.
. ^
" *~~<
>*«.„
-------
-------
Hifi
't&fis
Agricultural Atrazine Use and Water Quality:
A CEEPES Analysis of Policy Options
S/;"iJii-toi.
US Environmental Protection Agency
Office of Program and Policy Evaluation
Water and Agricultural Policy Division
Agricultural Policy Branch
September 1993
Model Development and Technical
Assistance Provided by
Center for Agricultural and Rural Development
Resource and Rural Development
Resource and Environmental Policy Division
under Cooperative Agreement CR-816099-0.1 -I
and Robert Taylor of Auburn University
Andrew Manale, Project Officer
-------
m
-------
Executive Summery
An ecological-economic modelling
system was used to compare the risks
and benefits of national bans on the
use of the corn and sorghum herbi-
cide atrazine, as well as the entire
group of triazine herbicides. A major
advance of the system is its linking of
decisions on the choice of technology
to environmental and economic im-
pacts. It accomplishes this by linking
biogeophysical process models with
economic models. In response to al-
ternative policies, the system simu-
lates farmer substitution among
inputs, crops, and agricultural prac-
tices. Because impacts may occur at
different temporal and spatial scales
according to the medium through
which the damage occurs, the system
aggregate results temporally and spa-
tially.These results are linked, in turn,
to theeconomic benefits derived from
agricultural production in the corre-
sponding spatial unit. It provides as-
sessments of policy options at a local,
regional, and national scale.
Atrazine and the other triazine
herbicideswerechosenforthisanaly-
sis because they are the most widely
detected group of pesticides in sur-
face and groundwater. Current atra-
zine use in the Midwestern states
also represents nearly 12% of total
agricultural usage of pesticides in the
United States. Hence policies de-
signed to reduce corn and sorghum
herbicide use can lead to significant
reductions in overall pesticide use.
A ban on atrazine results in a
decrease of $937 million in producer
income and consumer effects. This
losswouldbecounterbalancedinpart.
by a sharp reductions in government
expenditures for corn price support
leading to an overall short-term eco-
nomic loss of $365 million for the
entire nation. Overall herbicide use
would not decrease with an atrazine
ban. Indeed, total triazine use would
increase by 27% on corn, though it
would decrease by 84% on sorghum.
Nontriazineherbicide use would also
increase as more acreage is treated
with nontriazine herbicides at higher
application rates. Use of the new low-
dosage sulfonylurea herbicides
would greatly increase, at least in the
short run.
A ban on atrazine alone, without a
significantshiftawayfromchemically-
intensive agricultural practices, could
exacerbate the current public health
concern regarding surface water qual-
ity. Though) exposures, in general, to
herbicides ingroundwaterwould gen-
erally decline and be of lower general
health concern than under current use
patterns in any region regardless of
tillage practice, there could be an ac-
tual increase in site-specific short-term
exposures to other triazine herbicides
that pose greater ground water con-
tamination risks. The increased use of
alternative triazine and nontriazine
herbicides from substitute weed con-
trol practices would lead to increases
in their concentrations in surface wa-
ter, with differential impacts by soil
and tillage type, sometimes ex-
ceeding the risk associated with
atrazine before a ban. '-.''_
With a triazine ban, the-overall -
econo:rniclosswouldbe$526inillion-
But under a triazine ban, human
health, risks from herbicide residues
in groundwater would decline sig-
nificantly. Health risk from herbi- .
cideexposureswoulddedine overall,
with exceptions for specific areas on
a snutller proportion of soils from
dicamba, bentazon, alachlor, and-
metolachlor and only under conven-
tional and conservation tillage. There
would be a major shift to the more
widespread use of the sulfonylurea
herbicides, resulting in concentra-
tions well below their human health
benchmarks. However, there are
major uncertainties associated with
these chemicals regarding pest resis-
tance and their hazardto aquatic and
nontarget terrestrial vegetationeven
at the extremely low concentrations
expected in surface water from run-
off or drift. .
Theconventionalparadigminad-
dressmg pesticide-related problems,
i.e.,bansorlicensingrestrictions,may
not b<2 the appropriate response for
the corn and sorghum herbicides, in
particular, or water-relatedprobiems
of pesticides in general without a
significant effort to change agricul-
tural practices that rely less on
synthetic pesticides. More geo-
graprucallytargetedmitigationmea-
sures short of bans may lead to supe-
rior environmental protection and
minimal impact on producers and
society on' a whole. . .
-------
ISs
-------
Contents
Introduction .'. ,
Methodology
Environmental Fate Component
Chemical Leaching..., .
Surface Water Runoff ..;..........
Air Transport
Agricultural Decision Component
Weed Competition
Input Substitution Model
Economic Behavior Model
Risk Characterization and Weighting
Results of Analysis of Atrazine and Triazine Bans.
Baseline Use of Triazine Herbicides
Economic Impacts —.,..,
Input Substitution Effects
Environmental Impacts
Aggregate Hazard Index...........:
1
Conclusion.......—
References.... —:.
j
S
...9
.11
.12
15
.15
.18
21
..23
:es.
seafAi'-ASXTt**? >• • •>;•
i
-------
Figures
figure 1. CEEPES sludy region for the atrazine analysis 2
Figure 2V The four components of CEEPES 6
Figure 3. Information flow in CEEPES '. 6
Figure 4. Process of melamodel development ..7
Figure 5. Flowchart of the major processes simulated by WISH 10
Figure 6. Example of WISH Output........ 11
Figure/. Percentage of corn acres treated by chemicals,
with and without primisulfuron and nkosulfuron —20
Figure 8. Acute surface water exposure for aquatic habitat in the study
area: conventional-, reduced-, and no-till practices 25
Figure 9. Hierarchal tree for alrazine policy impacts nodes 27
Figure 10. Weighting of hierarchical tree nodes 28
Figure 11. Aggregate Hazard Index Score for health and ecological impacts
weighted equally where exposure values are summed 29
Figure 12. Relative importance of health and ecological weights 29
Figure 13 a and b. Ordering of alternatives with different weights for
drinking water and dietary risks ...30
Figure 14 a and b. Plots of relative hazard index —..30
w®,
-------
Tables
Table 1. Models that comprise CEEPES -»8
Table 2. Current (Baseline) use of atrazine ,»•••— 16
Table 3. Percentage changes in crop acreages from
baseline for chemical restrictions ,—••• H'
Table 4. Percentage changes in per acre crop yields from ;
baseline for chemical bans ••«• 16
Table 5. Cost change per treated acre :. 17
Table 6. Acres treated in study region 17
Table 7. Aggregate economic (welfare) effects of atrazine 17
TableS. Price effects of atrazine and triazine
restrictions—selected commodities •»•••« 18
Table 9. Herbkide use in study region .-. 19
Table 10. Average application rate for tiiazines 19
Table 11. Changes in herbicide application rates with bans,
with and without the sulfonyhireas 21
Table 12. Exposure value distribution in surfuce water for
the three scenarios 22
Table 13. Herbkide-treated corn acres'resulting in 24-hour
surface water exceedances 24
Table 14. Relative impacts of policies on ground and surface waters 26
Table 15. Estimated percentages of acreage
under different tillage practices ,. !..,....:........: 26
>•*%?''
c..:;
III
-------
if
fc-X&K+r;
m
-------
Introduction
The United States Environmental
Protection Agency (EPA) regulates
pesticides through its Office of Pes-
ticide Programs under the authority
of the Federal Insecticide, Fungicide,
and Rodenticide Act (FIFRA) (7
USCA, section 136) and the Federal
Food, Drug, and Cosmetic Act
(FFDCA) (21 USCA, section 348).
FIFRA is a licensing statute that re-
quires that all pesticides used in the
United States be registered with EPA.
EPA may deny registration if the
requirements for registration have
not been met.
Meeting registration require-
ments entails-presenting evidence to
verify that, among others, "the pes-
ticide performs its intended func-
tions without unreasonable adverse
effect on the environment" when
used in accordance with widespread
and commonly recognized practice.
FFDCA authorizes EPA to establish
tolerances for pesticide residues that
remain on food or feed at the time of
marketing of the commodity.
Because most pesticides had been
registered before the introduction of
stringent data requirements, an
amendment to FIFRA in 1972 re-
quired that all registered pesticides
be reregistered in accordance with
new standards for registration—the
so-called "pesticide reregistration
program." An additional amend-
ment in 1988 called for the expedited
reregistration of pesticides to meet
current data requirements and test-
ing protocols.
EPA has traditionally responded
to evidence of unreasonable health
or environmental risks associated
with pesticides by refusing to regis-
ter a new pesticide or new uses of an
already registered pesticide; or by
either withdrawing the registration
of one or more uses of an existing
pesticide (in essence, banning the
pesticide) or restricting its use to a
certified pesticide applicator. In as-
sessing adverse effects onhealth and
the environment as a consequence
of its use, the Agency examines a
variety of exposures associated with
uses of the chemical and the crops to
which it may be applied. Given its
assessments of the intrinsic hazards
to humans and to animal and plant
species, it estimates the risk associ-
ated with each type of exposure: di-
etary, worker, drinking water,
ecological. A determination of sig-
nificant risk within at least one expo-
sure dimension puts the burden of
proof upon the registrant of the pes-
ticide to show that, in all its uses,.it
can meet EPA's risk standards or
that the risk, though greater than de
minimis, is offset by the economic
benefits that accrue to farmers and
society.
In conducting, the risk assess-
ment, EPA relies upon best available
information on how a pesticide is
being used, as well as on the pesti-
cides and practices that can serve as
substitutes in the event of loss of the
pesticide. Theresponsibility hasbeeri
upon the registrant to identify how
it could be used, as affected by label-
ling changes, and thecropsforwhich
it should be registered. A finding of
unreasonable risk usually leads reg-
istrants to request that all registra-
tions for crop uses thataremarginally
profitable be dropped in order to.
reduce the total exposure and hence
risk associated with all registrations
of the pesticide.
Generally the most profitable
uses' are also those associated with
use on major field crops. Therefore,
the array of pesticides available for
minoir use crops .(crops other, than
the major commodity cropslikecom,
wheat, and soybeans) has dimin-
ished, over time, while those uses
associated with the most significant
risks have been retained—those for
major field crops. This Joss of pesti-
cides available for minor-use crops
has allegedly inadvertently led to
greater reliance and use of a small
number of remaining pesticides reg-
istered for a particular use, resulting"
in problems of pest resistance and
greater loadings of these pesticides
into the environment. {Leonard
Gianessi, 1992].
The focus in EPA's Office of
Pesticide Programs has been pri-
marily on dietary and worker
sources of risks from pesticides—
-------
Agricultural Atrazlne Use and Water Quality
A CEEPES Analysis of Policy Options
risks that are clearly national in
nature. EPA's Science Advisory
Board in its Relative Risk Reduc-
tion Project1, however, recom-
mended that greater attention be
paid to ecological problems as-
sociated with pesticides. Also, the
recent promulgation of new
drinking water standards for pes-
ticides2 has resulted in greater
attention paid to health problems
associated with water exposures.
The current licensing para-
digm for the prevention of health
risks associated with pesticide
use may no longer be defensible
for m anagement of water-related
risks from agricultural uses of
pesticides. The variety of factors,
including weather, soil, and ag-
ricultural practices, that affect the
loadings of pesticides to surface
and ground waters requires a
very large amount of informa-
tion for national or state level
licensing decisions. The need to
generate and to evaluate the wide
range of information, in turn,
partly explains the delays in the
making of decisions by EPA and
hence the loss of public confi-
dence in the government regula-
tory program. Furthermore, the
geographic variability in risk is
likely to make national restric-
tions on use socially and eco-
nomically inefficient. This report
is an attempt to evaluate the ap-
propriateness of national ban/
no ban decisions for protecting
water quality from agricultural
uses of pesticides,-focusing on
triazine herbicide use on corn and
sorghum. It also illustrates the
usefulness of a new analytical
tool for multiobjective and mul-
tidimensional decision-making.
Given the need to take into
account the complexity of water
quality issues in assessing the
risks and benefits associated with
agricultural pesticide use, EPA
funded the development by Iowa
State's Center for Agricultural
Figure 1.
CEEPES study region for the atrazine analysis
and Rural Development (CARD)
of the Comprehensive Environ-
mental Economic Policy Evalua-
tion System (CEEPES). CEEPES
is an integrated modeling system
developed to estimate the eco-
nomic and environmental conse-
quences of alternative policies
affecting the use of pesticides.
CEEPES integrates diverse simu-
lation models comprising four
major components—policy, agrV
cultural and economic decisions,
fate and transport, and health and
ecological risk. " -
A major contribution of the
CEEPES system is the character-
ization of weed control technolo-
gies, resulting in the construction
of over 300 alternative weed con-
trol strategies for corn produc-
tion and over 90 strategies' for
sorghum. These strategies aim at
controlling both grasses and
broadleaf weeds. Each strategy
includes a primary and a backup
treatment (to deal with weeds
that survive the primary, treat-
ment); a set of herbicides applied
individually or in tank mixes; a
tillage practice (no till, reduced,
and conventional); chemical ap-
plication rates; an application
mode (broadcast, incorporated,
banded); a timing of application
(early pre-plant, pre-plant incor-
porated, pre-emergent, post-
emergent); and temporal
windows of application and ef-
fectiveness for both the primary
and the secondary strategies.
Thus, a ban on a particular herbi-
cide does not simply imply a
chemical-for-chemical substitu-
tion, but rather selection from
among an entire array of weed
control strategies that are poten-
tial substitutes. In addition, pro-
duction risk is incorporated into
the modeling system by simulat-
ing the impact of uncertainty re-
garding the weather on dates of
-------
Introduction
chemical application and the re-
sulting effectiveness, This ap-
proach is embodied in WISH
(Weather .Impact Simulation on
Herbicides). A more detailed de-
scription of WISH and its use in
decision making under uncer-
tainty is provided in Bouzaher et
al. (1992a,b).
The CEEPES study region in-
cludes the Corn Belt and Lake States
regions, plus portions of the North-
ern Plains region and five other
United States Department of Agri-
culture farm productionregions (Fig-
ure 1), which are henceforth refered
to, collectively, as the Midwestern
Cornbelt. The region includes more
than 80% of US.corn acreage and
30% of sorghum acreage. Details on
the CEEPES system are .provided in
the section on methodology and are
available in more detail in the Cen-
ter for Agricultural and Rural
Development's (CARD) Research
Memos 3 and 4.
The system was used to evalu-
ate two policies regarding the use
of the herbicide, atrazine: an atra-
zine ban onVcorn and sorghum
and a ban on the entire group of
triazine herbicides of which atra-
zine is a member. This latter ban
is henceforth referred to as a tri-
azine ban and includes bans on
atrazine, cyanazine, and simazine
for corn and sorghum use.
Though the policiesand thesimu-
lations apply only to the Midwest-
ern Cornbelt in this analysis, the
implications, given that the bulk of
production for corn and sorghum is
covered by the study, area,, apply
nationwide. Bans that apply region-
ally, without scientific justification,
would raise questions of equity for
producers in the region and would
not likely stand. A regional rather
than a national analysis was con-
ducted because it is very difficult to
calibrate the models for corn-pro-
ducing areas outside the Midwest
and because the information on the
risks and benefits of the impact of
policies on marginal producing ar-
eas would not likely significantly
affect the conclusions of the analysis.
Atrazine was chosen because it is
the most widely detected -pesticide
in surface and groundwater. Belluck
et al. (1991) note that the detection
rate is 10 to 20 times more frequent
than the next most frequently de-.
tected pesticide. The levels detected,
often exceeding the Federal drink-
ing water standard of 3 parts per
billion, (ppb), have led the US EPA
and state agencies to review policies
to control or ban atrazine use. CUTT
rentuse of atrazine in themidwestern.
states of the United States is esti-
mated, at 52 million ppunds of active
ingredient (USDA,-1991), account-
ing foir nearly 12 percent of the total
agricultural usage of pesticides in
the U.S. (US EPA, 1992a).
The results in this report are ag-
gregated over the entire study re-
gion, though most data-generated
by the analysis are available on a US
Department of Agriculture produc-
tion region, .state, watershed, or
county basis. A subsequent report,
in which policies involvinjg targeted
watershed or county restrictions on
use are analyzed, will present the
results by watershed and county.
-------
Agricultural Atrazine Use and Water Quality
A CEEPES Analysis of Policy Options
-------
Methodology
Environmental polities relating to
agricultureinherentlyreflectairade-
off between the need for production
practices that minimize the costs of
production and control of agricul-
tural chemicals that are introduced
into the environment by those prac-
tices. Besides producing bountiful
crops, agricultural activities can
sometimes lead to soil erosion, pes-
ticide and nutrient contamination in
surface and ground waters, worker
exposures to pesticides, pesticide
drift, and atmospheric transport.
Adverse impacts include losses of
ecosystemsand their functions,pub-
lic uncertainty about the safety of
the food and drinking water supply,
declines in plant and animal num-
ber and diversity, and reductions in
soil fertility.
On the other hand, regulations
or controls can have undesired eco-
nomic effects: incremental loss of
pesticides leading to increased costs
to producers, higher prices of agri-
cultural products for consumer
goods, and reduced export earnings
from agricultural products. Unin-
tended adverse environmental ef-
fects can even occur if policies to
protect ground and surface water
quality from pesticides result in in-
creases in soil erosion through
greater reliance on soil tillage and
hence sediment loadings to surface
waters. Therefore, it is essential that
analytical tools allow policy-makers
to evaluate policies and practices in
terms of both the needs of commod-
ity production and protection of the
environment.
CEEPES simulates risk-benefit
trade-offs associated with non-point
source pollution from agricultural
productions. Itlinks biophysical with
economic modelling systems that
have been integrated over the di-
mensions of time and space. Four
components comprising the concep-
tual structure, illustrated in Figure
2, provide the necessary flexibility
for model and policy integration:
policy specification, fate and trans-
port, agricultural decisions, and
evaluative criteria relating to envi-
ronmental and humanhealth bench-
marks and economic outcomes
(impacts). To ensure congruence of
temporal and geographic scale, ''ex-
periments" with calibrated geo-
physical process models produced
response surfaces that have statisti-
cal integrity and known experimen-
tal and sampling error.
The information flow and con-
figuration of the system is shown in
Figure 3. Policy and regulatory space
affects the range of strategies avail-
able to the producer. The system is
configured to allow simulation of
policy interventions restricting or
enhancing producer behavior with
regard to production decisions. The
Resource Adjustment Modelling
System (RAMS) is a linear program-
ming model that simulates the profit
maximizing decisions of producers.
Producers choose an optimal mix of
crop arid crop rotations, chemical
inpubi, labor, tillage, and other fac-
tors to maximize net return. The
Weather ImpactSimulation for Her-
bicides model (WISH) identifies the
most efficient weed control -strate-
gies for corn and sorghum based on
timing and method of application,
efficacy of chemical combination,
and tillage. Agricultural -Land
Management Alternatives \vithNa-
merical Assessment Criteria (AL-
MANAC) is a process model that
simulates crop growth, weed com- .
petition, and the interactions of man-
agement alternatives. The fate and
transport models estimate loadings
and concentrations of contaminants
in the various environmental media,
such as water and air. These concen-
trations art* then either summarized
directly as a ratio to health or envi-
ronmental benchmarks or as a fre-
quency that a benchmark is likely to.
be exceeded under a scenario.
The models that comprise the
.various, components are shown in
Table. 1. They are described in detail
below in conjunction with the com-
ponent in which they function.
As schematically shown in Fig-
ure 4, various factors influence the
domain of natural phenomena,, as
represented by the object at the top.
of the figure, to produce measurable
outcomes of policy interest. A com-
puter physical process model simu-
lates the phenomena for a specific
u tivit i-|y
rizes economic benefits and envi-
ronmental risk indicators to allow
comparisons of policy scenarios. Ag-
gregation of concentrations and risk
come in applying CEEPES to a large
region include the wide range of
variation in temporal and spatial
scales of different models requiring
between USDA and EPA with re-
gard to data specification and devel-
opment could improve accuracy for
policy-making at a smaller geo-
-------
Agricultural Atroiine Use and Wafer Quality
ACEEPESAnolysis of Policy Option*
Figure 2.
The four components of CEEPES
igrkultura! Atrazine Use and Water Quality
* CEEPES Analysis of Policy Options
• ?ablel. ^ , I 1
fllodels tnat comprise CEEPES -
ft •
Acronym
Name
Citation
Comment .
Economic Decision/Production Models
RAMS
WISH
ALMANAC
AGSIM
Resource Adjustment
Modeling System
Weather Impact Simulation
for Herbicide
Agricultural Land
Management Alternatives
with Numerical Assessment
Criteria
Agricultural Sector
Integration Model
Bouzaher et al. (1992)
Bouzaheretal. (1992)
Jones and OToole
(1986)
R.C.Taylor (1987)
Li.near programming model that
simulates profit-maximizing
behavior
Systems approach model for
weed control strategies •
Process that simulates crop .
growth and weed competition
t . •
Econometric model of national
& international markets for 1 0
' agricultural sectors
Environmental Fate & Transport. Components
RUSTIC
STREAM
B LAYER
PAL
Risk of Unsaturated/
Saturated Transport and
Transformation of Chemical
Concentrations
Stream Transport and
Agricultural Runoff of
Pesticides for Exposure: a
Methodology
Boundary Layer. Model
Point, Area, and Line
Source
Dean, etal. (1989)
Donigian, et al (1986)
McCorcle, M.D. (1988)
Peterson etal (1987)
Model that partitions mass of
pesticides into various media;
Groundwater model for
pesticides (root zone; vadose
zone) .
Surface water model for
pesticides
Atmospheric transport model
Short-range air transport model
- :
-------
Methodology
graphic scale, and facilitate the use
of CEEPES for other crops and
regions.
Environmental Fate
Component
Chemical Leaching:
Herbicide leaching in the soil root and
vadose zones was simulated with the
PRZM and VADOFT components of
the US EPA RUSTIC model, the
PRZM (Plant Root Zone Model) com-
ponent of RUSTIC partitions the mass
of the pesticide into amounts avail-
able for volatilization, runoff, and
leaching. The amount available for
leaching becomes the input into the
VADOFT component that moves the
mass from the root zone through the
vadose zone. We did not estimate
lateral flow with RUSTIC. Statistical
sampling of soil, climatic, pest man-
agement, and chemical parameters
representativeofthestudyregionwere
used as inputs to RUSTIC to estimate
groundwater concentrations. Thus
using statistical procedures, we select
a sufficient number of data for each
key variable, within the range of val-
ues for the variable within the study
region, that affect the concentrations
in groundwater (or other medium) to
achieve the desired level of statistical
accuracy andreliability.Weestimated
peak and average pesticide concen-
trations at depths of 12 and 15 meters,
which are assumed representative of
vulnerable,shallowgroundwater and
are typical depths for the water table
and rural domestic drinking water
wells, respectively, in the study area.
Surface Water Concentrations:
Pesticide loadings to surface runoff
from RUSTIC serve as input into the
STREAM methodology to estimate
chemical concentrations in surface
water. STREAM is a screening-level
tool for estimating in-stream solution
and streambed pesticide concentra-
tions. It is based upon 10-year simula-
tion runs of the HSPF (Hydrological
Simulation Program—FORTRAN
Uohanason, R.C., et al] river basin
model for representative watersheds
of four main crop producing areas.
STREAM estimates are generally
within a factor of ten of actual moni-
toring values.
AirTransporfc
Monitoring by the United States Geo-
logical Survey3 and others have found
pesticides in rainwater in areas where
they have not been applied. To evalu-
ate how the use of pesticides in agri-
cultural activities may affect
ecosystems at various distances from
the source of application, we have
included both a short- and long-range
air transport component to CEEPES.
Preliminary work, which is not
discussed in this report, has been per-
formed with the Boundary Layer
Model (BLAYER), a long-range nu-
merical model and PAL, a short-range
Gaussian plume model, to simulate
short and long-range air transport of
pesticides.4 Volatilization data are
obtained from RUSTIC, which is run
for 10 years of weather data, 4 soil
types, and 12 chemical properties.
Output data consist of the amount of
the chemical volatilized and the
amount available for volatilization, in
the top 0.5 cm of soil, on each day
following May 9, or the day on which
the chemical is applied, until June 30.
Agricultural
Decision
Component
Weed Competition:
The USDA .ALMANAC model is
used to simulate weed-crop compe-
tition. Statistical response functions
of yield impacts were developed
with ALMANAC based on a sanv- .
pling design of crop, weed, soil, cli-
matic, and pest management
parameters representative of the
study region. These response func-
tions are used in WISH. A flowchart
of processes in WISH, is shown in.
Figure 5.
In Figure 5, the major land re- ;
source area (MLRA) defines spatial .
areas by grouping of soil types.;.
MLRAs are aggregated up to a pro-
ducing area (PA) -corresponding
roughly to the boundaries of a hy-
drologic area or watershed through
a process of weighting. PAs are dis-
cussed in greater detail below. Pqst-
emerg;ent refers to whether or not
the herbicide is applied, before or
after the corn plant has emerged.
The effectiveness window corre-
sponds totheperiod of timein which.
the farmer can apply the herbicidje
and have it be effective in reducing
weed pressure. The variable weather
is the most important factor in deter-
mining the size of the effectiveness
window.
Input Substitution Model:
The WISH model simulates the effi-
cacy a nd cost of over 300 alternative
weed control strategies for corn and
over 90 for sorghum. It simulates
likely weed control management
based on herbicide efficacy, weather
conditions, timing of application and
its effectiveness, mode of applica-
tion, soil texture/targeted weeds,
and observed farming practices. An
example of output for one PA is
shown in Figure 6.
Am optimal strategy is one
which has high mean net revenue
per acre and low variance, i.e.,
less uncertainty regarding de-
sired outcome given the uncer-
tainties of weather. As shown in
Figure 6, the plot of the status
quo (SQ), or baseline, use of atra-
-------
Agricultural Atrazine Use and Water Quality
A CEEPES Analysis of PoGcy Opfions
Figure 5.
Flowchart of the major processes simulated by WISH
MLRAs and
associated weather
stations
Weather Generator Dally
Rt,Tt,Wt
Herbicide Strategy Table \—
Chemical costs
Chemical rates
Application costs'
Labor requirements
Labor costs
Cultivation requirements
Crop file:
•Planting data
•Tillage Information
Read Inputs |
I For each of over 300 herbicide strategies
Simulate over SO years |
lit
Ye«
Primary activity all
non-chemical?
Start primary application
Keep track of non-chemical
activity separately
No
Herbicide Strategy entirely
post-emergent?
While application window open:
•Check weather for field conditions
•Check for Incorporation or spray
•Record acreage treated
Yes
For Broadlands and
Grasses separately
Check weather for
planning and crop
emergence
|
While effectlveness
-------
Methodology
1 Figure 6.
Example of WISH Output
Plot of yield loss (buJacre) vs. cost ($/acre) for conventional tillage
herbicide strategies applied to a clay soil for one PA . : .
30.
~ ^ Triazineban .
2°! \ \ .-. . . : , - •
| : \ + V * + " •
-* i **=^~ — '•*-*- — •• > +
I \t** "* *+!** "*-^ . 4? + " '
•in- ^^ +*^ , *^- \ - "
I Atrazineban A \ \ +
\\ —Status quo *
0 10 20M 30 40
Policy Atrazineban Status quo Triazineban
Note: The lines represent frontiers of strategies tor each scenario that provide
the lowest expected yield loss tor a given expected cost
zine under conventional tillage,
which plots closest to the origin
(that is, the lowest yield loss at
the lowest cost), is superior to
that of alternatives that either
substitute for atrazine or the tri-
azines (AB & TB) or involve a
change in agricultural practice.
This is the consequence of the
relatively low current price of
atrazine and its reliability in con-
trolling weeds relative to substi-
tute pesticides and practices that
may involve a greater cost due to
increased labor requirements in
lieu of herbicide use.
Economic Behavior Model:
RAMS simulates themannerinwhich
policies affect the economic behavior
by translating' information on pesti-
cide policy into constraints on farmer
pest control strategies. It is a regional,
short-term, profit-maximizing, linear
programming model of agricultural
production, defined at the PA level.
PAsarehydrologicregionsrepresent-
ing aggregated subareas defined by
the Water Resources Council5. There
are 105 PAs in the continental United
States, excluding Alaska. RAMS esti-
mates the economicimpact of alterna-
tive agricultural and environmental
policies in terms of acreage planted,
rotation, tillage practice, chemical in-
put, net return, yield, and cost of pro-
ductionforallPAsinthestudyregion.
For this analysis of corn and sorghum
herbicides, 27 PAs were included.
These results serve as inputs into a
national econometric model, AGSIM,
developed by Robert Taylor of Au-
burnllniversity [Taylor (1987), (1991)
and Penson and Taylor (1992)], to es-
timate changes in measures of eco-
nomic welfare (producer income,
domestic consumer effect, and gov-
ernment outlays).
The AGSIM model ties together
econometrically estimated demand
and supply equations for major crops
and livestock through market clear-
ing identities. Thus, the model solves
for the set of crop and livestock prices
that simultaneously ensure that sup-
ply meets demand (i.e., dear all mar-
kets), in. a given year for given
exogenous factors, that is, factors ex-
traneous to the market, such as gov-
ernment policy. Due to the dynamics
of supply and .demand, market dear-
.ing prices are recursively obtained for
eachy ear simulated.Producer income
is calculated as the sum of net crop
income and net livestock income. The
domestic consumer effect is defined
as the sum of changes in outlay of
livestock consumers, domestic crop
food consumers, and the oil and meal
markets.Govemmejitoutlaysencom- •
pass the sum of farm program .pay-
ments (i.e., commodity deficiency
11
-------
Agricultural Atrazlne Use and Water Quality
A CEEPES Analysis of Policy Options
payments)andexpendituresforGov-
ernment milk purchases.
Risk
Characterization
and Weighting
The analysis used the ratio of pre-
dicted concentration to human
health or ecological benchmarks
(or, alternatively, the likelihood
that a Benchmark level would be
exceeded) as the measure of risk
of adverse impact resulting from
environmental exposures. The
greater this ratio is, the greater
the risk that we would predict
for exposure to a particular pes-
ticide. Any value approaching or
exceeding unity is of concern. The
decision-maker must decide, us-
ing whatever information is
available that is specific to a
chemical, whether or not the risk
depicted in the baseline scenario
is of sufficient magnitude to war-
rant action'. CEEPES can only as-
sist in determining whether or
not any action or policy to ad-
dress the risk in the baseline sce-
nario is likely to improve or
worsen the overall environmen-
tal impact.
Since pest control strategies
often call upon more than one
pesticide being used in a tank
mix and since different farmers
use different strategies and hence
pesticides, surface waters may
contain a mixture Of pesticides at
any given point in time. Ground
waters may also contain a mix-
ture of pesticides where a variety
of pest control strategies has been
used over a number of years. For
the purpose of comparing policy
alternatives, we characterize the
risk associated with exposures to
these mixtures of pesticides in a
particular medium, by adding the
ratios of predicted concentrations
to benchmarks for each pesticide
in the mixture, in accordance with
EPA guidelines on assessing risk
from mixtures of chemicals. A
final risk value for a category of
environmental impact for a par-
ticular policy, such as, human
health to ecological impact is ar-
rived at through a process of as-
signing wieghts on the basis of
relative importance. The method
by which we arrive at weights is
described below.-
Minimizing total environmen-
tal risk requires the comprehen-
sive evaluation of the fate of
chemicals as they move through
a medium, from one medium to
another, the manner and types of
exposures to receptor popula-
tions through contact with each
medium; and the adverse health
and ecological effects that are
likely to result. A pesticide may
leave little or no residue in food
and hence pose minimal dietary
risk to sensitive populations, but
nevertheless partition itself in ag-
ricultural runoff where it can con-
taminate streams and rivers and
pose aquatic risks. Conversely, it
can pose little or no likelihood of
contaminating water at all, yet,
through volatilization and drift
pose risks to agricultural work-
ers or nearby residents of towns
and even be transported great
distances to the surface of lakes
and rivers hundreds of miles
away. '
Balancing environmental
risks with the benefits that ac-
crue from the use of pesticides
therefore involves weighing the
risks associated with one medium
or to one receptor with those as-
sociated with other media or re-
ceptors and trading off benefits
for reductions in one or more
impacts. These trade-offs often
concern objects of dissimilar na-
ture or risks of different duration
or severity. One pesticide that is
particularly effective in a .specific
use and manner of application,
for example, may pose a threat to
workers. Yet its less effective
substitutes—were it to be banned—
threaten wildlife because the appli-
cation of the pesticides in the form of.
granules, for instance, increases the
likelihood that birds mistake them
for pebbles and eat them. Alterna-.
lively, a choice may be between an
acutely toxic pesticide, that can kill "
or injure a worker immediately or a
substitute that poses uncertain long-
term adverse health consequences
from minute exposures.
Where options or choices
present different types of risk.and
hence are not directly compa-
rable, value judgments must be
made in deciding their relatiye
importance. Should economic
gain be weighed against environ-
mental risk? How should threats
to birds.be weighed against risks
to humans or ecological habitats?
Consensus can more easily be;
reached on how these decisions
should be made if the. following
holds true: 1) technical and pro-
fessional estimates are clearly
separated from value judgments
and 2) value judgments are ap-
propriately made by policy-mak-
ers to whom the responsibility of
political judgment has been
given.
Weighting is not intrinsically
a scientific exercise, but rather a
political process that reflects per-
sonal values or group prefer*
ences. Nevertheless, science can
be brought to bear to facilitate
the process. In this analysis, we
demonstrate one method that the
field of operations research has
recently provided by which such.
weighting may occur. The
weighting procedure uses the
12
-------
Methodology
Analytical Hierarchy Process
(AHP) as operationalized in the
computer program Expert
Choice*. A model is specified that
relates key risk variables or cat-
egories of impacts to the goal of
an analytical exercise. In this case,
the goal would be least total en-
vironmental impact or risk. The
relationshipjjetween variables is
decided by weights. The weights,
or estimates of relative impor-
tance or user preference, are elic-
ited from the user through
pairwise comparisons of options.
Information, such as monitoring
and field data on environmental
levels of contaminants and
knowledge of isiiirface and
groundwater impairments, is
drawn upon in making the quali-
tative comparisons. 3umming
weighted risk variables produces
an aggregate environmental haz-
ard index value that is then com- •
pared against economic benefits.
1 US EPA, Science Advisory Board, Reducing Risk: Setting Priorities and Strategies for Environmental Protection, SAB-EC-!HM)21,Sejttemberl991.
1 US EPA, Office of Drinking Water, Phase // Fact Sheet: National Primary Drinking Water Regulations for 38 Inorganic anil Synthetic GteanJc
Chemicals, January 1991. . .
1 '
3 Memorandum from Donald Goolsby, U.S. Geological Survey, to Andrew Manale, US EPA, see also USGS Selected Pajvers on Agricultural
Chemicals in Water Resources of the Midcontinental U.S., Open file report 93-418. '
4 Work conducted under contract by Michael McCordle, SAIC, Springfield, Virginia. . . - :
5 U. S. Water Resources Council, Water Resources: Regions and Subregions for the National Assessment of Water and Related Land Resources,
U.S. Govt. Printing Office, Washington, D.C. (1970). ' . .
6 Expert Choice™, based on the Analytic Hierarchy Process, by Decision Support Software, Inc., 4922 Ellsworth AWJ., PittsJmrgh, PA.
-------
Agricultural Atrazfae Use and Water Quality
A CEEPES Analysis of Policy Options
14
-------
Results of Analysis of
Atrazine and Triazine Bans
Baseline Use of
Triazine Herbicides
The pesticide usage data used to
calibrate, that is, to define the bound-
aries of, baseline herbicide use in
CEEPES were obtained from Re-
sources for the Future (RFF)
(Gianessi and Puffer, 1991). Baseline
atrazine use estimated by CEEPES is
approximately 39 million pounds ac-
tive ingredient (a.i.) on corn and
about 3.3 million pounds a.i. on sor-
ghum (Table 2). The use of all triaz-
ines combined is about 60 million
pounds a.i. for corn and 3.3 million
pounds a.i. for sorghum. The RFF
and CEEPES estimates differ be-
cause, whereas the RFF data repre-
sent a single snapshot of herbicide
use, CEEPES provides estimates for
any given year based uponmeteoro-
logical and other' factors affecting
use.
Economic Impacts
For both an atrazine and a triazine
ban> corn acreage and corn yields
are predicted to decline, with soy-
bean acreage and soybean yields in-
creasing (Table 3 & 4). For the
atrazine ban, corn acreage would
decrease by 3 percent from the
baseline of 72.6 million acres, while
soybean acreage would increase by
4.1 percent from the baseline of 44.2
million acres. Sorghum acreage in-
creases slightly (0.7 percent for an
atrazine ban and 1.9 percent for a
triazine ban) offsetting some of the
production loss due to yield de-
creases. Corn yield decreases by 2.8
percent for an atrazine ban and 4.1
percent for a triazine ban (Table 4).
Yield decreases for sorghum are
much larger at 5.7 percent and 6.8
percent for the two scenarios.
The cost of weed control per acre
would increase .between $6.00 and
$8.00 for corn and less than $1.00 for
sorghum (Table 5). In corn, banning
atrazine requires the use of more
costly weed control strategies to
achieve a comparable level of con-
trol. In sorghum, banning atrazine
leads to heavier reliance on compa-
rably costly, but less effective strate-
gies. Table 6 summarizes the changes
in acres treated with triazines for
both corn and sorghum. We sepa-
rated the atrazine-treated acres ac-
cording to whether or not atrazine
would be applied at a rate of no
more than 1.5 Ib/acre rate (50.3 mil-
lion acres of corn in baseline). [The
distinction between'rates of 15 lb/
acre and rates greater than that is the
consequence of heavy clay soils gen-
erally receiving greater application
rates than lighter sandier soils. Thus,
we predict a bimodal distribution of
atrazine application rates that has
beenconfirmedby surveys.] We note
a significant increase in the use of
cyanazine for sorghum and simazine
for corn under an atrazine ban. This
can be partly attributed to our cur-
rent asisumption of no significant
crop injury from these" two triazine"
herbicides. Convincing evidence
from weed scientists may cause us
to change this assumption.
The national economic welfare
measures (changes in producer in-
come, domestic consumer, effect, and
governmentoutlays)-associatedwith
yield and cost impacts of an atrazine
or triazine ban-in the study area were
estimated using the AGSIM Mode}.
Because of the importance of the .
study area to corn and sorghum pro-
duction (more than 80% of corn acre-
age and 30% of sorghum lies in the
study area), changes in yield and
cost affect prices nationally. Thus,
though corn or sorghum production
practices outside the study area
would not be affected by the poli-
cies, growers in these areas would •
be affected by changes in prices paid
for the commodities. Because most
of the corn acreage is captured by
the analysis-, we would not expect a
significant difference inresults if the
policies were applied nationally. We
present both short-term (1993-96)
and long-term (2005-2008) annual
economic welfare effects for the na-
tion in Table 7. : .
In 'the short term, the average an-
nual decreases in total economic wel-
fare for the nation would .be about
$365 million under an atrazine ban
and $526 million under a triazine ban.
With tin atrazine ban, crop producers
in the Corn Belt bear a large share of
IS
-------
Agricultural Atrailne Use and Wafer Quality
A CEEPES Analysis of Policy Options
• Table 2. j j. |
Current (Baseline) use of atrazine and all triazines in the study region (in million Ibs
prop
Com
Sorghum
Chemical
Atrazine
All Triazines
Atrazine
All Triazines
RFF, 19911
39.9
58.7
6.3
6.3
CEEPES2 Baseline
38.9
60.7
3.3
3.3
O.L)
NAPIAP Study*3
50.6
72.3
4.1
4.1
•NAPJAP(1992). .'•••-
'V«Ju« report odt for tho CEEPES study region. .
^•Juoroportsd is for the CEEPES study region. , . . - . - - -
*Vdua raportad letor 12 iridwestem states. "
Percentage changes in crop acreages from baseline for chemical restrictions
1
Corn .
Sorghum
Barley
Cotton
Hay
Oats
Soybeans
•Wheat
Atrazine
Ban
-3.0%
0.7%
0.0%
0.0%
-1.6%
1.3%
4.1%
2.6%
Triazine
Ban
-2.7%
1.9%
0.0%
0.0%
-0.0%
2.3%
3.9%
•1.1%
Percentage changes in per acre crop yields from baseline for chemical bans
6
Corn
Sorghum
Barley
Cotton
Hay
Oats
Soybeans
Wheat
Atrazine
Ban
-2.8%
-5.7%
-0.006
0.00%
2.3%
0.29%
0.20%
-1.72%
Triazine
Ban
•4.1%
-6.8% • .
-0.007 ."•'.
0.00%
.95%
-0.69% '
0.24%
-0.15%
16
-------
Results of Analysis
• TableS. : i II 1
Cost change per treated
Corn
Sorghum
acre
Atrazine Ban
$6.70
$0.64
Trlazine Ban
$8.25
$0.12 .
• Table 6. ! ! I I I I
Acres treated in study region
Current Use in CEEPES Percent Change
Atrazine (>1. 5 Ibs/acre):
Atrazine (<1.5 Ibs/acre):
Cyanazine
Simazine
Corn
Sorghum
Corn
Sorghum
Corn
Sorghum
Corn
Sorghum
(mil. acres)
22.9
1.1
50.3
2.1
. 39.4
<0.1
5.8
Atrazine Bain
-100%
-100
-100
-100
-46
>200
>200
Triazine Ban
-100%
-100
-100
-100
-100
-100 .
-100
Aggregate econonk (welfare) effects of atrazine and triazine bans (million $)
Economic Effects
Atrazine Ban
Short Term Long Term
Triazine Ban
Short Term • Long Term
Producer Income
Domestic Consumer Effect
Government Outlays
-531
-406
(-572)
-312
-421
(-18)
-673
-548
(-695)
-423
-553
(-21)
Total Economic Effect -365 -715 -526 -955
Source: AGSIM
Note: Producer income is the change in the portion of net revenues for the nation that accrues to producers of com and
sorghum, domestic consumer effect is the change in consumer surplus for consumers in the US as a consequence of
changes in the price of com and sorghum, and government outlays is defined as the value of government transfer
payments that accrue to producers through price support programs for corn and sorghum. .
-------
Agricultural Atrazine Use and Water Qualify
A CEEPES Analysis of Policy Options
the burden. Producer income from
cropsis reduced by $234millionin the
region. For a triazine ban, some of the
losses in the Corn Belt are offset by
higher com prices and the loss'in pro-
ducer income of $168 million is less
than With the atrazine ban. Some re-
gions within the studyareawouldsee
an increase in producer income, par-
ticularly those where little corn and
sorghum are grown. Com and sor-
ghum producers outside the study
area would benefit from the increase
in prices for the two commodities.
Under the tworestrictions,significant
short-term decreases occur in govern-
ment expenditures, while losses cc-
curinnet lives tockincomeduemainly
to the combined effect of a decrease in
com production and an increase in
com and sorghum prices.
Long-term impacts may not be
as meaningful for the current analy-
sis because no information was in-
cluded on new, potentially more
effective, weed control technologies
like biological controls or new chemi-
cal substitutes. Under the current
assumptions, however, total eco-
nomic impacts would be of the same
magnitude as in the short run.
. Table 8 shows the corresponding
commodity price effects both for the
short and the long run. In the short
run, prices increase for corn (between
6.4 and 8.4 percent), sorghum (be-
tween 10.4 arid 12.2 percent), all hay
(1 percent), and hogs (0.8 to 1.1 per-
cent); prices decrease for soybeans
(about 1 percent), oats, wheat, and
barley. In the long term, price im-
pacts would be lessened, particu-
larly for corn, sorghum, and soy-
beans.
Input Substitution-Effects
The use of herbicides under the var i- •
ous scenarios is indicated by the dis-
tribution of corn and sorghum acres
treated by different herbicide strate-
gies (Appendix B). In the baseline,
more than 65 percent of corn acres
and more than 60 percent-of sor-
ghum acres are treated with a mix of
strategies containing atrazine. Un-
der an atrazine ban, more than 57
percent of corn acres and more than
15 percent of sorghum acres would
be treated with a mix of strategies
containing at least one triazine her-
bicide. Under a triazine ban, 27 perr
cent of corn acres and 9 percent of
sorghum acres would be treated with
• TableS. |; !
Price effects of atrazine and triazine restrictions-
selected commodities (% change)
. • Atrazine Ban
Price Effects
Corn
Sorghum
Soybeans
Oats
All Hay
Wheat
Barley
Hogs
Cows
Broilers
Turkeys
Eggs
Milk
Catf
Short Term
+6.4
+10.4
-1.2
-0.8
+0.8
-0.8
-0.4
+0.5
-0.1
+0.2
+0.3
+0.2
0.0
-0.1
Long Term
+2.6
+6.0
-0.5
-0.5
-0.0
-0.3
+0.4
+0.5
-0.1
+0.4
+0.6
+0.4
+0.1
-0.2
Triazine Ban
• Short Term
+8.4
+12.2
-1.3
-0.9
+1.1
+0.4
-0.7
+1.0
•0.2
+0.3
+0.4
+0.2
0.0
-0.2
Long Term
+3.5 '
+6.7
-0.1
-0.6
-0.0
. -0.3
+0.1
+1.3
-0.1
+0.6 . '-
+0.8 /
+0.4 •'.'..'
+0.1
-0.3.
18
-------
Results of Analysis
• Table 9. ! i i : I 1
.Herbicide use in study region -
(.
:•'•'• Atrazine:
All Triazines:
Non-triazines:
_
* All Herbicides:
Corn
Sorghum
Corn
Sorghum
Corn
. Sorghum
Corn
Sorghum
Current Use in CEEPES
(mil. Ibs. a.l.)
38.9
3.3
60.7
3.3
53.7
9.2
112.7
12.5
Percent Change
Atrazine Ban
-100
-100
27
-84
97
31
60
1
TrlazIneBan * .
-100
• -100
-100 .- -
-100 •
>200
31
49
-4
I
• Table 10. : I I
Average application rate for triazines
Cyanazine: Corn
Sorghum
Simazine: Corn
Change (Ibs. a.i.)
Atrazine Ban
0.94
-0.03
1.34
Change
Atrazine Ban
•249%
-4%
1313%
rotary hoe and row cultivation as
the main strategy. In addition, more
than 50 percent of corn acres would
be treated with strategies involving
alachlor, metolachlor, butylate,
EPIC, dicamba, and 2,4-D, and more
than 90 percent of sorghum acres
would be treated with strategies in-
volving alachlor, metolachlor,
dicamba, 2,4-D, and propachlor.
The newly registered, low dosage
sulfonylureaherbicides,nicosulfuron
(Beacon) and primisulfuron (Accent),
which are used in ounces per acre and
are believed to be of much less human
health consequence than other corn
and sorghum herbicides, could possi-
bly play a very significantrole in weed
control strategies in lieu of atrazine or
the other triazines.Nevertheless, there
is concern regarding rapid develop-
ment of weed resistance, crop injury
under certain conditions, and offsite
adverse impacts on nontarget crops
and plants. Thus, in light of the uncer-
tainty regarding the extent to which
farmers will accept the risks associ-
ated with theiruse,weexamined weed
control strategies with and without
them. The use of nicosulfuron and
primisulfuron in backup strategies
increases considerably when atrazine
and all triazines are banned. If
nicosulfuron and primisulfuron were
not allowed, as illustrated in Figure 7,
atrazine or triazine bans would result
in a much greater reliance upon
nontriazine herbicides that are gener-
ally applied post-emergent. Post-
emergent backup strategies that
include bentazon, bromoxynil,
pendimethalin, dicamba, and 2,4-D in
various combinations "would be sub-
stituted.
We summarize herbicide use in
the study region in Table 9. With an
atrazine ban, total triazine use in-
creases by 27 percent in corn and
decreases by 84 percentin sorghum.
The average application rate for
cyanazine and simazine cm corn in-
creases by 249 percent and.133 per-
cent, respectively, given an atrazine
ban (Table 10), whichtranslates into"
an increase in active ingredients by
0.94 libs and 134 Ibs per acre for
cyanazine and simazine. Inaddition,
we observe large increases in
nontriazine and total herbicide use
under both policy scenarios because
the acres treated by nontriazines in-
creas€',andthesubstituted weed con-
trol strategies, with the exception of
-------
Agricultural Atrazine Use and Water Qualify
A CEEPES Analysis of Policy Options
Percentage of com acres treated by chemicals, with and without primisulfuron and nicosulfuron
With Accent and Beacon , • |,~1V :/.,;. Without Accent and Beacon
Atrazlne
Atrazine <1.5
Nicosulfuron
Dlcamba
••••
Primlsutfuron
•••IMBI
BHOmESOCJOl
Cyanazlne
Bromoxynil
Bsntazon
MQtolachlor
Alachlor
Slmazlne
9B3XS3
Pendimethalin
Propachtor
Qlyphosate
EPIC
mmsa
2,4-d
0 10 20 30 40 50 60 70 80 90
Baseline I
Atrazine
Atrazine<1.5
Nicosulfuron
Dicamba
Primisulfuron
Cyanazine
Bromoxynil
Bentazon
Metolachlor
Alachlor
Simazine
Pendimethalin
Propachlor
Glyphosate
mtz • .
EPTC
2,4-d
0 10 20 . 30 40 50 60 70 80 90
Atrazine ban E
20
-------
Results of Analysis
• Table 11. 1 1 i 1
Changes
with the *ul
'.%r
SQ 0.98
AB 0.00
TB 0.00
without the
SQ 0.93
AB 0.00
TB 0.00
Note:Atr = at
metolachlor, i
50 years of w
applied when
in herb'
Ifonylur
atr
n rate
Pri
0.001
0.002
0.003
0.00
0:00
0.00
,dto = d
idimethe
therdoe
s with
cya
0.39
1.35
0.00
0.33
1.10
0.00
camba,
ilin, pro
s not pe
bans, i
bro
0.01
0.01
0.02
0.01
0.02
0.02
pri = pri
= propa
rmit app
with ai
ben
0.03
0.04
0.04
0.03
0.05
0.05
misulfur
chlor, gt
lication
id witl
met
0.78
0.59
0.48
0.77
0.59
0.47
on, cya
^=giyp
n every
lout th
ala
0.87
0.60
0.54
0.87
0.61
0.54
= cyana
losate,
year, th
e sulfo
sim
1.25
2.22
0.00
1.30
2.16
0.00
zine, bro
ala = ala
e values
nyluroi
pen
0.11
0.74
0.00
0.03
0.33
0.00
= brorn
chlor. R
are low
IS
pro
1.34
0.00
0.00
10.17
0.00
0.00
Dxynil, b
ates ref
er Ulan 1
gly
0.00
0.00
0.71
0.81
0.76
0.75
en = be
ectthe
he avei
EPTC
2.86
2.81
-2.82
.2.87
2.82
2.83
ntazon,
average
age rate
2.4D
0.07
0.23
0.23
0.13 .
022
0.20
met-
use over.
that is '
those employing the sulfonylureas,
entail relatively high application
rates.
Table 11 provides more detail on
the changes in application rates for
15 different herbicides for the Corn
Belt (Appendix A lists these 15 her-
bicides by their chemical and trade
names).
An atrazine ban causes a rate
increase for both cyanazine, si-
mazine, and other chemicals such as
dicamba, bromoxynil, bentazon,
pendimethalin, and 2,4-D. And
though cyanazine application rates
increase significantly, the total acre-
age upon which it is used decreases.
Thus, total pounds of pesticide use
would be a misleading measure of
water quality in this case. Total
pounds used would increase only
slightly yet loadings to surface wa-
ter in some 'areas, as will be dis-
cussed later, increase very
significantly.
In sorghum, banning atrazine
leads to heavier reliance on compa-
rably costly, but less effective strate-
gies. Average application rates for
other triazines also increase.
We provide a summary of shifts
in rotations in the CEEPES study
region in Appendix C. The major
rotation shifts occur between con-
tinuous corn, which decrease 16 and
16.5 percent under an atrazine and
triazine ban, and corn-soybean rota-
tion, which increase by 11 and 10.4
percent under the two restrictions.
These changes are importantbecause
they occur for the two most frequent
rotations. By shifting to a corn-soy-
bean rotation, savings in nitrogen
fertilizer and insecticide applications
are possible, which is a benefit to
water quality.
Environmental Impacts
Environmental indicators complete
the picture of the societal impacts of
an atrazine and a triazine ban in the
study region. Since a single average
indicator of water quality across the
study region would be almost mean-
ingless, we present results indicat-
ing both "relative risk" to humans
and aquatic life, and the spatial dis-
tribution of these indicators identi-
fying the most vulnerable soils. In
addition, results are separated by
tillage, surface water and ground-
water, and chemical. •
The peak and average chemi-
cal concentration levels found in _
surface and groundwater are
transformed into a unitless mea-
sure of risk that we call exposure
value. Pesticide-specific bench-
marks for human health and.
aquatic habitat are used to weight
the relative importance of pesti-
cide concentrations.
The term "exposure value" is
used to prevent confusing such val-.
ues with estimates of absolute risk.
Uncertainties in the data and in the
methodology do not allow the gen-
eration of precise point estimates or
probabilities. Instead, the purpose
of the estimates is solely for screed
ing- level comparison of policies and
practices arid serving as rough indi-
cators of water quality.
Using a benchmark for environ-
mental hazards, such as drinking
water Maximum Contaminant Lev-
els (MCLs) for long-term exposures
and 10-day Health Advisories for
short-term exposures, we calculate
the exposure value for each chemi-.
21
-------
Agricultural Atrazine Use and Water Quality
A CEEPES Analysis of Policy Options
cal in the following way for both
peak and average long-term levels:,
Expoture Value (hazard weighted exposure) -
predicted concentration
environmental benchmark
The exposure value normalizes con-
centration levels, thereby allowing
us to compare risks across herbi-
cides. If the exposure value exceeds
unity, the concentration exceeds the
benchmark and warrants further in-
vestigation; it does not necessarily
indicate that there is a significant
environmental impact. A chemical
detected in ground or surface water
represents a greater risk the larger
the exceedance of the benchmark.
Since the values represent averages
of peak and long-term values over
large areas, a value below the bench-
mark, conversely, does not indicate
that there is no problem at specific
sites.Notethatmore reliance should
be placed on relative differences be-
tween exposure values than on ab-
solute values [US EPA, (1992b)].
Table 12 presents the human
health exposure values for surface
water from peakloadings, by chemi-
cal and tillage, under baseline use
and under an atrazine ban and a.
triazinebanincorn production. Each
line in this table represents.the per-
centofsoiltypespresentinthestudy.
area that are associated, with high
levels of pesticides in surface water
runoff that could result in concen-
trations of possible health concern.
For example, the first rpto in Table
1 — ^ — : '-~. : 1
• Table 12. I ! 1
Exposure distribution in surface water for the three scenarios
(% soils wilh concenlrations exceeding EPA Benchmarks)
Conventional tillage Reduced tillage No-tillage
BASELINE
Atrazine
Atrazine <1. 5 Ib.
Dteamba
Cyanazlne
Bentazon
Metolachlor
Alachlor
Simazine
Propachlor
Atrz_S!ow_Decay
Atrz_Fast_Decay
ATRAZINE BAN
Dtcamba
Cyanazine
Bentazon
Metolachlor
Alachlor
Simazine
TRIAZ1NE BAN
Dicamba
Bentazon
Metolachlor
Alachlor
43.10
14.75
18.14
24.91
2.39
3.01
20.83
86.95
6.65
28.31
11.73
35.08
37.57
67.74
0.06
4.70
96.68
26.66
51.41
0.00
0.67
Note: Bromoxynil, Butylate, Glyphosate, Nfcosulfuron,
had zero probability of exceedance
42.87
15.80
0.00
0.92
45.90
8.91
32.65
67.66
38.44
- 26.06
0.01
i.21
90.15
0.00
26.93
31.61
42.57
5.60
84.96
13.59
0.00
Pendimethalin,
8.08
2.38
0.00
, . 0.00
0.00
. 0.00
0.00
40.12
••
4.26
1.13
0.00
0.00
•• , •
12.63
18.28
40.05
0.00 "
••
o.oo • .
0.00
Primisulfuron, and 2,4-D
in all three scenarios.
22
-------
Results of Analysis
12 shows that atrazine concentra-
tion levels exceed the short-term
benchmark of 100 ppb (the drinking
water Health Advisory level for
short-term exposure) in 43,43, and 8
percent of the soils in the study re-
gion, if the soils were cultivated un-
der conventional tillage,reduced till,
or no till systems, respectively. Un-
der an atrazine ban a higher propor-
tion of soils would lead to loadings
to surface resulting in chemical con-
centrations exceeding the bench-
mark under all three tillage systems
(e.g., dicamba, cyanazine, simazine,
and bentazon with conventional till-
age; dicamba, cyanazine, and
metolachlor with reduced tillage;
and metolachlor and alachlor with
no till). Note that for groundwater,
all average concentrations are well
belowthe long-term exposurebench-
marks for all soils and all tillage sys-
tems, under both an atrazine and a
triazineban. Thus, this information
could be used in targeting areas,
given the prevalence of certain soil
types, where ground or surface wa-
ters may be vulnerable to herbicide
leaching or runoff. AppendixDpro-
vides the details of predicted con-
centration and exposure levels for
each chemical in the study regionfor
the three scenarios, in corn
production.
An alternative way to use
CEEPES information to identify
practices associated with surface
water problems with herbicides is to
examine the percent of total herbi-
cide-treated corn acreage in a tillage
system resulting in die exceedance
of a benchmark. As Table 13 indi-
cates, exceedance of the atrazine
benchmark is currently more likely
to occur as a consequence of weed
control on land under reduced till-
age than under conventional tillage.
The opposite is predicted for
cyanazine and simazine. Banning
atrazine would lead to the predic-
tion that more land under reduced
tillage would be associated with
cyanazine exceedances than under
conventional tillage, whereas si-
mazine use under conventional till-
age would lead to the greatest
percentage of land associated with
high levels of herbicide runoff. Ban-
ning the triazines would greatly re-
duce the amount of land associated
with exceedances of surface water
benchmarks corn herbicides.
As shown in Figure 8, the ex-
posure values for aquatic veg-
etation—which are used as
indicators of ecological impacts—
from corn production for the two
policies tend to be high. The ma-
jority of aquatic exposure values
exceed the aquatic benchmarks,
which have either been only pro-
posed as standards by EPA or
have been derived according to
EPA guidelines, often by more
than a factor of more than 20. As
mentioned before, these numbers
are to be used for comparison
purposes, not as estimates of ab-
solute risk.
Aggregate Hazard Index
Developing an estimate of impact
per medium:
The modelling results, as well as
monitoring data1, indicate that hu-
mans and aquatic flora and fauna
are potentially exposed to a complex
mixture of pesticides in water sup-
plies. Adverse impacts associated
with policies and agricultural prac-
tices are therefore, for the sake of
comparisons, characterized in terms
of the aggregate exposure. To evalu-
ate theeffect of exposure to mixtures
of pesticides in each medium, we
treat the exposure values in each
medium in two ways: 1) the ratio of
predicted pesticide concentrations
to benchmark values are summed
across all pesticides by acreage un-
der different tillage practices for each
policy option and presented for both
acute and chronic exposures and
2) the highest exposure value is se-
lected from among those of all pestir
cides predicted in a medium by
tillage practice for both types of ex-
posures (Table 14). This is done in
order to derive a range of possible
impacts since little is known about
the additivity or synergism of mul-
tiple exposures. The weighted sum, .
or hazard index for a particular me-,
dium and health or ecological end- ,
point, was arrived at by multiplying
the percentage of total acreage de-
voted to a type of tillage practice
(Table 15) by either the sum of the
exposure values or the highest expo-
sure value in the medium. Acute
impacts derive from peak concen-
trations of herbicides averaged over
the study area and compared to
benchmarks for health or ecological
risks from short-term exposures and
chronic from averageconcentrations
compared to benchmarks for long-
term exposures.
Even though conventional till-
age is practiced on far more land
than reduced or ho till, more herbi-
cides are typically substituted for
less tillage to control weeds. Thus, a
shift out of conservation tillage from
an atrazine or triazine ban would
result in less intensive herbicide use
overall but aggregate use would in-
crease as more corn is grown under
conventional tillage. The surface and
ground wafer impacts from pesti-
cide use associated with conven-
tional tillage would also tend to
increase. •
An atrazine ban would reduce
the impact of long-term (chronic)
exposures from pesticide use in corn
production on ground water. Acute
risks could slightly increase with the
use of other triazine herbicides that,
though less persistent, are greater
leachers. A triazine ban would sig-
nificantly reduce both acute and
chronic risks as pesticides that tend
23
-------
Agriculture! Atrazine Use and Water Quality
A CEEPES Analysis of Policy Options
•Table 13. ' ! i! 1
Herbicide-treated corn «res resulting in 24-hour surface water exceedances:
exceedance relative to total cbra acres by tillage
Conventional tillage
....
BASELINE
Atrazine
Atra<1.5
DIcamba
Cyanazlne
Bentazon
Metolachlor
Alachtor
Simazlne
Propachtor
Atrz_Stow_Decay
Atrz_Fast_Decay
ATRAZINE BAN
Dtepmba
Cyanazlne
Bentazon
Metolachlor
Alachtor
Simazlne •
TRIAZINE BAN
DIcamba
Bentazon
Metolachtor
Alachtor
Note: Bromoxynil,
Acres
1,495,000
576,500
537,300
2,323,800
66,900
51,000
960,000
452,500
700
NA
NA
688,000
369,400
520,000
4,600
715.000
• 7;106,000
379,000
336,500
0
102,000
Buctril. Glyphosate,
Pct(%)
16.8
5.0
6.3
25.7
1.0
0.5
9.7
34.6
0.2
28.8
11.2
10.0
. 25.6
10.7
0.04
6.6
79.4
8.5
11.1
0
0.5
Nicosulfuron,
Reduced Tillage
Acres
555,600
3,431,900
0
222,000
1,162,600
84,700
9,000
354,000
1,800
NA
NA
645
3,966,400
0
600,000
471,000
3,500
36,000
11,200
137,300
Q
Pendimethalin,
Pct(%)
33.8
28.3
0.0
2.4
17.5-
4.2
1.6
29.3
2.4
56.9
0.0
0.05
44.1
0.0
12.2
12.7
2.2
1.1
4.2
4.3
0.0
Primisulfuron,
Acres and percentage of
No Tillage
Acres
51,000
34,600
0
0
0
0
0
36,400
0
.NA
NA
0
0
0
23,200
21,800
92,000
0
0
0
0
Pct(%)
•
6.8 ' .
2.8
0.0
"0.0
0.0.
0.0
0.0
10.6 .
0.0
8.6
2.3
0.0
0.0
6.0
7.3 .
6.8
13,6
0.0
0.0
0.0
0.0
and 2,4-D had zero exceedances.
to leach less or are less persistent are
substituted for atrazine and the other
triazines. There would be more till-
age of the soil to control weeds and
lesscornwouldbegrownthatwould
require atrazine use under bothbans.
For surface waters, the acute im-
pact on health may actually increase
with an atrazine ban with little
change for chronic exposures to her-
bicides. With a triazine ban, the im-
pact on health would diminish
dramatically for both short- and
long-term exposures. The combina-
tion of herbicides in surface waters
suggest a more pronounced impact
on health of the ecosystem with a
hazard index for baseline use sub-
stantially above unity. Banning atra-
zine may, in fact, worsen the impact,
and banning the triazine herbicides
altogether suggests that the health
of the ecosystem could actually de-
cline further.
The impact of the policy options
on soil erosion was examined under
twoscenarios—with and without the
low-dose sulforiylurea herbicides1—
in light of the relatively little agricul-
tural experience with this new class
of herbicides. Atrazine, in part be-
cause of its persistence, is very effec-
tive in the control of weeds,
especially where the soil is not tilled
as in no-till corn production. With-.'
out atrazine, the amount of no-till
24
-------
Results of Analysts
Figure 8.
Acute surface water exposure for aquatk habitat in She study area:
conventional-, reduced-, and no-till practices
Bf^.gss^^:*..-^^^^^
A. Conventional tillage
Exposure relative to aquatic habitat benchmark
~^ggg||*f»
", X
Alachlor
Atrazine
Atrazine <1.5
Bentazon
Bromoxynil
Butylate
Cyanazine
Dicamba
Metolachlor
Nicosulfuron
Primisulfuron
Propachlor
Simazine
2,4-d
) 20 40 60 80 1OO 120 140 160
±-.
•*
Policy
Baseline
^mmmm^
Atrazine Ban
Triazine Ban
Alachlor
Atrazine
Atrazine <1.5
Bentazon
Bromoxynil
Cyanazine
Dicamba
Metolachlor
Nicosulfuron
Primisulfuron
Propachlor
Simazine
2,4-d
O 1O 20 30 40 5O
B. Reduced tillage
Exposure relative to aquatic habitat benchmark
Alachlor
Atrazine
Atrazine <1.5
Bentazon
Bromoxynil
Cyanazine
Dicamba
Metolachlor
Nicosulfuron
Primisulfuron
Simazine
2,4-d
10
15 2.0
C. No-tilD
Exposure relative to aquatic habitat benchmark
25
-------
Agricultural Atrazine Use and Wafer Quality
A CEEPES Analysis of Policy Options
• Table 14. i 1
Relative Imparts of policies on ground and surface waters
POLICIES
BASELINE CONVENTIONAL
REDUCED TILL
NOTILL
Weighted sum
ATRAZINE BAN CONVENTIONAL
REDUCED TILL
NO TILL
Weighted sum
TRIAZINEBAN CONVENTIONAL
REDUCED TILL
NOTILL
Weighted sum
GROUND WATER
Acute Chronic
1.5/0.9 0.09/0.08
2.8/2.3 0.2/0.2
12/0.9 0.09/0.08
1.8/1.3 0.12/0.11
2.8/2.7 0.04/0.04
0.05/0.03 0.0006/0.0005
0.14/0.14 0.002/0.002
2.0/1.9 0.028/0.028
024102 0.001/0.0006
0.02/0.02 0.0001/0.00009
0.001/0.001 0.00002/0.00001
0.18/0.15 0.0007/0.0004
SURFACE WATER
Acute Chronic
2.48/0.7 14.2/4.4
.1.86/0.7 13.4/9.8
0.75/0.41 4.7/3.5
257/0.69 13.7/5.8
3.98/2.7 15.8/8.4
1.34/0.8 7.3/4
0.95/0.54 3.1/1.7
. 3.19/2.13 135/7.0
1.02/0.37 6.5/4.7
054/0.1 0.88/0.6
0.009/0.0036 0.07/0.04
0.17/0.07 0.6/0.4
ECO
189/30
131/58
33/21
169/37
203/53
114/40
33/17
175/49
344/123
57/23
5.9/3.3
258/93
EROSION
W/SU
••
..
1.000
••
- 1.038
1.014
w/o
1.000
••
1.009
••
...
1.018
. ,
Note: The first number In a box represents the sum of the pesticide exposure values in a medium. The second is the highest ;
•exposure value for any pesticide predicted in a medium.
Table 15.
Estimated percentages of acreage under different tillage practices under the 3 policy scenarios
Baseline Atrazine Ban Triazlne Ban
Conventional Tillage 0.6992 0.7031 0.7054
Reduced Tillage 0.2746 0.2708 0.2684
No Till 0.0262 0.0261 0.0262
corn acreage would be expected to
decrease, leading to more conven-
tional tillage and thus more soil ero-
sion. The Conservation Compliance
provisions of the Farm Bill, how-
ever, restrict how much land, par-
ticular those lands that are highly
erodible, that can be taken out of
conservation tillage. The loss of the
triazine herbicides, which have been
heavily relied on in shifting produc-
tion from conventional to conserva-
tion tillage, would partly force the
adoption of new technology—in the
form of thesulfonylureaherbicides—
to meet the demands of Conservation
Compliance.Thesulfonylureas,which
are very effective in post-emergent
use, allow the farmer to reduce the
amount of tillage that would other-
wise be necessary to control weeds.
Weighting and aggregating
impacts:
In deciding whether or not the risks
of the use of atrazine on corn out-
weigh its social benefits, EPA must
not only determine the magnitude
of these risks and benefits, but also
whether or not curtailment of its use
leads to a net environmental gain or
loss under different economic and
policy assumptions.If theuse of atra-
zine on corn is banned in all areas of
the country where ground water is
susceptible to contamination, for
example, will the switching of farm-
ers to alachlor or the other alterna-
tives lead to unacceptable worker
health risks or levels of contamina-
tion of surface water? Are fish popu-
lations potentially placed at greater
risk and do the benefits of use still
outweigh this risk? How should
ground and surface water problems
associated with atrazine be traded
off against dietary risks possibly as-
sociated with the greater reliance on
the use of a herbicide that is -more
likelytoleaveresiduesoncorn?What.
26
-------
Results of Anolysis
do we have to believe about how
toxic dicamba is, in light of the iden-
tified data gaps, such that we would
decide to discourage switching to it
in the event that atrazine is banned?
How can we compare risks from this
functional group of pesticides when
used in different agricultural prac-
tices? The atrazine problem can be
structured to answer any or all of
these questions. This will be demon-
strated in the following specifica-
tion exercise.
Identification of the goal is the
first step in the process of specifying
the model. The goal of the analysis is
here defined as the best balance be-
tween lowest total environmental risk
and loss in economic benefit, both to
producers and society as a whole, in
combatting an economic pest prob-
lem. We structure the analysis to
identify the pesticides (and/or prac-
tices) that contribute the most (or
conversely) theleasttoenvironmen-
tal risk and economic benefit.
Specification of the risk compo-
nent is the next step. The decision-
making structure in OPP lends itself
best to separate analyses of human
health risk and adverse ecological
impact. The first branching point or
node in the hierarchical tree occurs
with the division of the total uni-
verse of risk into the categories of (1)
human health risk and (2) ecological
risk (Figure 9). This separation al-
lows initial comparison of pesticides
on the basis of only one category of
environmental impact. Thus, at an
early stage of the analysis, we do not
have to make comparisons between
pesticides that may pose human toxi-
cological risks and those that kill
fish, for example. Only after all the
information has been entered and
the problem completely structured
do we confront the larger question
of the trade-off between the identi-
fied human and identified ecologi-
cal risks. A pesticide that poses both
human health and ecological risks
Figure 9.
Hierarchal tree for atrazine policy impacts
Lowest
Risk
Health
Ecological
Drinking water
Dietary
I
Aquatic
I .
Sediment
1
GW
Peak
Ave. Peak
SW
etc.
etc.
Ave.
etc.
base AB TB
1.8 2.0 0.18 (summed)
1.3 1.9 0.15 (highest single value)
will be ranked according to how it
compares with its alternatives
against both criteria. Furthermore,
the weighting of ecological versus
human risks does not have to be <
explicitly made, but rather made
through a series of what-if questions
that identify the weighting that
would result in a reordering of poli-..
cies or practices. !
The next branches on the hierar- ;
chical tree in Figure 9 are the envi-.
ronmental impacts. The impacts for;
human health risk for this analysis
are dietary and drinking water risks.
For ecological risk, the impacts were
chosen to be risk to aquatic ecosys-
tems and sediment loadings to sur-
face water—the latter being a proxy
for the ecological problems associ-
ated with sediment loadings.
Clearly, worker risk, impacts of long-
range transport, impacts on terrestial
specieis, and bioaccumulation in fish
could also have been included. They
were not done so for the purpose of.
simplifying this analysis. Each hu-
man health impact is further differ-
entiated by type of exposure: peak
level, resulting in short-term or acute
exposures, and average levels lead-.
ing to long-term or chronic expo-
sures.
. The weighted values obtained
from Table 14 for. the policy alternar
tives that are to be compared serve
as inputs to the leaves of the tree.
Thus, 1.8, summed exposure values,
and 1,3, highest value for any pesti-
.cide in the medium, are the input for
baseline acute exposures to ground-
water sources of drinking water in
the category of human health, 2.0/
1.9 for an atrazine ban, and 0:18/.
0.15 for a triazine ban. We create an
initial weightingby settingeach node
of the tree to 0.5. The one exception
is the setting of 0.1 to peak exposures
for ground water and 0.9 to the cor-
responding average exposures be-
cause of the apparently low Agency
concern regarding short-term expo-.
27
-------
Agricultural Atrazine Use and Water Quality
A CEEPES Analysis of Policy Options
sures to pesticides in ground water
versus long-term exposures (Figure
10).
With these summed exposure val-
ues and the weights for impacts, we
obtain an initial index score for the
policy^ttematives asshowninFigure
11. Baseline receives the score 0.387,
an atrazine ban 0.414, and a triazine
ban0.199.Wherewedonotsum expo-
sure values and use highest value ob-
tained for anypesticideinthemedium,
the scores are baseline 0.41, atrazine
ban 037, and triazine ban 0.2.
The key weight affecting" the or-
dering of the alternatives is that for
health impacts (Figure 12). Changing
this weight to less than 02, where
exposure values are summed (024,
where they are not summed), results
in a reordering of priorities. Thus, the
decision-maker must decide whether
or not ecological impacts associated
with corn herbicide use'greatly out-
weigh (by a factor of 3 to 5) those
associated with human health in de-
ciding whether or not to ban the atra-
zineor the triazineherbicides together.
In fact, the more important ecological
impacts are deemed to be, the less
preferable becomes a triazine ban.
A similar sensitivity analysis
could entail reweighting the relative
importance of drinking water and
dietary risks, as shown in Figures'
13a and 13b. One must believe that
the magnitude of risks associated
with dietary risks must greatly ex-r
ceed those of drinking water (0.62 to
0.38) to get a reordering of alterna-
tives with respect to the health crite-
rion. A similar sensitivity analysis
could be conducted for the weights
for each of the nodes of the tree.
In the next step, the aggregate
hazard index is compared to the es-
timates of economic benefit. Eco-
nomic benefit is drawn from the
information in Table 7, whereby we
use only the short-term measures.
An atrazine ban results in the eco-
nomic .loss of $365 million and a
triazine ban $526 million.
A plot of economic benefit against
our estimate of risk (the relative haz-
'ard index), as shown in Figures 14a
and b below, suggests that, if the sta-
tus quo, as shown by baseline, is not
desired, then an triazine ban is supe-
rior to a simple ban on atrazine. An
atrazine ban buys little reduction in
the hazard index, yet a roughly 50%
increase in cost for a triazine ban leads
to a halving of the risk. However, as
indicated above, the^reater the deci-
sion-maker decides to weight ecologi-
cal impacts relativetoimpacts affecting
human health, the less preferable a
triazine ban is since its ecological im-
pact is greater. Thus, the rationale for
banning either atrazine alone or the
triazines as a group diminishes the
more important the ecological impact
of herbicide use is relative to human
health.
Weighting of hierarchical tree nodes
Lowet Risk
HealthfO.51
EcoloaicalfO.5^
1
prinking water(O.S)
I I
DietaryfO.5) AquaticfO.5)
lii
GWfO.5)
swfoja
1
Peak(0.1) Ave.
SedimenUO.S)
.ESS
isjaxi
i g §
o.S.5
"11
T
0.9) Peak(0.5)
-
1
I z£ Sis SiE
28
-------
Results of Analysis
Figure 11.
Aggregate Hazard Index Store for health and ecological Impacts weighted equally where
exposure values are summed
h
Criteria
Figure 1 2.
Relative importance of health and ecological weights in rank order of alternatives
Sensitivity with respect to goal nodes below:
Goal
.1 .2 .3 .4 .5 .6 .7 .8 .9
Priority of HEALTH
Atrazine Ban
Baseline
Triazine Ban
29
-------
Agricultural Atrazlne Use and Water Quality
A CEEPES Analysis of Policy Options
Figure 13a.
Ordering of alternatives with drinking water and
dietary risks weighted equally (exposure values summed)
Criteria
|TriazjneBan(.017)
Figure 13b.
^^^^^^•i^^^^^^^HIH^^M^^^B^^^^"^"^^^H*^^^~'^~"1 •
Relative importance of drinking water and dietary
risks at point of Indifference between accepting of
status quo (baseline) and atrazine ban (exposure
values summed)
Criteria
iTriazine Ban (.017)
Figure 14a.
Plot of relative hazard index (using summed exposure
values) against loss In economic benefit
600 -
500 -
400
300
200
100
0
TrUzine ban
Alnzine ban
J_
_L
Baseine
0 0.1 0.2 0.3 0.4 0.5
Relative hazard index
Plot of relative hazard index (using a maximum
exposure value per impact) versus economic benefit
1
600 -
500 -
400
300
200
100
Triazine ban
Atrazine ban
J_
0.1 0.2 0,3 0.4
Relative hazard index
0.5
30
-------
Conclusion
Our CEEPES analysis leads to a num-
ber of key conclusions. The overall
goal of reducing environmental risk
from agricultural non-point source
pollution would not be met from an
atrazine ban without a concomitant
effort to move producers towards
other, more environmentally pro-
tective practices. A ban on the triaz-
ine herbicides would be preferable
to a ban on atrazine should the im-
portance of reducing human health
risks be considered more important
than protecting the ecological integ-
rity of hydrologic systems.
With an atrazine ban we esti-
mate a decrease in both producer
income and consumer effects of $937
million. Government expenditures
for commodity price supports, how-
ever,, would decrease significantly.
Thus, the producer and consumer
losses offset in part by the reduction
in government transfer payments
leads to an overall short-term eco-
nomic loss of $365 million for the
entire nation.
Overall herbicide use would not
decrease with an atrazine ban. In-
deed, total triazine use would in-
crease by 27% on corn, though it
would decrease by 84% on sorghum.
Nontriazine herbicide use would
also increase as more acreage is
treated with nontriazine herbicides
at higher application rates. Use of
the new low-dosage sulfonylurea
herbicides would greatly increase,
at least in the short run. Given uncer-
tainty regarding pest resistance to
these new herbicides and phy totoxic
effects on non-target plants, an alter-
nate scenario wherein these herbi-
cides are not available suggests a
major shift to post-emergent use of
other, non-triazine herbicides that
could pose potential problems to
surface waters.
We also observe minor shifts in
tillage for all crops due to the con-
straint posed by the Conservation
Compliance. We did, however, pre-
dict a significant shift away from
continuous corn to corn-soybean
rotation.
Theenvironmentwould not nec-
essarily be better of f with an atrazine
ban. Though exposures, in general,
to herbicides in groundwater would
generally decline and be of lower
general health concern than under
current use patterns in any region
regardless of tillage practice, there
could be an actual increase in site-
specific short-term exposures to
other triazine herbicides that pose
greater ground water contamination
risks. The increased use of alterna-
tive triazine and nontriazine herbi-
cides from substitute weed control
practices would lead to increases in
their concentrations in surface wa-
ter, with differential impacts by soil
and tillage type, sometimes exceed-
ing the risk associated with atrazine
before a ban. Because the atrazine
ban would result in decreased pro-
ducer and consumer surplus and
declines in surface water quality,
there could be an overall decfeasein
welfare. Thus, a ban on atrazine
alone, without a significant shift
away from chemically intensive ag-
ricultural practices, couldexacerbate
the current public health concern
regarding surface water quality.
With a triazine ban, the overall
economic loss would be $526 mil-
lion, resulting from"further rcduc- .
tions in producer income and
consumer effects, offset by reduc-
tions in government expenditures.
But under a triazine ban, human
health risks from'herbicide residues
in groundwater would decline sig^-
nificantly. Health risk from herbi-
cide exposures .through surface
waters would decline overall, with
exceptions for specific areas on a
smaller proportion of soils from
dicamba, bentazon, alachlor, and
metolachlor and only underconven-
tional and conservation tillage!
There would be a major shift to
the more widespread use of the
sulfonylurea herbicides,Tesultingin
concentrations well below their hu-
man health benchmarks. However;
there sire major uncertainties associ-
ated with these chemicalsTegarding
pest resistance and their hazard to
aquatic and nontarget terrestrial veg-
etation even at the extremely low
concentrations expected in surface
water from runoff or drift. The more
important one deems the ecological
impacts associated with herbicide
31
-------
Agricultural Atrazine Use and Water Quality
A CEEPES Analysis of Policy Options
use on corn and sorghum, the less
preferable is a ban on triazine herbi-
cides, or atrazine in particular, in
light of the possibly'greater ecologi-
cal risks associated with chemical
alternatives to atrazine use.
An'irnportantassumption under-
lying this analysis which should be
reconsidered is that farmers will not
opt out of Federal commodity pro-
grams.Theconsequencefor produc-
ers of leaving the commodity
programs is that they are no longer
bound to adhere to requirements re-.
garding conservation" tillage prac-
tices on highly erodible land. The
loss of atrazine and/or the other tri-
azineherbiddes could lead to greater
reversion to conventional tillage
practices. More widespread use of
conventional tillage for the control
of weeds would, in turn, lead to
greater soil erosion and hence water
quality problemsresultingfrom sedi-
mentation. This scenario could be
examined by CEEPES by relaxing
the constraint that the percentage of
producers subject'to Conservation
Compliance remains constant. Nev-
ertheless,thequestionablenessofthe
assumption does not detract from
the overall conclusion drawn from
the analysis that bans on atrazine
alone or the triazines as a group
would not likely result in significant
water quality benefits.
We are addressing several issues
associated with the comprehensive-
ness of the analysis. First, work is
under way with the weed scientists
of the study region to determine and
incorporate yield impacts of crop
injury. In addition, an effort is under
way toreviewtheweedcontrol strat-
egies developed for corn and sor-
ghum and evaluate their efficacy
based on local conditions in the vari-
ous states. Second, we are revisiting
the potential adoption of nicosul-
furon and primisulfuron. Given that
both chemicals are relatively new,
more analysis is warranted with re-
gard to the likelihood of their adop-
tion. Third, there are several other
policy options such as restrictions
on use, technology, and timing of
application that will be explored in
the near future.
One can infer from the results
that the conventional paradigm in
addressing pesticide-related prob-
lems, i.e., bans or licensing restric-
tions, may not be the/appropriate
response for the corn and sorghum
herbicides, in particular, or water-
related problems of pesticides in
general. More geographically tar-
geted mitigation measures short of
bans may lead to Superior environ-
mental protection and minimal inv-
pact on producers and society on a
whole. The nature of these mitiga-
tion measures will be explored in
future analyses with CEEPES.
"Some measures that we are ex-
ploring include a mix of incentives
and controls to encourage the more
efficient use of chemicals and the
adoption of agricultural practices
that are more protective of water
quality—that is, to manage resources
better. The controls that are being
considered are the following: atra-
zine label restriction to 1 Ib a.i. per
acreorless;nppreemergentuse(atra-
zine or herbicides in general); re-
quirement of vegetative filter strips
along surface water bodies; impact
of crop rotations on yield and envi-
ronmental risk. These and other poli-
cies and practices will be examined
in future studies usingCEEPES, with
results presented by watershed.
32
-------
References
Belluck, S. Benjamin, and Tawsori (1991). Groundwater Contamination by Atraziine and Its Metabolites: Risk
Assessment, Policy, an Legal Implications. Pesticide Transformation Products, Fate and Significance in the Environ-
ment. (L. Somasundaram and J. Coats, eds.). Washington,D.C: American Chemical Society. :
Bouzaher, A., D. Archer, R. Cabe, A. Carriquiry, and J. Shogren. (1992a). "Effects of Environmental Policy on Trade-
offs in Weed Control Management." CARD Working Paper 91-WP81, November. (Forthcoming in Journal of
Environmental Management.) • "
Bouzaher A., R. Cabe, D. Archer, and J. Shogren (1992b). "Characterization and Choice of Weed Control
Technology Under Weather Uncertainty." Unpublished manuscript. Center for Agricultural and Rural Devel-.
opment, Iowa State University.
NAPIAP (1992). "The Effects of Restricting or Banning Atrazine Use to Reduce Surface Water Contamination in •
the Upper Mississippi River Basin." National Agricultural Pesticide Impact Assessment Program. June.
Penson, John B. and C. Robert Taylor (1992). "United States Agriculture and the General Economy: Modeling the
Interface." Agricultural System, Vol. 39, pp. 33-66. •
Gianessi, Leonard P. and C. A. Patton (1992). "Reregistration of Minor Use Pesticides: Some Observations and
Implications." Resources for the Future.
Gianessi, Leonard P. and C. A. Puffer (1991). "Herbicide Use in the United States: National Summary Report." •
Resources for the Future.
Johananson, R.C., Imhoff, J.C., Kittle, J.L., and Donigian, AS. (1989). Hydrological Simulation Program—FORTRAN,
EPA 600/3-89-066. . - . .
Taylor, C. Robert (1987). "AGSIM User's Manual." Illinois Agricultural Economic Staff Paper 87E-394. Department of
Agricultural Economics, University of Illinois at Urbana-Champaign.
Taylor, C. Robert (1991). "Deterministic vs Stochastic Evaluation of the Aggregate Effects of Price Support
Programs: A Large-Scale Simulation Comparison." Staff Paper ES91-2, Aprii. Department of Agricultural
Economics and Rural Sociology, Auburn University.
Thurman, E. M., D.A. Goolsby, M.T. Meyer, and D.W. Kolpin, "Herbicides in Surface Waters of the Midwestern
United States: the Effect of Spring Flush," Environmental Science & Technology, pp 1794-1796 (1991-).
US. Department of Agriculture (199 i). Agricultural Chemical Usage: 1990 Field Crops Summary. Economic Research
Service, Ag. Ch. 1(91). March. . . ..
US. Environmental Protection Agency (1992a). "Voluntary Label Changes for Atrazine.? EPA News-Notes #21-'
(May), 6-8. U.S. Environmental Protection Agency (1992b). Framework for Ecological Risk Assessment. EPA/630/ "
R-92/001. February. r
.'.'"' '! • - • " '•- " .33'-
-------
Agricultural Atrazine Use and Water Quality
A CEEPES Analysis of Policy Options
34
-------
Appendix A
Herbicides Included in CEEPES Configuration for
Atrazine and Water Quality
-------
Agricultural Atrozlne Use and Water Quality
A CEEPES Analysis of Policy Options
36
-------
Appendix
A
II Herbicides included in CEEPES configuration for atrazine and water quality |
Code
LAS
ATR
BAS
BUG
SUT
BLA
BAN
ERA
ROU
DUA
ACC
GRA
BEA
PRO
RAM
PRI
TFD
Chemical
Alachlor
Atrazine
Bentazon
Bromoxynil
Butylate
Cyanazine
Dicamba
EPIC
Glyphosate
Metolachlor
Nicosulfuron
Paraquat
Primisulfuron
Pendimenthalin
Propachldr
Simazine
2,4-D
^^H^^^^H^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^H
Trade Name
Lasso
AAtrex
Basagran '.
Buctril
Sutan
|Bladex
Banvel
Eradicane
1 'i
Round-up
Dual .
Accent
Gramaxone
Beacon .
. Prowl
Ramrod
Princep
2,4-D
-------
Agricultural Atrazine Use and Water Quality
A CEEPES Analysis of Policy Options
38
-------
Appendix B
Shifts in Weed Control Strategies
-------
Agricultural Atrazlne Use and Water Quality
A CEEPES Analysis of Policy Options
40
-------
Appendix
B
•^^^^M^^^^^^^^^^^^^^^^MHH^^^^^^^^^^^^^^^BMMMH^^^^H^^B^^
Percentage of corn acres treated, baseline
Strategy
Number
66
156
157
158
170
69
70
219
107
67
Primary Strategy
Atrazineb-Bladex preemergence
Atrazine'-Lasso preplant inc.
AtrazinevDual preplant inc.
Atrazineb-Bladex preplant inc.
Atrazine2-Bladex preplant inc.
Atrazine'-Lasso preplant inc.
AtrazineMBIadex preemergence
Atrazinefc-Bladex preemergence
Sutan preplant inc. & Banvel post
Oual-Banvel preemergence
Atrazineb-Bladex preemergence
•Atrazine applied at a rate > 1.5 Ib/acre.
bAtrazine applied at a rate < 1.5 Ib/acre.
Secondary Strategy
Atrazineb-Basagran postemargence
Atrazineb-Banvel preemergence
Atrazineb-Buctril preemergence
Accent preemergence
Beacon preemergence
• Atrazine* preemergence
Atrazineb-Basagran preemergence
Atrazineb-Banvel preemergence .
Atrazine'-Buctril preemergence
Atrazineb-Basagran preemergence
Atrazineb-Banvel preemergence
Atrazkie'-Buctril preemergence
2,4-D preemergence
Banvel-2,4-D preemergence^
Accent preemergence ;
Atrazineb-Basagran preemergence
Atrazineb-Banvel preemergence
Atrazineb-Buctril preemergence
Bladex preemergence '
Prowl-Bladex preemergence
Percent
Treated
16.0
11.7
9.5
7.6
-5.3
3.9
4.4
3,8
3J6
3.4
41.
-------
Agricultural Atrazine Use and Wafer Quality
A CEEPES Andys'* of Policy Options
Figure B.I.
Percentage of corn acres treated^ baseline
Percentage of acres treated
67 107 219 70 69 170 158 157 156 6.6 all others
Weed management strategy
42
-------
B
Percentage of com acres treated, atrazine ban
Strategy Number Primary Strategy
185 Princep preplant incorporated
127
225
99
223
219
183
191
Bladex-Prowl preemergence
Sutan preplant inc. & 2,4-D post Lasso
preplant inc. & 2,4-D post Dual
preplant inc. & 2,4-D post
Bladex-Lasso preemergence
Bladex-Dual preemergence
Sutan preplant inc. & 2,4-D post
Lasso preplant inc. & 2,4-D post
Dual preplant inc. & Banvel post
Princep preplant incorporated
Princep preplant incorporated
Bladex-Lasso preemergence
Bladex-Dual preemergence
Secondary Strategy
Accent postemergence
Beacon postemergence
Accent postemergence
Beacon postemergence
Accent postemergence
Beacon postemeirgence
Accent postemergence
Beacon postemergence
Accent postemergence
Beacon postemergence
Banvel postemergence
Buctril postemergence
Basagran postemergence
Banvel postemergence
Buctril postemergence
Basagran postemergence
Accent postemergence
Beacon postemergence
Percent Treated
.22.8
15.6
11.9
11.3
.11.1
3.8
3.8 -
3A
43
-------
Agricultural Atrazine Use and Water Quality
A CEEPES Analysis of Policy Options
Figure B.2.
Percentage of corn acres treated, atrazine ban
Percentage of acres treated
191 183 219 223 99 225 127 185 all others
Weed management strategy
44
-------
Appendix
B
Table B.3.
Percentage of com acres treated, triazine ban
Strategy
Number
225
Primary Strategy
Sutan preplant & 2,4-D post
Lasso preplant inc. & 2,4-D post
Dual preplant inc. & 2,4-D
144 Rotary hoe and row cultivation
223 Sutan preplant inc. & Banvel post
Lasso preplant inc. & Banvel post
Dual preplant inc. & Banvel post
111 Dual-Banvel preemergence
Secondary Strategy Percent
Treated
Accent postemergence 43.9
Beacon postemergence
Accent postemergence 27.0
Beacon postemergence
Banvel postemergence
Buctril postemergence 9.2
Basagran postemergence
Accent postemergence
Beacon postemergence
4.3
219 Sutan preplant inc. & Banvel post
Lasso preplant inc. & Banvel post Accent postemergence 4.3
Dual preplant inc. & Banvel post • Beacon postemergence
133 Lasso preemergence & Banvel post Accent postemergence 3.9
Dual preemergence & Banvel post Beacon postemergence
52 Accent postemergence
Beacon postemergence
None
3.1
45
-------
Agricultural Atrazine Use and Water Quality
A CEEPES Analysis of Policy Options
Percentage of com acres.treated, triazine ban
Percentage of acres treated
521 133 219 111 223 144 225 all others
Weed management strategy
46
-------
Appendix
B
Table B.4.
Percentage of sorghum acres treated, baseline
Strategy Number Primary Strategy
1061•'.' Lasso preplan! inc. & Banvel post.
Dual preplan! inc. & Banvel post.
1059 Lasso preplan! inc. & Banvel post.
Dual preplan! inc. & Banvel post.
1050 Atrazine* preplan! incorporated
1042 Prowl-Alrazineb postemergence
1054 A!razineb-Dual preplan! inc.
1090 Atrazine* postemergence
•Atrazine applied at a rate > 1.5 Ib/acre.
bAtrazine applied at a rate < 1.5 Ib/acre.
Secondary Strategy
2,4-D postemergence
Banvel 2,4-D postemergence
Prowl-Atrazine* postemergence
2,4-D postemergence
None
Prowl-Atrazine''postemergence
None
Percent Treated
35.3
22^0
12.2
10.3
10:2 "...
5.9
Figure B.4.
Percentage of sorghum acres treated, baseline
Percentage of acres treated
40
30
20
1090 1054 1042 1050 1059 10611 all others
Weed management strategy
47
-------
Agricultural Atroiine Use and Water Quality
A CEEPES Analysis of Policy Option*
Table B.5.
Percentage of sorghum acres treated, atrazine ban
Strategy Number Primary Strategy
'• '1061 Lasso preplan! Inc. & Banvel post.
Dual preplan! inc. & Banvel post.
1066 Lasso preplant Inc. & 2,4-D post.
Dual preplant inc. & 2,4-D post.
1025 Bladex-Ramrod preemergence
1039 Lasso preemergence & Banvel post,
Dual preemergence & Banvel post.
Ramrod preemergence & Banvel post.
Secondary Strategy Percent Treated
2,4-D postemergence 60.7
Banvel, 2,4-D postemergence .
2,4-D post-emergence 15.9.
Banvel 2,44-D '.
2,4-D post emergence . 15.5
Banvel, 2,4-D postemergence •• .
2,4-D post emergence 7.8
Banvel 2,4-D postemergence
Percentage of sorghum acres treated, atrazine ban
Percentage of acres treated
70
60
1039 1025 1066 1061
Weed management strategy •
48
-------
B
Table B.6.
Percentage of sorghum acres treated, triazine ban
Strategy Number Primary Strategy
1061 Lasso preplan! inc. & Banvel post.
Dual preplan! inc. & Banvel post.
1066 Lasso preplan! inc. & 2,4-D post.
Dual preplant inc. & 2,4-D post.
Lasso preemergence & Banvel post.
. 1039 Dual preemergence & Banvel post.
Ramrod preemergence & Banvel post.
1043 Rotary hoe and row cultivation
Secondary Strategy ' Percent Treated
2,4-D postemergonce BO-2
Banvel 2,4-D postemargence .
2,4-D postemergonce 15.7 •
Banvel 2,4-D postemargence •
2,4-D postemergonce
Banvel-2,4-D postemergence
None '
B.B
Figure B.6.
Percentage of sorghum acres treated, triazine ban
•Percentage of acres treated
70
60
1066
Weed management strategy
1061
-------
Agricultural Atroiine Use and Water Quality
A CEEPES Analysis of Policy Options
50
-------
Appendix C
Shifts in Rotations
-------
Agricultural Atrazme Use and Water Quality
A CEEPES Analysis of Policy Options
52
-------
Appendix
c
• Change in the distribution of crop rotation; ! I I I
I^BB^H
, .V. «
' BAR BAR SOY
BAR BAR SMF
BAR CRN CRN CRN
BARCSL
BAR HLH HLH CRN CRN
BAR SOY
BARSUNSMFBWT
BARSWT
BARSWTSUNSWT
CRN
CRN CRN CRN OTS NLH NLH
CRN CRN CRN SWT
CRN CRN CRN WWT HLH
CRN CRN CRN HLH HLH HLH
CRN CRN OTS HLH HLH
~ CRN CRN OTS HLH HLH HLH
CRN CRN OTS NLH NLH
CRN CRN OTS NLH NLH NLH
CRN CRN SOY
CRN CRN SOY DTS HLH
CRN CRN SOY SRG
CRN CRN SOY WWT
CRN CRN SOY WWTHLH
CRN CRN WWT HLH HLH
CRN CRN WWT HLH HLH HLH
CRN OTS HLH HLH HLH
CRN OTS NLH NLH NLH
CRN OTS WWT
CRNSRG
CRN SOY
CRN SOY CRN WWT HLH HLH
CRN SOY OTS NLH NLH
CRN SOY WWT
CRN SOY WWTHLH HLH HLH
CRNSWT
CRNSWTSWT
CRN WWT
CRN NLH NLH
CRN NLH NLH NLH
m
2
4
8
11
21
70
78
„ 80
81
100
107
109
111
115
122
123
125
126
131
132
136
137
138
144
145
162
168
170
178
. 186
189
196
201
203
210
215
218
231
232
Baseline
mil. ac
4.1903
0.2606
0.3952
0.0639
4.4649
0.0585
1.1884
0.1244
2.6112
18.0974
0,2161
0.0027
0.1552
0.7133
0.2444
2.3169
1.1646
0.2579
15.3956
2.6669
0.5747
0.5524.
2.0080 -
1.7473
0.6041
0.8612
0.1723
0.1019
0.1165
51.6816
- 1.3258
0.5885
12.9918
3,4480
0.2238
2.1908
07467
0.2277
0.4235
ATRZ.BAN
% chng
0.000
0.000
0.000
0.000
0.000
1.144
0.000
0.000
0.000
-16.010
-59.545
-6.520
111.915
:
0.006
-79.920
-1.025
108.965
-14.495
0.227
-24.878 ,
o.opo"
-8.512
0.351
0.000
149.892
0.000
169.026
7.955
11.116 j
0.000 ; '
-3.163 ;
-2.874
21.207
39.355 i
27.264
-15.374 ;
0.000
0.000
ij
TRIZ_BAN
% chng
,
0.000
0.000
0.024
0.000
0.000
. 1:144
0.000
0.000
0.000
-16.552
0.000
. ' •- - '•'.'••'
199.823
'
0.025 .
-25.077
0.212
.
-12.901
0,201
-24.878
0.000
-15.344
0.976
0.000
116.890
0.049
0.000
" 7.955
-10.409
0.000
42.863 '
-2.800
20.580 - -'
38.925
33.223
0.000
o.obo
0.000, " •
53
-------
Agricultural Atrazine Use and Water Quality
A CEEPES Anolysis of Policy Options
CSL , ' -
CSLCSLCSLSWT
CSL CSL OTS HLH
•'i: - CSL CSL OTS HLH HLH
• CSL CSL OTS HLH HLH HLH
CSLCSLOTSNLHNLH
CSL CSL SOY
CSLCSLWWTNLHNLH
CSL OTS
CSL OTS HLH HLH
CSL OTS HLH HLH HLH '
CSL OTS HLH HLH SWT
CSL SOY CSL OTS NLH NLH
CSL SOY HLH HLH HLH HLH
CSL SWT
COTSOV
HLH HLH HLH HLH SRG SOY
OTS HLH HLH HLH
OTS HLH HLH HLH SRG SRG
OTS NLH NLH NLH
OTS NLH NLH NLH SOY
OTS NLH NLH SSL SOY SSL
OTSOTSSMF
SRG
'SRG SOY
SRG SOY SOY
SRGSOYWWT
SRG SOY HLH HLH HLH HLH
, SRGWWT
SRG NLH NLH
SSL SSL WWT NLH NLH
SSL WWT
SOY SWT SWT
SOY WWT
SOY WWT SOY
SOY WWT WWT WWT
SOY NLH NLH NLH NLH
SMFSWT
SMFWWT
SWT HLH HLH HLH
WWT
HLH HLH HLH HLH
NLH NLH NLH NLH
235
239
242
243
244
246
250
252
259
261
262
265
276
277
280
325
339
350
354
366
372 .
379
384
409
415
416
417
419
424
432
434
439
452
454
458
459
462
463
471
478
490
503
508
0.5276
0.4141
0.9167
1.6314
1.3533
0.2556
0.4473
0.0053
0.5418
0.2843
- 1.6779
0.4283
0.0874
0.4449
0.4112
2.4551
0.9416
0.8094
0.4518
2.2941
0.1380
0.0204
0.7267
1.2112
1.4834 '
1.9871
0.2250
0.2029
3.3043
0.0240
0.0031
0.0997
1.0513
0.2606
. 0.8846
1.6778
0.0431
5.8720
0.6293
0.0519
2.8487 .
18.2503
18.0575
2.427
0.042
-0.324
-0.154
12.500
6.656
-41.129
3.367
-0.418
-12.559
-3.060
-
-23.647
0.000
-0.111
0.000
15.14
48.51
9.82
-10.40
0.000
5.11
6.83
0.03
-15.79
-9.06 '
1.03
-0.09
7.50
1017.94
2.47
0.000
0.000
-5.44
10.63
0.000
-61,45
-3.05
-
0.000
-12.63
0.26
-6.29
4.812 .'•-_••'
-98.685
-0.324.
-0.108
15.486
5.940
-60.806 • \ '
3.367
-0.373
-11.765
5.601
-
-21.877
0.000
0.123
0.000
0.000
-22.48
9.14
-16.43
0.000
6.55
5.69
-0.01
-12.36
-9.06
1.03
-0.52
11.08
1017.94
' 2.47 •
0.000
-•• o.ooo
-5.40
10.26
0.000
-61.45
-4.20 .
• ' " '.'.'.
781.34
-5.54
-0.09
-3.95'
54
-------
Appendix D
• H
Ground and Surface Water Concentrations and Exposure
Values for 15 Herbicides in the Study Region for Baseline,
Atrazine Ban, and Triazine Ban Scenarios in Corn Production
-------
Agricultural Atrozine Use and Water Quality
A CEEPES Analysis of Policy Options
56
-------
Appendix
D
\
• Ground and Surface Water Herbicide Concentrations (in parts per billion)
POLICY=Baseline CROP=Corn
Conventional Tillage
CHEMICAL*
Atrazine
Atrazine<1.5
Nicosulf uron
•'': Dicamba
• ;Primisutfuron
Cyanazine
Bromoxynil
Bentazon
Melolachbr
Alachlor
Simazine
Pendimethajin .
Propachlor
Butylate
2,4-D
Atrz-Med-Oecay
Atrz-Slow-Decay
Atrz-Fast-Decay
AVG12
0.24904
0.23983
0.00000
0.00064
0.00138
0.00017
0.00000
0.00681
0.00019
0.00009
0.23551
0.00000
0.00001 '
0.00983
0.00003
0.47923
2.83069
0.00009
PK12
2.3674
2.7014
0.0000
0.1246
0.0130 .
0.0325
0.0000
0.1796
0.0133
0.0147
1.8963
- 0.0000
0.0011
2.7832
0.0145
4.3875
10.8769
0.0085
AVG15
0.001296
0.001804
0.000000
0.000000
0.000006
0.000000
0.000000
0.000010
0.000000
0.000000
0.002580
0.000000
0.000000
0.000000
0.000000
0.002944
0.056920
0.000000
PK15
0.009379
0.017331
0.000000
0.000002
0.000026
0.000008
0.000000
0.000147
0.000009
0.000001
0.019724
O.OOOOOO
0.000000
0.000004
0.000000
0.022901
0.099106
0.000000
PKSTRM
34.9458
17.8924
0.0175
29.4817
0.1757
.63.7835
0.0309
2.9032 . ,
17.1911-
27.7008- ••'..'•
34.7559 -
.0.0000
6.3120
28.1301
15.7291
59.0422
52.0976
25.4502 .
Reduced Tillage • .
CHEMICAL
Atrazine
Atrazine<1.5
Nicosulfuron
Dicamba
Primisulfuron
Cyanazine
Bromoxynil
Bentazon
Metolachlor
Alachlor
Simazine
Pendimethalin
Propachlor
2,4-D
Atrz-Med-Decay
Atrz-Slow-Decay
Atrz-Fast-Decay
AVG12
0.60158
0.52893
0.00000
0.00064
0.00090
0.00005
0.00000
0.02229
0.00004
0.00001
0.22449
0.00000
0.00001
0.00004
1.10987
6.32963
0.00054
PK12
7.0364
5.3189
0.0000 '
0.1688
0.0094
0.0093
0.0000
0.5347
0.0031
0.0022
1.7034
0.0000
0.0037
0.0226
10.8803
26.2752
0.0515
AVG15
0.004093
0.003573
0.000000
0.000000
0.000003
0.000000
0.000000
0.000028
0.000000
o.oooooo •
0.002092
0.000000
0.000000
0.000000
0.007317
0.099055
0.000000 .
PK15
0.024113
0.02304
0.00000
0.00001
-Q.00002
0.00000
0.00000
0.00030
0.00000
0.00000
0.01350
0.00000
0.00000 .
0.00000
0.0423(5
0.14595
0.00000
PKSTRM '
65.918
• 51.643
0.006
19.254
0.062
19.215
0.072
7.704
10.160
3.891
22.949
0.000 .
11.067
6.071
133.502
117.672
23.729
No-tillage
CHEMICAL
Atrazine
Atrazine<1.5
Nicosulfuron
Dicamba
Primisulfuron
Cyanazine
Bromoxynil
Bentazon
Metolachlor
Alachlor
Simazine
Pendimethalin
Glyphosate
2,4-D
Atrz-Med-Decay
Atrz-Slow-Decay
Atrz-Fast-Decay
AVG12
0.25210
Oil 0878
0.00000
0.00001
0.00088
0.00000
0.00000
0.00047
0.00002
0.00000
0.18579
0.00000
0.00000
0.00000
0.35380
2.05537
0.00013
PK12
2.73901
0.98029
0.00000
0.00359
0.00857
0.00055
0.00000
0.00986
0.00125
0.00057
1.11248
0.00000
0.00000
0.00087
3.23373 -
7.92776
0.01297
AVG15
0.002933
•0.000925
0.000000
0.000000
0.000006
0.000000
0.000000
0.000001
0.000000
0.000000
0.003074
0.000000
0.000000
0.000000.
. 0.003656
0.062238
0.000000
PK15
0.023260
0.005138
0.000000
0.000000
0.000024
0.000000
0.000000
0.000008
0.000001
0.000000
0.012677
0.000000
0.000000 -
0.000000
0.024598
0.093926
0.000000
PKSTRM
26.8218
13.7617
0.0029
0.2557
0.0300
1.7765
0.0006
0.0634
0.8591
0.9827.
15.2363
0.0000
0.0000
0.2505
45.8989
40.4743
8.0125
*see legend, page 58
57
-------
Agricultural Atrazine Use and Water Quality
A CEEPESAnolysis of Policy Options
• Ground ond Surface Water Herbicide Concentrations (in parts per billion); i •
POLICYsAtrazine Ban CROP=Cora , - "
Cnventlonal Tillage
CHEMICAL*
r-'>Jkx>sutfuron
*• DIcamba
Prtmlsulfuron'
Cyanazine
Bromoxynll
Bentazon
Metolachtor
Alachtor
Simazine
Butylate
2,4-D
AVG12
0.00000
0.00114
0.00265
0.00017
0.00000
0.00592
0.00029
0.00015
1.37028
0.01052
0.00009
PK12
0.000
0.2313
0.0281
0.0343
0.0000
0.1301
0.0220
0.0244
10.9284
3.0583
0.0365
AVG15
0.000000
0.000000
0.000014
0.000000
0.000000
0.000008
0.000000
0.000000
0.013849
0.000000
0.000000
PK15
0.000000
0.000003
0.000061
0.000009
0.000000
0.000090
0.000033
0.000010
0.090643
0.000004
0.000000
PKSTRM
0.027
53.845
0.256
58.733
0.044 •
. 3.228 ,
13.308
21.492
134.997
28.729
42.60.3
Reduced Tillage
CHEMICAL
. Nlcosutfuron
DIcamba
Primlsutfuron
Cyanazine
Bromoxynil
Bentazon
Metolaehtor
Alachtor
Simazine
Pendlmethalin
2,4-D
CHEMICAL
NteosuHuron
DIcamba
Primlsutfuron
Cyanazine
Metolachlor
Alachtor
Simazine
Pendimethalin
Glyphosate
AVG12
0.000000
0.000418
0.002906
0.000078
0.000000
0.000016
0.000128
0.000040
0.018520
0.000000
0.000003
AVG12
0.000000
0.000001
0.001561
0.000001
0.000305
0.000058
0.084350
0.000000
0.000000
PK12
0.00000
0.10324
0.03080
0.01260
0.00000
0.00097
0.00887
0.00631
0.13033
0.00000
0.00145
PK12
0.00000
0.00022
0.01534
0.00012
0.02146
0.00875
0.53965
0.00000
0.00000
AVG15
0.00000000
0.00000000
0.00001459
0.00000000
0.00000000 .
0.00000000
0.00000003
0.00000000
0.00029247
0.00000000.
0.00000000
No-till
AVG15
0.0000000
0.0000000
0.0000131
0.0000000
0.0000006
0.0000000
0.0010061
0.0000000
0.0000000
PK15
0.0000000
0.0000012
0.0000550
0.0000006
0.0000000
0.0000000
0.0000044
0.0000004
0.0018712
0.0000000
0.0000000
PK15
0.0000000
0.0000000
0.0000478
0.0000000
0.0000456
0.0000041
0,0062077
0.0000000
0.0000000
PKSTRM
0.0235
14.2105.
0.2298
79.8721
0.0026
0.1847
23.2487
21.6838
2.3393
0.0000
1.0366
PKSTRM
0.0062
0.0260
0.0610 '
0.5952
16.6854
10.7396
26.8990
0.0000
0.0000
*see legend, page 58
58
-------
Appendix
D
• Ground and Surface Water Herbicide Concentrations (in parts per billion) 1 1
POlia=Triaiine Ban CROP=Corn
Conventional Till .
CHEMICAL*
Nicosulfuron
Dicamba
Primisutfuron
Bromoxynil
Bentazon
Metolachlor
Alachlor
Butylate
2,4-D
AVG12
0.000000
0.000849
.0.006444
0.000000
0.007219
0.000368
0.000187
0.028833
0.000347
PK12
0.00000
0.17073
0.07271 .
0.00000
0.19859
0.02846
0.03158
9.79993
0.17729
AVG15
0.000000000
0.000000007
0.000034789
0.000000000
0.000034937
0.000000535
0.000000074
6.000000637
0.000000002
PK15
0.00000000
0.00000203
0.00014400
0.00000000
0.00011283
0.00003907
0.00001069
0.00015354
O.OOOCI0092
PKSTRM .
6.072
45.053
0.655 ^
0.059
4.362
22.064
37.631
90,059
126.123
Reduced Tillage •
CHEMICAL
Nicosulfuron
Dicamba .
Primisutfuron
Bromoxynil
Bentazon
Metolachlor
Alachlor
2,4-D
AVG12
0.0000000
0.0007748
0.0060336
0.0000000
0.0000983
0.0000694
0.0000085
O.OOOD035
PK12
0.00000
0.16183
0.06930
0.00000
0.00590
0.00407
0.00099
0.00171
AVG15
0.000000000
0.000000004
0.000030323
0.000000000
0.000000003
0.000000021
0.000000000
0.000000000
PK15
0.00000000
0.00000238
0.00010681
0.00000000
0.00000014
O.OOOCI0247
0.00000009
0.00000000
PKSTRM
0.0451
.22.7040
0.4287
0.0162
1.1282
9.7194
1.6802
.1/1657 ;
No-tillage |
CHEMICAL
Nicosulfuron
Dicamba
Primisutfuron
Metolachlor
Alachlor
Gtyphosate
AVG12
0.0000000
0.0000362
0.0028049
0.0000049
0.0000026
0.0000000
PK12
0.00000 •
0.09751
0.28281
0.00403
0.00483
0.00000
AVG15
0.000000000
0.000000000
0.000022393
0.000000001
0.000000000
0.000000000
PK15
O.OOOCIQ0001
0.000000117
O.OOOQ83589
0.000000238
0.000000044
0.000000000,
PKSTRM
0.01009
1.08519
0.09718
0.18317 -
0.33859
.0.00000
*see legend, page 58
59
-------
Agricultural Atrazine Use and Water Quality
A CEEPES Analysis of Policy Options
j Ground and Surface Water Exposures (relative to benchmarks) j |
mwjm^l
CHEMICAL*
Atrazine
.. _ Atrazlne<1.5
'l-:' Nlcosulfuron
' DIcamba
Primlsulfuron
Cyanazlne
Bromoxynil
Bentazon
Metolachtor
Alachtor
Simazlne
• Pendimethalin
Propachlor
Butylate
2,4-D
Atrz-Med-Decay
Atrz-Slow-Decay
Atrz-Fast-Decay
CHEMICAL
Atrazine
. Atrazlne<1.5
• Nlcosulfuron
DIcamba
Primisulfuron
Cyanazlne
Bromoxynil
Bentazon
Metolachtor
Alachtor
Simazine
Pendimethalin
Propachtor
2,4-D
, Atrz-Med-Decay
Atrz-Stow-Decay
Atrz-Fast-Decay
CHEMICAL
Atrazine
Atrazine<1.5
Nlcosutfuron
DIcamba
Primlsulfuron
Cyanazine
Bromoxynil
Bentazon
Metolachtor
Alachtor
Simazlne
Pendimethalin
Gryphosate
2,4-D
Atrz-Med-Decay
Atrz-Slow-Decay
Atrz-Fast-Decay
^7Q|^H|
BM^H
AVG12
0.08301
0.07994
0.00000
0.00007
0.00001
0.00002
0.00000
0.00034
0.00000
0.00005
0.00673,
0.00000
0.00000
0.00020
0.00000
0.15974
0.94356
0.00003
Conventional Tillage
PK12
0.02367
0.02701
0.00000
0.00042
0.00006
0.00033
0.00000
0.00719
0.00013
0.00015
0.03793
0.00000
0.00000
0.00116
0.00001
0.04388
0.10877
0.00008
AVG15
0.00432
0.00601
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00074
0.00000
0.00000
0.00000
0.00000
0.00981
0.18973
0.00000
PK15
0.009379
0.017331
0.000000
0.000001
0.000013
0.000008
0.000000
0.000589
0.000009
0.000001
0.039448
7 0.000000
0.000000
0.000000
0.000000
0.022901
0.099106
0.000000
, -• "
PKSTRM .
0.34946
0.17892
0.00040
0.09827
0.00084
0.63784.
0.00004
0.11613
0.17191
• 0.27701
0.69512
0.00000
0.01803
0.01172
0.01430
0.59042
0.52098
0.25450
Reduced Tillage
AVG12
0.20053
0.17631
0.00000
0.00007 .
0.00000
0.00001
0.00000
0.00111
0.00000
0.00001
0.00641
0.00000
0.00000
0.00000
0.36996
2.10988
0.00018
AVG12
0.08403
0.03626
0.00000
0.00000
0.00000
0.00000
0.00000
0.00002
0.0000
0.00000
0.00531
0.00000
0.00000
0.00000
0.11793
0.68512
0.00004 '
PK12
0.07036
0.05319
0.00000
0.00056
0.00004
0.00009
0.00000
0.02139
0.00003
0.00002
0.03407
0.00000
0.00001
0.00002
0.10880
0.26275
0.00052
No-tillage
PK12
0.027390
0.009803
0.000000
0.000012
0.000041
0.000006
0.000000
0.000394
0.000013
• 0.000006
0.022250
0.000000
0.000000
0.000001
0.032337
0.079278
0.000130
AVG15
0.01364
0.01191
0.00000
0.00000
0.00000
0.00000
0.00000
0.00001
0.00000
0.00000
0.00060
0.00000
0.00000
0.00000
0.02439
0.33018
0.00000
AVG15
0.000978
0.000308
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000088
0.000000
0.000000
0.000000
0.001219
0.020746
0.000000
PK15
0.02418
0.02304
0.00000
0.00000
0.00001
0.00000
0.00000
0.00121
0.00000
0.00000
0.02699
0.00000
0.00000
0.00000
0.04236
0.14595
0.00000 •
PK15
0.023260
0.005138
0.000000
0.000000
0.00001 1
0.000000
0.000000
0.000031
0.000001
0.000000
0.025355
0.000000 .
0.000000
0.000000
0.024598
0.093926
0.000000
PKSTRM
0.65918
0.51643
0.00015
0.06418
0.00030
0.19215
0.00010
0.30815
0.10160
0.03891
0.45898
• o.ooooo
0.03162
0.00552
1.33502
1.17672
. 0.23729
.
PKSTRM
0.26822
0.13762
0.00007
0.00085
0.00014
, 0.01777
0.00000
0.00254
0.00859
0.00983
0.30473
0.00000
. 0.00000
0.00023
0.45899
0.40474
0.08013
•see legend, page 58
60
-------
Appendix
D
•"•- • - . ' •• 1
• Ground and Surface Water Exposures (relative to benchmarks) | j |
POLICY=Atrazine Ban
CROP=Corn
" ' • -
Conventional Tillage
CHEMICAL*
/'/Nicosulfuron
Dicamba
Primisulfuron
Cyanazine
Bromoxynil
Bentazon
Metolachlor
Alachlor
Simazine
Butylate
2,4-D
AVG12
0.000000
0.000127
0.000013
0.000019
0.000000
0.000296
0.000003
0.000073
0.039151
0.000210
0.000001
PK12
0.00000
0.00077
0.00013
0.00034
0.00000
0.00520
0.00022
0.00024
0.21857
0.00127
0.00003
AVG15
0.00000000
0.00000000
0.00000006
0.00000001
0.00000000
0.00000038
0.00000000
0.00000003
0.00039569
0.00000000
0.00000000
PK15
0.0000000
0.0000000
0.0000003
0.0000001
0.0000000
"0:0000036
0.0000003
0.0000001
0.0018129
0.0000000
0.0000000
PKSTRM
0.00061
•0.17948
0.00122 "
0.58733
0.00006
0.12914 . ' • - _
0.13308
0.21492
2.69993
0.01197
0.03873
Reduced Tillage
CHEMICAL
Nicosulfuron
Dicamba
Primisulfuron
Cyanazine
Bromoxynil
Bentazon
Metolachlor
Alachlor
Simazine
Pendimethalin
2,4-D
AVG12
0.00000000
0.00004650
0.00001384
0.00000868
0.00000000
0.00000080
0.00000128
0.00002006
0.00052914
0.00000000
0.00000004
PK12
0.0000000
0.0003441
0.0001466
0.0001260
0.0000000
0.0000386
0.0000887
0.0000631
0.0026066
0.0000000
0.0000013
AVG15
0.00000000
0.00000005
0.00000695
0.00000002
0.00000000
0.00000000 .
0.00000003
0.00000006
0.00083563
0.00000000
0.00000000
PK15
0.000000000
0.000000004
0.000000262
0.000000006
0.000000000
0.000000001
0.000000044
0.000000004
0.000037424
0.000000000
0.000000000 -
PKSTRM
0.00053
0.04737
0.001O9
0.79872
0.00000
0.00739
0.23249
0.21684
0.04679
0.00000
0.00094
No-tillage
CHEMICAL
Nicosulfuron
Dicamba
Primisulfuron
Cyanazine
Metolachlor
Alachlor
Simazine
Pendimethalin
Glyphosate
AVG12
0.0000000
0.0000001
0.0000074
0.0000001
0.0000031
0.0000289
0.0024100 .
0.0000000
0.0000000
PK12
0.000000
0.000001
0.000073
0.000001
0.000215
0.000088
0.010793
0.000000
0.000000
AVG15
0.000000000
0.000000000
0.000000062
0.000000000
0.000000006
0.000000014
0.000028745
0.000000000
0.000000000
PK1-8
0.00000000
0.00000000
0.00000023
o.ooopoooo
0.00000046
0.00000004
0.00012415
0.00000000
0.00000000
PKSTRM
o.ooo-w
0.00009
0.00029
0.00595
0.166S5
0.10740 ' :
0.5379B
0.00000 ' '
0.00000
"see legend, page 58
61
-------
Agricultural Atrazlne Use and Water Quality
A CEEPES Analysis of Policy Options
• Ground and Surface Water Exposures (relative to benchmarks) j I
POLlCYeTriazine Ban CROP=Cora '
Conventional Tillage .
; CHEMICAL*
* Nlcosulfuron
Dlcamba
Primlsutfuron
Bromoxynil
Bentazon
Metolachlor
Alachlor
Butylate
2,4-D
AVG12
0.00000000
0.00009436
0.00003069
0.00000000
0.00036095 '
0.00000368
0.00009343
0.00057665
0.00000496
PK12
0.0000000
0.0005691
0.0003462
0.0000000
0.0079435
0.0002846
0.0003158
0.0040833
0.0001612
AVG15
0.00000000001
0.00000000072
0.00000016566
0.00000000000
0.00000024685
0.00000000535
0.00000003709
0.00000001273
0.00000000003
PK15
0.0000000000.
0.0000000068
0.0000006857
0.0000000000
0.0000045131
0.0000003907
0.0000001069
0.0000000640
0.0000000008
PKSTRM
- 0.00164 ...'.-:
0.15018
0.00312
0.00008
0.17447 -
0.22064
0.37631
0.03752
•0.11466 ' .
Reduced Tillage
CHEMICAL
Nlcosulfuron
Dlcamba
Primisutfuron
Bromoxynil
Bentazon
Metolachlor
Alachlor
' 2,4-D
AVG12
• 0.000000000
0.000086086
0.000028732
0.000000000
0.000004914
0.000000694
0.000004230
0.000000050
PK12
0.00000001
0.00053943
0.00033002
0.00000000
0.00023589
0.00004068
0.00000994
0.00000156
AVG15
0.00000000001
0.00000000039
0.00000014440
0.00000000000
0.00000000015
0.00000000021
0.00000000015
0.00000000000
PK15 .
0.00000000003
0.00000000794
0.00000050861
0.00000000000
0.00000000545
0.00000002475
0.00000000095
0.00000000000
PKSTRM
0.001025
0.075680
0.002041 .
0.000023
0.045130
0.097194
0.016802-
. 0.001060
No-tillage
CHEMICAL
Nlcosulfuron
Dlcamba
Primlsutfuron
Metolachlor
Alachlor
Glyphosate
AVG12
0.000000000
0.000004024
0.000013357
0.000000049
0.000001300
0.000000000
PK12
0.00000001
0.00003250
0.00013467
0.00000403
0.00000483
0.00000000
AVG15
0.00000000000
0.00000000002
0.00000010663
0.00000000001
0.00000000003
0.00000000000
PK15
0.00000000003
' 0.00000000039
0.00000039804
. 0.00000000288
0.00000000044
0.00000000000
PKSTRM
0.0002293
0.0036173
0.0004628 .
0.0018317
0.0033859
0.0000000
•see legend, page 58
62
-------
Appendix
D
AV12(15):
PK12(15):
PKSTRM:
Benchmarks:
Average concentration of chemical (ppb) aM.2 (15) meters below
the surface .
Peak 1.2(15) meters below the surface •
Peak surface water concentrations (edge of Held loadings)
Drinking water maximum contaminant levels. (MCLs) on drinking
water health advisories . ...-•.
-------
Agrkultural Atrazine Use and Water Quality
A CEEPEStoolysis of Policy Options
64
-------
-------
f
------- |