and even in the cool summers of the Central Appalachian Mountains we
saw a striking influence of temperature. Since the drift studies were replicated five
times, we were able to determine the degree to which drift was correlated with
temperature and relative humidity for the dimethylamine salt of dicamba. The
maximum temperature during the study window was 82 degrees while most days
were in the mid 70’s. Over this narrow and cooler range of temperatures the extent
of crop injury and distance of herbicide movement was highly correlated with
temperature, with drift distance doubling over a range of 75 to 82 degrees. While
the experiment was thoughtfully designed, it had its limitations. First, the newer
diglycolamine formulation was only tested in one field season (2010) while the
older dimethylamine formulation was tested in 2009 and 2010. Therefore, our
ability to infer to temperatures outside the range we experienced was highly
constrained. The fact that the weather was cooler than normal during the 2010
window was something we had to live with given the complexities of conducting
field studies like this. Those complexities included treating the plants at the right
growth stage and fitting the experiments in between rains and other inclement
weather. In spite of less than ideal, cooler weather, we saw and reported a strong
temperature effect on vapor drift and highlighted this finding in the paper.
The study described above was a “small” plot study, a study that revealed
14

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 20 of 29

some worrying realities. While the newer formulation reduced vapor drift, vapor
drift was observed for new and old formulations AND while the experiments were
conducted under cooler conditions, we saw a strong positive correlation between
vapor drift and temperature. Why do I highlight “small” plot study in the first
sentence of this paragraph? I do so because while they have their place in
controlled component research, there are serious scaling deficiencies associated
with small plot work. What follows is an attempt to describe the challenges
associated with scaling results from a small plot to a typical farm field, a farmstead
and a matrix of farms. As plots go, our experimental plot was fairly large
compared to “tiny” lab-based humidome experimental systems. The treated area of
our plots was 60 feet by 60 feet. That is to say we sprayed a square patch of
soybean field measuring 60 feet by 60 feet or 3,600 square feet. A typical
Midwestern soybean field would fill a quarter section of land (a section is 640
acres, a quarter section is 160 acres). An acre is 43,560 square feet, so a typical
field is 6,969,600 square feet. Thus, our treated area was 0.052% of a typical field
size in area. Why does this present a scaling problem? Imagine the application of
an herbicide to a field as a 3-dimensional plume (pesticide movement modelers
make this assumption all the time). The volume of the dicamba plume over a
typical field would be nearly 2,000 times greater than that over a small plot. Why
15

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 21 of 29

do I highlight this constraint of small plots? The probability of the plume moving
much further and at phytotoxically damaging concentrations is a function of the
plume size. Therefore, in order to understand the risk of dicamba injury at a scale
relevant to modern soybean farming in the US you would need to follow our work
up with a replicated series of field studies where many whole fields were treated
with dicamba over a broad range of temperature, soil moisture and relative
humidity conditions and where injury to adjacent plants was carefully monitored.
E. Problems with EPA’s assessment.
To my knowledge such work was never done and therefore couldn’t have
been used in the EPA assessment of XtendiMax vapor drift. My understanding is
that EPA based its assessment of XtendiMax vapor drift primarily on two
Monsanto field studies, each of which involved just one application of dicamba to
fields 3.4 and 9.6 acres in size.18 This is shockingly insufficient. As I indicated
above, the weather at the time of application matters as does the weather that

18

Monsanto Company, Field Volatility of Dicamba Formulation MON 119096
Following a Pre-Emerge Application Under Field Conditions in the
Southeastern USA, Monsanto Company submission to USEPA, 49888501, p. 12
(03/30/2016), available at https://monsanto.com/app/ uploads/2017/09/Ex.-261.pdf; Monsanto Company, Field Volatility of Dicamba Formulation MON
119096 Following a Post-Emerge Application Under Field Conditions in Texas,
Monsanto Company submission to US EPA, 49888503, p. 12 (03/30/2016),
available at https://monsanto.com/app/uploads/2017/09/Ex.-27-1.pdf.
16

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 22 of 29

follows over the days and weeks after application. What should have been done is
a sufficient number of experimental fields the size of real farm fields should have
been treated and the after-effects studied. Anyone who has conducted field
research would know that to tackle a multidimensional problem like this (where
the response variable vapor drift is dependent on a number of landscape position
and weather factors) would require landscape-scale studies involving field tests on
between 60 and 100 fields. Not having taken this step to conduct and analyze these
studies was irresponsible on the part of the company. The fact that such work
wasn’t required by EPA to inform its decision on whether or under what conditions
to register XtendiMax was a fatal flaw in the Agency’s review process.
The point I raise about the size of the treated area is profoundly important,
particularly if the practice is to be widely adopted by farmers. I just walked
through the issue of scaling from a small plot to a field, but it really only starts
there. Imagine a farmer who is treating an entire farm with dicamba instead of one
field. With a compressed planting period and herbicide spray window, the reality is
that many hundreds to thousands of acres on one farm will be sprayed in a
compressed window of time, say days to a week, and many neighboring farmers
are doing the same. All of a sudden it’s not a plume arising from one 160 acre
field; rather, it is the aggregate of many, many fields. Some describe this
17

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 23 of 29

phenomenon as atmospheric loading.19 Effectively, it is the cumulative effects of
the large-scale application of a pesticide. Remember, 25 million acres of dicambaresistant crops were planted last year, and many millions of acres were treated with
Xtendimax herbicide. My lab is just concluding a spatial analysis of the Midwest
arable cropping region. Imagine, many counties in soybean growing country are
comprised of 80% or more corn and/or soybean. That is to say that 80% of the land
area is planted to corn or soybean. The point is that the spatial extent of this
practice and therefore the spatial loading of the herbicide is enormous. When the
herbicide drifts, the drift arises from a very large treated area. I come back to a
point I raised earlier in the brief: increasing the area treated and amount of
herbicide applied increases the likelihood that herbicides appear in in non-target
sites. 20 We’ve known this for nearly twenty years.

19

20

E.g. Tom Barber, as quoted in David Bennett, Dicamba tests showing similar
results from scattered locations, Delta Farm Press (Sept. 6, 2017),
http://www.deltafarmpress.com/soybeans/dicamba-tests-showingsimilar-results-scattered-locations; see also Ludovic Tuduri et al., A review of
currently used pesticides (CUPS) in Canadian air and precipitation. Part 2:
Regional information and perspectives. 40 Atmospheric Environment 1579–
1589 (2006).
Jack E. Barbash et al., Distribution of Major Herbicides in Ground Water of the
United States, U.S. Geological Survey Water-Resources Investigations Report
98-4245 (1999).
18

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 24 of 29

F. It is nearly impossible to follow EPA’s XtendiMax label.
One final point. I have grave concerns about EPA’s practice of
developing/approving herbicide labels like the XtendiMax label when it’s nearly
impossible to follow the label instructions. EPA scientists should know better. To
expect farmers and commercial applicators to adhere to the narrowly defined
combinations of wind speed and temperature restrictions laid out in the 2017 and
proposed 2018 XtendiMax labels is entirely unrealistic and irresponsible. In much
of the soybean and corn growing regions in the US, strictly following this matrix of
temperature, wind speed and direction rules and fitting the applications in between
rains and other constraints limit application opportunities by 60-80%. In carefully
reading and rereading the XtendiMax label constraints, I continue to be surprised at
the unrealistic assumptions that must have been made while drafting the label.
Given the current structure of the crop production system, it’s not possible to cover
the ground necessary to effectively do dicamba weed control yet conform with the
weather constraints of the label. The implementation of such unrealistic labels sets
farmers and applicators up to fail. In addition to the unrealistically constrained
wind speed restrictions, the notion that rules could be based on a “prevailing” wind
direction as a means of minimizing adjacent susceptible crop injury is deeply
flawed. We saw this first hand in a series of studies we conducted measuring and
19

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 25 of 29

modeling the direction of wind dispersed seed dispersal. In a paper touted by the
journal editors to be the most comprehensive ever published, we documented that
while there is a prevailing wind direction, 30% of the time (in our replicated
studies) the wind blew in exactly the opposite direction.21 Here again, to base safe
buffer distances on an assumption about “prevailing” wind direction sets farmers
up for failure.
G. EPA should not have approved XtendiMax, and the 2018 label will not
fix the problem.
Taken together, XtendiMax should never have been registered for use on
dicamba-resistant crops in 2017. The necessary experiments to assess the risk of
vapor drift were never conducted. Frankly, Monsanto prohibited drift-related
testing of XtendiMax prior to commercialization in 2017, and EPA registered the
herbicide without the data needed to adequately assess the risk of adjacent crop
injury (or other plants). The scaling issue is widely known to modelers and
quantitative agronomists and is an issue that should have been addressed prior to
labeling. There was every reason to believe there was a significant risk of
widespread vapor and spray drift with XtendiMax and yet the label was granted.

21

Joseph T. Dauer, David A. Mortensen & Mark J. VanGessel, Temporal
and spatial dynamics of long-distance Conyza canadensis seed dispersal, 44
Journal of Applied Ecology 105-114 (2007).
20

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 26 of 29

Frankly, it’s my view that market forces encouraged the “let’s see how it goes the
first time around and we can tweak the label later if we have a problem” approach
to labeling and stewarding XtendiMax. Unfortunately, EPA supported the
approach by granting the label. The 2018 label is unlikely to improve things. I say
this first because the needed field and farmstead-scale studies still have not been
conducted, and second because I see no label amendments that address the critical
issue of XtendiMax vapor drift. Even if it were the case that this dicamba-based
weed management was a robust solution to the widespread glyphosate-resistant
weed problem – which it is clearly not22 – the risks associated with non-target crop
and non-crop plant injury were far too great to have proceeded to register
XtendiMax for the 2017 field season, and I feel the same about the 2018 field
season for the same reasons.

22

Mortensen et al. (2012); J. Franklin Egan, Bruce D. Maxwell, David A.
Mortensen, Matthew R. Ryan & Richard G. Smith, 2,4-Dichlorophenoxyacetic
acid (2,4-D)-resistant crops and the potential for the evolution of 2,4-Dresistant weeds, 108 PNAS E38 (2011).
21

 Case: 17-70196, 02/16/2018, ID: 10767632, DktEntry: 78, Page 27 of 29

CONCLUSION
XtendiMax should never have been approved for use on dicamba-resistant
crops in 2017. The EPA approved the herbicide without the necessary data, and the
resulting drift-related injuries during the first crop season were disastrous.
Therefore, I support the Petitioners’ request for declarations that the EPA violated
FIFRA and the ESA. Further, the EPA’s approval of XtendiMax should be
vacated, with a remand for further proceedings.

Respectfully submitted,
s/ David A. Mortensen
David A. Mortensen, Ph.D.
Amicus Curiae
Filed by:
s/ Jesse A. Buss
Jesse A. Buss
Oregon State Bar No. 122919
WILLAMETTE LAW GROUP
411 Fifth Street
Oregon City OR 97045-2224
Telephone: 503-656-4884
Facsimile: 503-608-4100
Email: jessebuss@gmail.com
Counsel for Amicus Curiae

22