Consultation on the Effects of Climate Change 

on Pesticide Exposure Assessment Models

November 10, 2010Table of Contents

  TOC \o "1-3" \h \z \u    HYPERLINK \l "_Toc274839036"  I. Background	 
PAGEREF _Toc274839036 \h  3  

  HYPERLINK \l "_Toc274839037"  II. Description of Approach	  PAGEREF
_Toc274839037 \h  6  

  HYPERLINK \l "_Toc274839038"  III. Screening Review	  PAGEREF
_Toc274839038 \h  8  

  HYPERLINK \l "_Toc274839039"  A. Human Health Exposure Models	 
PAGEREF _Toc274839039 \h  8  

  HYPERLINK \l "_Toc274839040"  1. Dietary Exposure	  PAGEREF
_Toc274839040 \h  8  

  HYPERLINK \l "_Toc274839041"  2. Handler Exposure	  PAGEREF
_Toc274839041 \h  10  

  HYPERLINK \l "_Toc274839042"  3. Post-Application Exposure	  PAGEREF
_Toc274839042 \h  12  

  HYPERLINK \l "_Toc274839043"  B. Ecological Exposure Models	  PAGEREF
_Toc274839043 \h  14  

  HYPERLINK \l "_Toc274839044"  1. Aquatic Exposure	  PAGEREF
_Toc274839044 \h  14  

  HYPERLINK \l "_Toc274839053"  2. Terrestrial Exposure	  PAGEREF
_Toc274839053 \h  22  

  HYPERLINK \l "_Toc274839054"  3. Spray Drift	  PAGEREF _Toc274839054
\h  23  

  HYPERLINK \l "_Toc274839055"  C. Summary of Screening Review	  PAGEREF
_Toc274839055 \h  24  

  HYPERLINK \l "_Toc274839056"  IV. Case Studies	  PAGEREF _Toc274839056
\h  25  

  HYPERLINK \l "_Toc274839057"  A. Human Health Case Studies	  PAGEREF
_Toc274839057 \h  26  

  HYPERLINK \l "_Toc274839058"  1. Dietary Exposure Evaluation ModelTM
(DEEMTM)	  PAGEREF _Toc274839058 \h  26  

  HYPERLINK \l "_Toc274839059"  2. Post-application: Agricultural
Workers	  PAGEREF _Toc274839059 \h  29  

  HYPERLINK \l "_Toc274839060"  B. Ecological Case Study: PRZM/EXAMS	 
PAGEREF _Toc274839060 \h  31  

  HYPERLINK \l "_Toc274839062"  1. Physical/Chemical and Environmental
Fate Properties	  PAGEREF _Toc274839062 \h  31  

2. Pesticide Use and Other Inputs	  PAGEREF _Toc274839061 \h  32 

  HYPERLINK \l "_Toc274839063"  C. Summary and Conclusions	  PAGEREF
_Toc274839063 \h  33  

 I. Background

The Office of Pesticide Programs (OPP) is seeking advice from the FIFRA
Scientific Advisory Panel (SAP) on a systematic approach to determine if
there is a scientific need to modify OPP’s approach to assessing
exposure from pesticides in response to the effects of climate change. 
In this SAP, OPP is focusing on data inputs and parameters used in its
human and environmental exposure models.  Changes in the Earth's
climate, such as increased temperatures and frequency and severity of
extreme weather events, have and are projected to continue to affect
cropping patterns and/or pest pressure.  Changing weather patterns may
also alter the behavior of pesticides in the environment.  OPP is
seeking advice on approaches for determining to what extent climate
change may effect pesticide exposure, whether current models are
sufficiently robust (applicable) to adequately predict any changes in
exposures, and how to identify priority areas for additional inquiry.

As an overview to the potential effects of climate change, the meeting
will begin with three presentations on some of the broader climate
issues.

The first of these presentations provides an overview of climate change
initiatives within EPA and across Federal Agencies, followed by two
discussions on areas where climate change may affect pesticide use.

EPA is working to fulfill its mission to protect human health and the
environment in a world in which the climate is changing.  Many of the
outcomes EPA is trying to attain (e.g.; clean air, safe drinking water,
safe food) may be sensitive to changes in weather and climate.  Until
now, EPA has been able to assume that climate is relatively stable and
future climate will mirror past climate (an assumption of
“stationarity”).  But with climate changing more rapidly and in
unprecedented ways, the assumption of stationarity may need to change.

EPA must anticipate and plan for future changes in climate.  It must
“mainstream” considerations of climate change into its programs and
rules to ensure they will remain effective under future climatic
conditions. Administrator Lisa P. Jackson explicitly acknowledged this
in her seven priorities for EPA’s future when she stated, “we must
also recognize that climate change will affect other parts of our core
mission, such as protecting air and water quality, and we must include
those considerations in our future plans.”

Given this imperative, EPA’s new Strategic Plan explicitly calls for
the Agency to integrate climate adaptation into its programs and rules:

The Agency must incorporate the anticipated, unprecedented changes in
climate into its programs and rules, drawing on the critical information
and tools provided by EPA researchers, to continue to fulfill statutory,
regulatory, and programmatic requirements.

The Strategic Plan also calls for EPA to collaborate with the
Interagency Climate Change Adaptation Task Force and the US Global
Change Research Program (USGCRP).  On October 5, 2010, the Task Force
delivered a report to the President entitled, “Progress Report of the
Interagency Climate Change Adaption Task Force:  Recommended Actions in
Support of A National Climate Change Adaptation Strategy” . 
Consistent with EPA’s efforts to “mainstream” climate adaptation,
the Task Force has recommended that adaptation planning be encouraged
and mainstreamed across the entire Federal Government, and Federal
agencies implement adaptation planning.  EPA is also a member of the
USGCRP, which is conducting scientific research, modeling and
assessments to evaluate the potential impacts of climate change on the
United States and evaluating alternative adaptation strategies.

The U.S. Department of Agriculture (USDA) has been evaluating the
relationship between climate change and its effects on agriculture for a
number of years.  A May 2008 report  describes a range of current and
predicted effects.  The effects of climate change on agriculture are
already apparent and expected to continue for the foreseeable future. 
These effects are both direct and indirect and vary by type and
intensity among regions of the country.  Agricultural production,
including crop and livestock, is sensitive to the direct effects of
temperature and rainfall, as well as indirect effects resulting from
increased pest pressure and pest resistance.  These changes can affect
the use of pesticides.

The Center for Disease Control (CDC) tracks the occurrence of vector
borne disease (VBD) in the U.S.  In recent years, changes have been
observed in the distribution and incidence of VBD, including the
northward movement of Lyme disease and Eastern Equine Encephalitis. 
These trends are potentially associated, in part, with a changing
climate and may be addressed through adaptive strategies, such as
greater reliance on pesticides and integrated vector control.  

As will be evident from these presentations, OPP’s work is only a very
small part of the suite of the climate change efforts that need to be
considered in the coming years.  Even within the context of pesticide
assessment, this SAP meeting addresses a fairly narrow aspect of the
potential effects climate change might have on the science of pesticide
risk assessments.  In large part, this consultation is intended to
provide OPP with input to assist it in determining the impact that
climate change may have on the magnitude and extent of pesticide
exposure and its ability to evaluate those exposures.

A team of OPP scientists and risk managers reviewed a wide range of
pesticide exposure assessment inputs and models to identify the inputs
in exposure assessment models in light of present and predicted effects
of climate change.  As discussed in further detail in section II, OPP is
initially focusing on exposure assessment tools because it has the
capacity to directly address these areas. 

OPP is using exposure assessment to illustrate some possible approaches
for considering potential effects of climate change in pesticide risk
assessment.  OPP recognizes that there are other issues involving
climate change and pesticides that are not addressed in this
consultation.  These issues include

changes in risk assessment methods or components which could be
influenced by climate change considerations, but which would encompass a
significant number of other issues as well.  These areas include, for
example, ecosystem modeling, greater use of probabilistic assessment
methodologies, and consideration of human exposure resulting from
volatilization.  These areas are being or would be better addressed in
other venues and the impact of climate change could be incorporated
directly into their development.

scientific issues which are not unique to pesticides.  For example, some
chemicals may exhibit different types of toxicity and may behave
differently in the environment under different climatic conditions. 
Such differences, however, would not arise exclusively for pesticides,
but may be similar for many chemicals in the environment.  Independent
of this consultation, OPP is reaching out across the Agency to begin to
address these cross-program issues.

benefits of pesticides, including the control of pests that carry vector
disease as well as agricultural pests.  Climate change could result in
conditions conducive to pest population growth, the expansion of pests
into new geographical areas, and the development of resistance.  Climate
changes may also influence the efficacy of alternative pest control
practices, including non-chemical alternatives.  EPA sees this as an
important area for future consideration and we plan to work with
partners from other parts of the Agency, USDA, CDC and other experts. 

This SAP meeting will focus on issues pertaining to exposure modeling. 
We are seeking advice to help us determine the extent our exposure
estimates might be influenced by changes in the inputs values as climate
changes. 

The background paper is organized as follows.  Section II provides a
brief description of OPP’s approach for considering how OPP could
evaluate the impacts of climate change on the elements to or other
aspects of its exposure models.  Section III describes the results of
OPP’s screening review of most exposure models used in conventional
pesticide risk assessment.  Section IV describes case studies that focus
on the model inputs and parameters that are potentially affected by
climate change and how the results may be affected by climate change. 
Section IV also provides a summary of preliminary conclusions.

II. Description of Approach

To a large extent, the risk assessment paradigm relies on historic data,
including data on how a pesticide is used, to evaluate current and
future pesticide exposures.  An important issue for OPP is to determine
whether there is a need, and if so, how to modify these models to
reflect future changes in climate within the time frames consistent
pesticide regulatory review cycle of at least every 15 years.

To help break the effort into manageable pieces, OPP prioritized various
components of risk assessment.  We began by focusing on the area that is
primarily within OPP’s control, exposure assessment models.  Exposure
models were selected because many are unique to pesticide assessment,
whereas hazard (toxicity) assessment tools are not solely within OPP’s
purview.  Methodological developments, e.g., probabilistic ecological
risk assessments, should consider the effects of climate change, but are
better considered in broader venues.  Moreover, the more immediate
manifestations of climate change, the increased likelihood and magnitude
of extreme weather events, may have the greatest influence on the
environmental fate and transport of pesticides. 

DEEM™) and mathematical algorithms such as the ones used for worker
and residential exposure assessments.  For each model, the team
conducted a screening review and made a preliminary determination as to
whether elements within the model are likely to be affected by climate
change.  For the purpose of this review, “elements” include both
inputs and parameters, where inputs could be results of a study or other
data that represent a specific scenario and parameters are constants
within a model that quantify linkages between inputs and outputs of the
model.  OPP selected several models as examples to solicit advice from
the SAP on how to evaluate the variability and magnitude in estimated
pesticide exposure, given the variability and magnitude of predicted
climate change.

OPP hypothesizes that some aspects of climate change will directly
affect the fate and transport of pesticides and their residues, and
therefore that exposure models which predict fate and transport may need
to change to account for the impact of climate change.  Our focus is
primarily on increasing temperatures and changing rainfall patterns,
especially the potential for more frequent occurrences of high rainfall
(runoff) events.  OPP also considers indirect effects of climate change;
that is, effects on pesticide exposure that result because of changes in
pesticide use patterns induced by changes in cropping patterns and pest
pressure.  

To summarize, the approach OPP followed to evaluate its exposure models
with respect to climate change involves three steps.

An examination of the exposure models to identify those elements that
are likely to be affected by climate change, either directly or
indirectly.  These elements include the data entered into the model and
the parameters in the model.

An assessment of how critical the changes may be to the results of the
model, i.e., in magnitude, variability, uncertainty and direction.  OPP
is particularly concerned about the possibility that assessments would
underestimate the current and future risks posed by pesticides.  This
step allows OPP to prioritize any needed modifications in the model or
the data used in the model.

A determination of how best to address OPP’s priority needs:  whether
existing data can be parsed to account for changes, existing research
will answer our questions, or whether new research or new data are
needed.

All models discussed in this document have been peer reviewed by the
SAP.  Therefore, OPP is not asking this SAP to examine the fundamental
models per se, but to consider how well the existing models might
perform in light of changed environmental conditions that may result
from climate change.  We are seeking advice from the SAP on the overall
approach to the preliminary evaluations we have completed regarding the
case studies, as well as asking for suggestions on methods to estimate
the influence of climate change on estimates of exposure both in terms
of variability and absolute values.

III. Screening Review

For the screening review, OPP examined the exposure models used in human
health and ecological risk assessments and made preliminary evaluation
as to which data inputs and model parameters are likely to be affected
by changing weather patterns.  The screening review helped guide the
selection of case studies for a more focused consideration of the inputs
that may be affected by climate change.

The two primary criteria for selecting case studies were: 1) the
likelihood that climate change may affect important elements of the
exposure assessment, and 2) the frequency with which the model is used
by OPP.  OPP also placed a higher priority on models that share
important inputs with other models.  Based on these criteria, three
models were selected, two human health models and one ecological model. 
From the human health models, OPP selected the Dietary Exposure
Evaluation Model™ (DEEM™) and the post-application occupational
exposure algorithm.  OPP selected the Pesticide Root Zone Model/
Exposure Analysis Modeling System (PRZM/EXAMS) from the ecological
models.    

This section briefly discusses each model screened and describes the
rationale for selecting certain models as the case studies.

A. Human Health Exposure Models

™ and occupational post-application as case studies.

1. Dietary Exposure 

Several models are used to assess dietary risks, depending on the
purpose of the specific assessment.  The models used include 

Dietary Exposure Evaluation Model™ (DEEM™)

Lifeline

Cumulative and Aggregate Risk Evaluation System (CARES)

Stochastic Human Exposure and Dose Simulation (SHEDS)

Calendex™.

The most commonly used dietary model is DEEM™, which provides
deterministic estimates of dietary pesticide exposure for chronic
dietary risk assessments and a probabilistic assessment for acute
dietary risk assessments.  DEEM™ predicts exposures resulting from
consumption of both food and drinking water.  

 Calendex™ is used to examine exposures to pesticides across various
pathways and routes (e.g., dermal exposure through turf uses) for
aggregate exposure assessment.  The Agency currently uses Calendex™ -
Food Consumption Index Diaries to perform this aggregation.

Although the workings of each model vary, the key inputs used in dietary
exposure assessment are similar or the same across the models.  Dietary
exposure is determined by the sum of residues in and on food and in
drinking water, given consumption of water and treated foods. 
Consumption, and body weight of the person consuming the food, can be
varied to focus on different sub-populations, including children.  That
is, dietary exposure is represented by the following generalized
equation:

 .

Where exposure is measured as milligrams (mg) of pesticide residues per
kilogram (kg) of body weight per unit of time, which is typically a day.
 Each of these inputs is discussed briefly below.

a. Residue Levels in Food 

Residue levels in food are measured in mg of pesticide per kg of food. 
Climate change could affect residues both directly and indirectly.  Some
changes, such as more intense rainfall events, could reduce residues on
food because more residues are washed off, but could increase residues
in drinking water through increased loading in runoff.  Indirectly,
changes in agricultural production and pesticide use patterns could
alter the level and distribution of residues on food or in drinking
water.  

b. Percent Crop Treated

For some refined assessments, measures of residues are weighted by
estimates of the percent of the commodity treated.  The percent of a
crop treated with a particular pesticide is influenced by market
factors, agricultural production practices, and the presence of pests at
economically damaging levels, as is the percent of an area treated
within a watershed.  Thus, depending on where and when the pesticide is
used, the percentage crop treated could either increase or decrease as a
result of climate change.

c. Consumption Values 

Consumption values are currently derived using data from USDA’s
Continuing Survey of Food Intakes by Individuals (CFSII), but OPP will
be updating the consumption information shortly using “What We Eat in
America – NHANES” survey.  Although dietary patterns may change over
time, factors unrelated to climate are expected to drive any such
changes.  Changes in food and water consumption would be captured as the
Agency updates its inputs file using the newest data.  OPP is working
closely with NHANES to develop systems for quickly incorporating new
consumption data as it is made available.

d. Body Weight 

Body weight is not likely to change as a result of climate change, but
body weight is captured in the consumption survey so it is updated when
new consumption data are incorporated.

In summary, two important elements into dietary exposure models could be
affected by climate change -- residues and percent crop treated.  Of all
of the dietary models, DEEM™ is the one most frequently used. 
Further, the important inputs in DEEM™ are similar, and many cases the
same, as those in all other dietary exposure models.  Therefore, DEEM™
was selected as a case study.  Section 4 presents a more in depth
discussion of the sources of data on dietary residues found on food and
drinking water and the percent crop treated.

2. Handler Exposure 

A handler is an individual who mixes, loads and/or applies pesticides. 
Depending on the labeled uses, OPP routinely conducts both occupational
and residential handler exposure assessments considering dermal and
inhalation routes.  For occupational handlers, OPP conducts assessments
for a range of scenarios, including those in agriculture, structural
pest control, painting, lawn care, and more.  Residential handler
assessments include residential lawn, gardens, pets, swimming pools,
paints, foggers, crack & crevice, treated articles, and rodenticides.

Changes in rainfall patterns would be unlikely to affect indoor
exposure.  Similarly, any influence of temperature increases would
likely be mitigated by indoor temperature controls.  Therefore, indoor
residential tools such as the Multi-Chamber Concentration & Exposure
Model (MCCEM) and the Screening Consumer Inhalation Exposure Software
were not further considered in this exploratory screening effort. 
Indirectly, climate changes that result in greater pest pressure could
increase indoor use.  However, OPP concludes that the effects of climate
change are more likely to be associated with activities that occur
outdoors. 

Regardless of whether the pesticide is handled on the job or by a
homeowner, exposure is based on similar factors that determine the
amount of exposure associated with a task and how much pesticide is used
for a specific type of treatment.  The major inputs used to estimate
handler exposure are unit exposure, the amount of pesticide handled in a
day (determined by application rate and area treated), the proportion of
the pesticide entering the body and the individual’s body weight. 
Handler exposure is represented by the following generic equation:

 

where exposure is measured in mg of pesticide residue per kg of body
weight per day.

An additional input of handler exposure models, percent dermal
absorption, may be influenced by increasing temperatures associated with
climate change.  Toxic effects associated with dermal exposures are
ideally measured through a route specific dermal toxicology study.  When
a route specific study is not available, an oral toxicology study can be
used, adjusted with a percent dermal absorption factor.  Although
percent dermal absorption is sometimes shown as an element in the
exposure algorithm, it is only used in cases where an oral toxicity
study is used to estimate risk.  Thus, it is fundamentally a
toxicological consideration.  As an adjustment to the available
toxicology study, it is more appropriately addressed with other
toxicological aspects of risk.  Thus, percent dermal absorption is not
considered for this exposure-focused SAP.

a. Unit Exposure 

Unit exposure measured in mg of residue per pound (lb) of pesticide
active ingredient (AI), is the amount of pesticide with which an
individual comes into contact while performing a particular task (e.g.,
mixing a powder concentrate into water, loading a backpack sprayer,
applying a pesticide with a backpack sprayer).  For each task, unit
exposure is largely a function of the formulation and application
equipment, along with types of packaging and protective equipment.  It
is not likely to be affected by climate change; changing rainfall
patterns would not likely to be a factor and there would be no indirect
effects through changes in pesticide use patterns since values are task
specific.  

b. Amount Handled per Day 

The amount handled per day is the number of pounds AI mixed, loaded or
applied by an individual in a day.  This value is a function of the use
being assessed, for example a lawn treatment.  In general, the amount is
calculated as the maximum allowable application rate multiplied by the
size of the area to be treated.  Climate change is not likely to affect
the values OPP uses for this variable.  The area treated is a function
of the site (e.g., crop) and application method; that is, it is a
function of technology, not climate or weather.

c. Body Weight 

Body weight is not likely to change as a result of climate change.

In summary, the elements of the handler exposure models are unlikely to
be affected by climate change.  The possible exception is percent
absorption into the body, which may be better addressed at a broader
level along with toxicology-related issues.  Thus, OPP did not select
these models for further review at this time.

3. Post-Application Exposure 

OPP conducts both occupational and residential post-application exposure
assessments.  As with handler exposure, OPP is not evaluating climate
change effects associated with indoor exposures at this time.  For
outdoor uses, OPP estimates exposure resulting from performing work
tasks or normal activities in and around the home.  Post-application
inhalation exposure is being evaluated in association with
volatilization in another process.  Thus, we focus on models used to
assess the amount of pesticide exposure from dermal contact.  Models
estimate exposure from direct contact with treated surfaces as well as
certain secondary exposures, such as hand-to-mouth exposure for
children.  Treated surfaces refer to surfaces, such as crops, plants,
painted decks, lawns, and pets, that have pesticide residues due to
direct application or indirect deposition.

The major inputs of post-application exposure are the amount of
pesticide on the treated surface, the proportion of the pesticide on the
treated surface that is transferred to the skin as a result of the
activity being performed, and body weight.  Some of the specific inputs
discussed below are unique to residential post-application exposure
assessment.  Post-application exposure may be represented by the
following generic equation:

 

where exposure is measured in mg of pesticide residue per kg of body
weight per day.

As mentioned previously, an additional element of the exposure models,
percent dermal absorption, may be influenced by increasing temperatures
associated with climate change.  However, as an adjustment to the
available toxicology study, it is more appropriately addressed along
with other toxicological aspects of risk.

a. Surface Residue 

Surface residue measured as mg of AI per square centimeter (cm2),
depends on: 1) the amount of pesticide applied to or that drifts onto
the treated surface, and 2) the amount of degradation that occurs over
time after an application.  Climate change, including increasing
temperature and shifting rainfall patterns, could affect surface
residues, particularly influencing the rate of degradation.  The effect
on surface residues could either increase or decrease as a result of
climate change.

b. Transfer Coefficient  

Transfer Coefficient (TC) is the amount of treated surface in cm2 that
an individual contacts per hour while performing a task.  Like Unit
Exposure in the handler equation, the TC is determined by the specific
activity.  High contact activities (e.g.; harvesting apples) have higher
TCs than those involving low contact activities (e.g.; hoeing two week
old lettuce).  OPP believes it is unlikely that climate change would
alter the values used in exposure assessments.

In residential exposures, OPP is also concerned about incidental oral
exposure by infants and children.  These models utilize a number of
parameters including surface area of the hand and frequency of hand to
mouth contact or object (e.g., toys) to mouth contact.  These latter
parameters are unlikely to be affected by climate change.

c. Exposure Time 

Exposure time is the duration of the activity during which an individual
may be exposed to residues on a given day.  In occupational settings,
this would be the hours worked per day performing a task; in residential
settings, it is the amount of time spent in activities such as gardening
or playing on a lawn.  The duration of occupational tasks is not
expected to increase as a result of climate change.  The effect on
residential activities is unclear.

d. Body Weight 

Body weight is not likely to change as a result of climate change. 

Surface Residue is an important component of post-application exposure
assessments and is the input most likely to be affected by climate
change, both directly by conditions that influence degradation and
indirectly by changes in application rate.  The agricultural
occupational post-application assessment model was selected as a case
study due to its high frequency of use by OPP and because its inputs are
representative of those used in the residential assessments.  Moreover,
farm-worker exposure is an important Environmental Justice
consideration.

B. Ecological Exposure Models

OPP environmental exposure models estimate the potential exposure of
plants and animals to pesticide residues in aquatic and terrestrial
environments.  The models also provide estimates of residues in drinking
water for use in the dietary exposure models.  OPP examined the inputs
to these models and selected PRZM/EXAMS as a case study.  Below are
brief descriptions of the aquatic models screened, followed by a
discussion of the components of models and a preliminary assessment of
whether the components may be affected by climate change.  A more
detailed description of aquatic and terrestrial models can be found in
the Agency’s website.  

1. Aquatic Exposure 

OPP uses simulation models to predict pesticide concentrations in
surface and ground water for use in both human health and aquatic
ecological exposure assessments.  Some models are generic, in that they
do not consider differences in climate, soil, topography, or crop in
estimating pesticide exposure.  Other models are more refined (higher
tier) and estimate concentrations under spatially and temporally
specific site conditions.  The results of the models can be used for
assessing both acute and chronic risks to various biological endpoints.

Rice Model

  HYPERLINK "http://www.epa.gov/oppefed1/models/water/" \l "rice"  The
Tier I Rice Model  is simple soil-water partitioning model.  It
estimates concentration in surface water based on the amount of
pesticide applied and the organic carbon partition coefficient Koc of
the chemical applied into a rice paddy for a specific volume.  The
single, screening-level concentration calculated with this model
represents both short- and long-term surface water exposure and can be
used for both ecological risk assessments for aquatic organisms and for
human drinking water exposure assessments for human health risk
assessment.

GENEEC2 and FIRST

GENeric Estimated Environmental Concentration (GENEEC2) is a Tier I
surface water model used for estimating concentrations of pesticides in
a "standard farm pond" scenario following a single, large
rainfall/runoff event.  The FQPA Index Reservoir Screening Tool (FIRST)
is a Tier I model used to estimate concentrations of pesticides in
drinking water from surface water sources.  Structurally, FIRST is very
similar to GENNEEC2, except the dimensions of the water body and
surrounding fields are larger, representing a drinking water reservoir. 
Both models are meta-models of the Pesticide Root Zone Model (PRZM) and
the Exposure Analysis Modeling System (EXAMS) models.

PRZM/EXAMS

The Pesticide Root Zone Model (PRZM) simulates chemical movement in
unsaturated soil systems within and immediately below the plant root
zone. It is a Tier 2 model that generates estimates of the amount of
pesticide in runoff and spray drift from an agricultural field.  PRZM
has capabilities for modeling the effects of soil temperature,
volatilization and vapor-phase transport, irrigation, and biotic
transformation of applied pesticide.  PRZM is also capable of simulating
site specific transport.  PRZM is often linked with the Exposure
Analysis Modeling System (EXAMS), which simulates the processes that
occur in a water body situated next to an agricultural field.  EXAMS
takes the estimates of runoff and spray drift generated by PRZM and
estimates the concentration of a pesticide and its degradates in an
aquatic environment in benthic sediment as well as in water.  Each PRZM
simulation is conducted for multiple years (usually 30 years) using
historical meteorological data to provide a probabilistic exposure
characterization for a crop scenario for aquatic and human health risk
assessment.

SCI-GROW

Screening Concentration In Ground Water (SCI-GROW) is the primary, Tier
1 model used to estimate pesticide concentrations in vulnerable ground
water for drinking water risk and environmental assessments. SCI-GROW
model was developed by fitting a linear model to ground water
concentrations with the Relative Index of Leaching Potential (RILP) as
the independent variable; as such, SCI-GROW is an empirical model.  The
RILP is a function of aerobic soil metabolism and the soil-water
partition coefficient.  The ground water concentrations are taken from
90-day average high concentrations from Prospective Ground Water
studies. Currently, OPP is developing a Tier II ground water model that
will include temporal (e.g. number of years of metrological data) and
spatial (e.g. regional crop scenarios) inputs that could be affected by
climate change. Results of this SAP review will be considered in the
model development so that the impacts of climate change on the model are
recognized.

KABAM

The Kow-based Aquatic BioAccumulation Model (KABAM) is a higher tier
model used to estimate bioaccumulation of hydrophobic organic pesticides
in freshwater aquatic food webs and subsequent risks to mammals and
birds that consume contaminated aquatic prey.  This model is based upon
work by Arnot and Gobas (2004), who parameterized it based on
polychlorinated biphenyls (PCBs) and a number of hydrophobic organic
pesticides (e.g., lindane, DDT) in the Great Lakes.  The model relies
upon a chemical’s octanol-water partition coefficient (KOW) to
estimate uptake and elimination constants through respiration and diet
of organisms in different trophic levels of the aquatic community.  The
model uses Environmental Effects Concentrations (EECs) from PRZM/EXAMS
to calculate pesticide levels in tissue.

The aquatic exposure models differ significantly in complexity and in
the inputs used between Tier 1 and higher tier models and even within
the Tier 1 models.  The Rice Model, for example, is a simple equation
relating the application rate to water concentrations given the soil and
sediment partition coefficient, Koc or Kd, on the other hand, PRZM/EXAMS
is a computer model that simulates a number of processes by which
pesticides move over and through soil into water.  At a basic level,
however, all models, except KABAM, translate an amount of pesticide
applied to an area of land, given physical and chemical properties of
the pesticide and the pesticide’s ability to move through the
environment, into an estimate of concentration in the environment. 
Depending on the model, other factors can also be considered.  KABAM
begins with the estimate of pesticide concentration in the environment
and estimates the amount of pesticide that may accumulate in animal
tissue throughout the food chain.  Because of the commonality across
aquatic models, the similar components will be discussed together.

a. Pesticide Use Information

Table 1 presents the pesticide use information used in each of the
aquatic exposure models.  The inputs may include the application rate,
the number of applications made, the interval in days between
applications, the application method, and the depth of soil
incorporation.  As shown in Table 1, GENEEC, FIRST, and PRZM/EXAMS
include all these variables while the Rice model and SCI-GROW use only a
subset.

Table 1.  Pesticide Use Information for Aquatic Models

Input 	GENEEC	FIRST	Rice	SCIGROW	KABAM	PRZM/EXAMS

Application Rate	Yes	Yes	Yes	Yes	NA	Yes

Number of Application	Yes	Yes	No	Yes	NA	Yes

Application Intervals	Yes	Yes	No	No	NA	Yes

Application Date	Yes	Yes	No	No	NA	Yes

Application method	Yes	Yes	No	No	NA	Yes

Pesticide Incorporation Depth	Yes	Yes	No	No	NA	Yes

NA = not applicable

The values used for Application Rate, Number of Applications, and
Application Intervals are all based on use directions found on pesticide
labels in order to assess the maximum amount of pesticide that can
legally be applied in the shortest period of time.  Label directions
specify the maximum application rate and typically the minimum
application interval.  The number of applications may not be explicitly
stated, but there may be a limit on the total amount of AI applied in a
season.  In some cases, however, there are no restrictions, and OPP
assesses the consequences of repeated applications of the maximum rate
at the minimum interval for the entire season, which may be the entire
year.  These label directions must be approved by OPP at the time of
Registration and again during Registration Review.  Therefore, climate
change will not, independently, result in changes to these values for
the exposure assessments.

Similarly, application method will not be altered as a result of climate
change.  OPP assesses application methods specifically allowed on the
product label, as well as all others not strictly prohibited in order to
estimate the maximum possible exposure.

Application dates, however, may be indirectly affected by climate change
if, for example, warming temperatures advance the cropping season or
permit earlier infestations of pests.  Climate change may also
indirectly affect the incorporation depth if, for example, changing
conditions induce insect pests to modify their behavior or weed seeds to
germinate at deeper or shallower levels, which in turn may lead farmers
to alter the placement of pesticides in the soil.

b. Physical and Chemical Properties

Table 2 presents the physical and chemical properties of the pesticide
that may be included in the exposure models.  With the exception of the
octanol/water partition coefficient (Kow), PRZM/EXAMS uses all the
properties listed while the Tier 1 models consider a limited number of
the variables.  KABAM is the only model that considers Kow, where it is
used to derive the time to reach a steady state.

Table 2.  Physical and Chemical Properties for Aquatic Models

Input 	GENEEC	FIRST	Rice	SCIGROW	KABAM	PRZM/EXAMS

Molecular weight	No	No	No	No	NA	Yes

Aqueous Solubility	Yes	Yes	No	No	NA	Yes

Vapor Pressure	No	No	No	No	NA	Yes

Henry’s Law Constant	No	No	No	No	NA	Yes

Octanol/Water Partition Coefficient (Kow)	No	No	No	No	Yes	No

NA = not applicable

Some of these properties may be altered by changes in average
temperatures and in precipitation.  It is well established that
temperature has an effect on vapor pressure, solubility, and Henry’s
law constant.  Partition coefficient (Kow) is also likely to be affected
by temperature.

c. Environmental Fate Properties

Table 3 lists the environmental fate properties that may be included in
various exposure models.  As with the other sets of variables, the Tier
1 models typically rely on a small subset of the environmental fate
properties to estimate water concentrations.  The Tier 2 models,
PRZM/EXAMS, rely on the whole suite of properties.  The variables do not
directly influence results in KABAM, but are considered indirectly
through the results of PRZM/EXAMS that are inputs into KABAM.

Table 3. Environmental Fate Properties for Aquatic Models

Input 	GENEEC	FIRST	Rice	SCIGROW	KABAM	PRZM/EXAMS

Hydrolysis	Yes	Yes	No	No	NA	Yes

Aqueous Photolysis	Yes	Yes	No	No	NA	Yes

Aerobic Soil metabolism	Yes	Yes	No	Yes	NA	Yes

Anaerobic Soil Metabolism	No	No	No	No	NA	Yes

Aerobic Aquatic Metabolism	Yes	Yes	No	No	NA	Yes

Anaerobic Aquatic Metabolism	No	No	No	No	NA	Yes

Soil and Sediment Partition Coefficient (Kd or Koc)	Yes	Yes	Yes	Yes	Yes
Yes

Foliar dissipation	No	No	No	No	NA	Yes

Volatilization from Foliage	No	No	No	No	NA	Yes

Plant Uptake Factor	No	No	No	No	NA	Yes

NA = not applicable

The dissipation of a pesticide from soil and water results from abiotic
degradation processes (i.e., hydrolysis and photolysis) and biotic
degradation processes (i.e., aerobic and anaerobic metabolisms). 
However, pesticide degradation in soils and water depends on complex
interactions between the physical and chemical properties and these
biotic and abiotic processes in soil/sediment and water under certain
temperature and moisture conditions.  Climate change could affect these
fate characteristics.

Soil partition coefficient (Koc) is generally influenced by the organic
matter content of soil.  Several researchers (Fallon et al, 2006 and
Grace et al, 2006) suggested that climatic shifts in temperature and
precipitation are likely to influence the amount of organic matter
present in soil, which may affect the adsorption/desorption capacity of
soil.

The dissipation and degradation processes of a pesticide on foliage are
also functions of complex interactions between physical and chemical
properties of the pesticide, photodegradation, plant uptake and wash-off
by rainfall.  Potential changes in temperature and rainfall patterns
could affect these interactions, and ultimately, the level of pesticide
concentrations in an aquatic system.

In summary, fate processes may be sensitive to temperature and moisture
conditions; therefore, changes in temperature and precipitation due to
climate change may alter the adsorption/desorption behaviors and
degradation/dissipation rates of some pesticides.

d. Other Inputs

Some of the models consider other variables, which are shown in Table 4.
 Application Efficiency and Spray Drift Fraction may be used to
determine the amount of a pesticide available for transport into surface
or ground water or transported via other routes.  These inputs are
derived from other models discussed in Section III.B.3.  KABAM also uses
the results of PRZM/EXAMS, the Estimated Environmental Concentrations
(EEC), as inputs.

Table 4. Other Inputs for Aquatic Models

Input 	GENEEC	FIRST	Rice	SCIGROW	KABAM	PRZM/EXAMS

Application

Efficiency 1	Yes	Yes	No	No	NA	Yes

Spray Drift Fraction 1	Yes	Yes	No	No	NA	Yes

Crop Specific

Scenario	No	No	No	No	NA	Yes

Percent Crop Area	No	Yes	No	No	NA	Yes

Site Specific

Meteorological data	No	No	No	No	NA	Yes

Water Column EEC 2	No	No	No	No	Yes	No

Pore Water EEC 2	No	No	No	No	Yes	No

1 May be estimated from spray drift models, AgDRIFT® and AGDISP.

2 Estimated Environmental Concentrations from PRZM/EXAMS output.

NA = not applicable

PRZM/EXAMS models various crop-specific scenarios, which include
detailed information on soil properties (chemical, physical and
hydrological), crop parameters, and agricultural practices, as well as
some non-agricultural scenarios.  These scenarios are representative of
areas of major agricultural production and of areas that, because of
soil types, may be particularly vulnerable to runoff into surface water
or leaching into ground water.

While climate does shape soil properties over a long period of time,
existing crop scenarios cover a range of geographical areas and are
likely to remain representative of soils on which pesticides may be
applied.  However, crop parameters and agricultural practices are likely
to be influenced by changes in climate.

Percent of Crop Area (PCA) is used to estimate pesticide concentrations
in drinking water.  Climate change may indirectly affect this variable
as changing temperature and rainfall patterns induce shifts in
agricultural production.

PRZM/EXAM also uses site-specific meteorological data, including
temperature, rainfall, and rainfall intensity.  Clearly, changing
climate patterns could have substantial effects on local meteorological.

e. Model Parameters

An important consideration is the presence of algorithms within the
models that quantify the linkages between inputs and outputs.  The PRZM
model has many algorithms (approximate 100+) and several are imbedded
into PRZM scenarios and many of these algorithms are temperature and
precipitation sensitive.  For example, the runoff curve number (CN), an
integral part of PRZM scenarios, is derived from an   HYPERLINK
"http://en.wikipedia.org/wiki/Empirical" \o "Empirical"  empirical 
algorithm used in   HYPERLINK "http://en.wikipedia.org/wiki/Hydrology"
\o "Hydrology"  hydrology  for predicting direct   HYPERLINK
"http://en.wikipedia.org/wiki/Surface_runoff" \o "Surface runoff" 
runoff  or   HYPERLINK
"http://en.wikipedia.org/wiki/Infiltration_(hydrology)" \o "Infiltration
(hydrology)"  infiltration  from   HYPERLINK
"http://en.wikipedia.org/wiki/Rain" \o "Rain"  rainfall .  This
algorithm is based on analysis of a large number of runoff events
studied over the 1950s through 1980s (NRCS, 1986).  The interaction of
hydrologic groups and land use treatment (cover) is accounted for by
assigning a CN value for average soil moisture condition to important
soil cover complexes for fallow, cropping, and residue parts of a
growing season.    

Many of the imbedded algorithms in the model are dependent to some
extent upon climatic factors such as temperature, humidity and rainfall.
 Most of the algorithms have the capabilities to incorporate changes due
to climate shifting, but it is unclear if more recent data would yield
significantly different algorithms, and if they would, how would they
affect the results of the exposure assessment.

As a result of the screening review, OPP selected PRZM/EXAMS as a case
study because: 1) several variables, especially among the physical and
chemical properties and the environmental fate properties, may be
affected by climate change; 2) it is commonly employed; 3) the inputs
for PRZM/EXAMS include all the inputs that are used in the Tier 1
models; and 4) results from the review of PRZM/EXAMS can be applied to
these other models. 

2. Terrestrial Exposure

OPP uses Tier 1 models to estimate pesticide concentrations on food
items eaten by terrestrial mammals, birds, amphibians, and reptiles. 
OPP also uses a Tier 1 model to assess exposures of terrestrial and
semi-aquatic plants located adjacent to areas treated with pesticides.  

TREX T-HERPS and TerrPlant

Terrestrial Residue Exposure (TREX) is a spreadsheet based model that
estimates pesticide concentrations on avian and mammalian food items. It
calculates the decay of a chemical applied to foliar surfaces for single
or multiple applications.  A first order decay function is used to
approximate the concentration at each day after initial application,
given the concentration resulting from the initial and additional
applications.  The concentration on foliage is calculated based on the
Kenaga nomogram (Hoerger and Kenaga, (1972), as modified by Fletcher
(1994).  Terrestrial Herpetofaunal Exposure Residue Program Simulation
(T-HERPS) is a simulation model that estimates exposures to terrestrial
reptiles and amphibians from pesticide use.  T-HERPS is similar to TREX
in that it uses the same underlying data set for determining initial
pesticide residues on food items.

TerrPlant provides screening-level estimates of exposure to non-target
plants from single pesticide applications.  The model considers habitats
adjacent to treated fields, including dry terrestrial habitats and
wetlands.  TerrPlant considers pesticide transport through spray drift
and runoff.

TREX and T-HERPS account for some dissipation from the application site
while TerrPlant is a function of the transport of the pesticide to
surrounding locations.

Table 5 provides a list of inputs into the terrestrial models.  All of
these pesticide use inputs, except for type of formulation, are included
in at least one of the aquatic models previously discussed.  As noted in
the discussion of aquatic models, values used in exposure assessments
are based on label restrictions, which would not change due to climate
change.  Formulation type, as well, is subject to OPP approval.  Thus,
none of these inputs except for foliar dissipation and aqueous
soluability would be influenced by climate change.

Table 5. Pesticide Use Inputs for Terrestrial Models

Input Description	TREX/T-HERPS	TerrPlant

Application Rate	Yes	Yes

Number of Application	Yes	Yes

Application Intervals	Yes	No

Type of Formulation	Yes	Yes

Foliar dissipation	Yes	No

Aqueous Solubility	No	Yes

Spray Drift Fraction	No	Yes

Runoff Fraction	No	Yes



Results of the TREX model depend on the foliar dissipation rate,
discussed under environmental fate properties for the aquatic exposure
models.  It is likely to be influenced by changes in temperature and
rainfall patterns.

Results of TerrPlant depend on aqueous solubility, discussed under
physical and chemical properties for the aquatic exposure models. 
Temperature is likely to affect this input.  TerrPlant also uses the
fraction of the pesticide transported via spray drift or runoff.  The
value used for spray drift may be a default value or it may be
estimated.  Estimation of spray drift is discussed in the following
section.  Runoff is very likely to be affected by changing rainfall
patterns, especially high-intensity events.

Several inputs into terrestrial models are also likely to be affected by
climate change.  These inputs are largely similar to those found in the
aquatic exposure models.  Therefore, OPP did not choose any of these
models for this consultation.

3. Spray Drift 

AgDRIFT® and AGDISP

OPP uses two models for estimating spray drift and deposition from
treated areas to adjacent areas where non-target organism may be
exposed; AgDRIFT® and AGDISP(AGricultural DISPersal).  Both models are
used for estimating drift from aerial applications through a mechanistic
approach. AgDRIFT is also used for estimating drift from ground boom and
air-blast applications.  The models cover a wide variety of application
methods, equipment and techniques that are used to apply pesticides. 
Climate change is not likely to significantly affect the type of
equipment used.

Pesticide use information is an input to these models, including the
maximum application rates, maximum number of applications per season,
and minimum application intervals permitted on the pesticide label are
typically used to yield the maximum exposure estimate from legal use. 
As noted previously, when such parameters are used, they can only change
with OPP approval.  

Meteorological conditions, such as wind speed, atmospheric stability,
relative humidity and temperature all have the potential to change in
light of climate change at any given time and place.  For wind speed and
atmospheric stability, however, many label requirements and/or best
management practices limit the conditions under which pesticide
applications are made.

The AgDRIFT® model contains empirical curves for ground and airblast
applications from which spray drift is estimated.  The parameters of the
curves and algorithms quantify pesticide drift under a defined
application scenario, given the amount applied and the application
method used.  If meteorological conditions change substantially and
require modifications to application equipment, these parameter values
may not longer accurately estimate the amount of drift.

Climate change could affect the spray drift models directly because the
meteorological data used in the models will need to reflect conditions
in the field.  Climate change could also affect the spray drift models
indirectly, if conditions change substantially from those on which the
models were parameterized.  We are not reviewing these models in greater
depth at this time, however, because the key components that may be
affected by climate change are represented in the PRZM/EXAMS case study.

C. Summary of Screening Review

The screening review was to examine whether inputs and parameters may be
affected by climate change.  The review suggests that some of the
components of OPP’s pesticide exposure assessments are may be affected
by climate change, but that there are differences among the models for
dietary exposure, occupational and residential exposure, and ecological
exposure.  Dietary exposure models are most likely to be affected by
climate change indirectly, through changes in pesticide use patterns as
humans and pests adapt to changes in climate.  Use patterns are unlikely
to affect other exposure models, however, because OPP evaluates exposure
at the limits of legal use.  The occupational and residential exposure
models and ecological exposure models are more likely to be directly
affected by climate change, where temperature and changing rainfall
patterns could alter the way a pesticide dissipates or degrades or moves
through the environment.

OPP needs to better understand the magnitude of effects climate change
may have on its exposure models in order to set priorities for
addressing any effects.  For this presentation, OPP has selected three
models for more thorough review.  For dietary exposure, we chose DEEMTM
because it is most frequently used and is representative of the other
dietary exposure models.  We selected the post-application exposure
worker model because it is commonly used, representative of other
post-application models, and raises potential environmental justice
concerns.  Finally, we selected the linked PRZM/EXAMS model, which is
frequently used and includes nearly all data inputs considered in other
ecological exposure models.  

IV. Case Studies 

In this section, OPP begins to evaluate the influence of climate change
to our exposure assessments in order to better identify any issues that
need to be addressed and evaluate the need for any revisions in the
context of the overall significance in exposure estimates.  Ultimately,
OPP wants to identify any inputs and parameters that could change
sufficiently during the maximum 15-year reliance on any given exposure
assessment to the extent that they would significantly increase OPP’s
exposure estimates. 

The importance of climate change to OPP’s exposure assessments will
depend on how sensitive the model results are to changes in the input or
parameter in question and the magnitude of change that may be induced by
alterations in climate.  In assessing the latter, it is important to
bear in mind the time frame over which different aspects of the climate
may change and the time period over which OPP relies on the outcome of
any particular assessment.  OPP reviews a pesticide’s registration at
least every 15 years, and often reviews more frequently.  OPP seeks to
understand if the rate of change or variability in temperature and
rainfall, and the timing of such changes will significantly increase our
exposure estimates over the time for which the Program relies on an
assessment.

The effects of climate change are also likely to differ by geographic
area.  Some areas may become hotter and dryer, whereas others may have
greater rainfall and more extreme weather events.  Both regionally and
nationally, climate change will be gradual relative to the maximum
duration between pesticide risk assessments.  The Agency exposure
assessments address a range of geographic conditions that currently
exist across the country.  The issue for OPP is whether climate change
will affect the range of conditions evaluated in OPP’s models faster
than OPP can update the input parameters from ongoing monitoring. 

A. Human Health Case Studies

1. Dietary Exposure Evaluation ModelTM (DEEMTM)

We previously identified two elements of DEEMTM that may be influenced
by climate change:  the residues found on food items or in drinking
water and the percent of the crop or area around sources of drinking
water that is treated with a pesticide.

 Evaluating Pesticide Residues on Food

Residues in or on food could either increase or decrease as a result of
climate change.  Warmer temperatures and increased rain events could
reduce residues by speeding degradation and/or removal.  Areas
experiencing less rainfall could have slower degradation of pesticide
residues.  Moreover, in cases where climate change results in higher
pest pressure, it is possible that pesticide application rates and the
frequency of application could increase, which in turn, could increase
pesticide residues on food.  Residues are an important element in the
outcome of an exposure assessment.

In a dietary exposure assessment, OPP can use three different values to
estimate the level of residues in food:  tolerance levels, field
residues, and/or monitored levels.  Tolerance levels and field residues
are both based on magnitude of residue field studies using maximum
labeled rates while monitored levels come from measurements taken from
food samples.

Magnitude of residue (MOR) studies are designed to determine the maximum
residues that could result in food or animal feed from the legal use of
the pesticide.  There are many types of MOR studies including crop field
trials, processing studies, livestock feeding studies, livestock dermal
application studies, application to stored commodities, residues in
foods as a result of direct treatment of irrigation water, and residues
in fish.  MOR studies are used to set tolerance levels as well as to
estimate of residues for dietary risk assessment.  This discussion will
focus on crop field trials, since they are most likely to be affected by
climate changes. 

Crop field trials are controlled studies where the pesticide is applied
to test plots under conditions that are likely to lead to the highest
residues:  maximum application rate, minimum interval between
applications, and minimum interval between the final application and
harvest, as specified on the label.  OPP guidance on crop field trials
(Guideline No. 860.1500) recommends the minimum number of field trials
for each crop as well as the locations for conducting the field trials. 
These locations represent the major growing areas of the crop, and the
number of trials in each region reflect the relative amount that region
produces.  For example, if half of a given commodity that the US
consumes is grown in the Northeast, half of the field trials should be
conducted in the Northeast.  The US, along with its NAFTA partners, has
created a North American Regional Map, which specifies the regions
described in the crop field trial guidance.  Some of the US regions
extend into Canada and Mexico.

OPP uses the results of the crop field trial data to establish a
tolerance level, which is generally higher than the highest residue
level found in the field studies.  FDA, which is responsible for
tolerance enforcement of plant-based food commodities, rarely finds
domestic commodities with over-tolerance residues.  This indicates that
the crop field trials accomplish what they are designed to do; that is,
to determine the upper bound of residues from legal use of the
pesticide.  In an initial screening dietary exposure assessment, OPP
assumes that all of the treated crops consumed in the US are at the
tolerance level, which is an overestimate.  Should this assessment
exceed the level of concern, the first step in refining the estimate is
to use the actual field residue data, as opposed to a high end estimate.

Generally, the field residue data will represent a variety of growing
conditions.  Because the underlying studies are conducted in a wide
variety of areas and conditions across the country, it is likely that
they will continue to be representative even with moderate changes in
climate over the coming few decades.  The Agency periodically reviews
the crop field trial guidance to reflect changing dietary consumption
patterns and changes in the major growing areas.  The last major change
to the 860.1500 guidance occurred in 1996.

Moreover, the pesticide use patterns in the trials represent those that
would result in the highest levels of residues.  Should there be a need
to increase application rates or reduce the interval between application
rates and harvest, the pesticide registrant would be required to submit
new crop field trials so the Agency could determine whether the changing
use conditions could result in increased pesticide exposure and also to
determine if the current tolerance level is adequate given the change in
how the pesticide can be applied.

Monitoring data are measurements of residues in and on samples of food
products.  The Agency frequently uses data from the USDA Pesticide Data
Program (PDP), which measures residues closer the dinner plate than
field trial data.  These data reflect actual use patterns in the field,
which could include lower application rates, less frequent applications,
and longer intervals between application and harvest than is specified
on the label.  Changes in the climate could alter environmental
conditions to be more favorable to pests, increasing pest pressure and
inducing agricultural producers to increase the use of pesticides. 
Therefore, the monitoring data obtained today may not be reflective of
residues in foods in the future.  

A key issue is the speed at which climate changes alter pest pressure
compared to how frequently PDP monitoring data are updated.  Monitoring
data are collected annually, although individual commodities may be
sampled less frequently.  By design, the commodities with higher residue
levels are monitored more frequently than those with relatively low
residue levels.  In addition, USDA works closely with OPP to identify
commodities that may be of particular regulatory interest and adjust the
commodities that are monitored accordingly.  Residue data for
commodities with the greatest potential to contribute to dietary
exposure are updated regularly to account for changes in use patterns
that could result from many factors including climate change.  Based on
the frequency of monitoring and the ability to modify the monitoring
program to focus on commodities of interest, OPP believes that PDP data
will continue to provide robust exposure estimates in light of climate
change.  For further detail on the frequency of monitoring by commodity
is available on the PDP web site.

 

b. Evaluating Pesticide Residues in Drinking Water

Pesticide residues or concentrations in drinking water are estimated
with the aquatic exposure models discussed in the Ecological Exposure
sections, Sections III.B. and IV.B.  Pesticide concentrations may be
affected by changes in climate, particularly temperature.  Change in
temperature can affect many of the physical and chemical properties as
well as the environmental fate properties of pesticides.  Concentrations
also depend on cropping patterns and pesticide use patterns.  However,
for modeling purposes, EPA relies on label restrictions, such as maximum
application rate and/or minimum interval between applications, to
estimate concentrations, and these values will not change without EPA
approval.  See Section IV.B. for more details.

c. Evaluating Percent Crop Treated

The Percent of a Crop Treated (PCT) with a particular pesticide can
change over time as a result of year-to-year variation in pest pressure,
weather conditions, and the availability of alternative pest control
methods, including products exiting the market and new products entering
the market.  Climate change may also affect the PCT if changing
conditions allow pests to spread more widely in space and time, leading
growers to treat relatively more acreage.  Changing conditions may also
result in shifts in cropping patterns, bringing crops into areas with
existing pest problems.  Changing conditions could also result in the
opposite effect, if conditions became less conducive to pests or
resulted in agriculture shifting away from areas with pest problems.

Currently, OPP uses PCT in refined risk assessments to estimate the
proportion of a commodity that would not have residues.  This refinement
can be applied with any of the three types of residue levels used (i.e.;
tolerance, field trial or monitoring) in the assessment.  The results of
the assessment can be sensitive to the PCT values used.

OPP has two methods for estimating PCT for use in the dietary exposure
assessment.  For new pesticide uses, OPP uses the PCT of the most widely
used pesticide (the market leader) within a similar class (e.g.,
herbicide, insecticide, fungicide) as a proxy for the PCT of the new
pesticide.  This is referred to as the projected percent crop treated
(PCTn).  This sets a very conservative upper bound, at least over a
five-year time period.  Refinements to the PCTn can occur if needed by
looking at the market leader for a specific pest/crop combination which
coincides with the new pesticide’s pest claims.  For existing
pesticide uses, OPP uses a ten-year average of the annual PCT of the
pesticide.  Data on PCT, for both methods, come from surveys conducted
by the National Agricultural Statistics Service (NASS) of the U.S.
Department of Agriculture.  Most crops are surveyed every other year. 
OPP supplements these data with data from private marketing surveys,
which are often performed annually.  Given climate change, these data
from past use, particularly for existing pesticide uses, may not reflect
future PCT.  As with residue data, a key question is the speed at which
changing pest and cropping patterns are expected to occur.

2. Post-application: Agricultural Workers

Surface Residue was identified to be the major input for predicting
post-application exposure for agricultural workers that may be affected
by climate change.  For this particular model, Surface Residues are
estimated as Dislodgeable Foliar Residues (DFR).

OPP estimates DFR as a function of time in order to assess the exposure
of workers who enter a field some days following a pesticide
application.  A basic formulation hypothesizes that some amount of the
total amount applied is immediately available on the surface of the
foliage and that residues decline over time and is represented as:

 

where DFR(0) is the proportion of the pesticide applied that is
immediately available to be dislodged (or transferred) from the plant
surface, D is the fraction of residue that dissipates daily, and t is
measured in days after treatment.  Data to estimate the parameters of
this function come from crop residue studies where foliage samples are
collected over time following application.  As with the field trials
mentioned in the previous section, these studies are conducted in
various geographical regions of the country to account for differences
in temperature and rainfall.

In assessing exposure, OPP assumes the maximum application rate
permitted by the label is used.  As the label rate is under OPP control,
it will not change independently due to changes in climate.  Thus, there
would be no indirect effects of climate change on the model.  The issue
is whether changes in climate would affect the parameters of the
equation, DFR(0) and D, such that data generated in the past would not
be relevant for assessing future exposure.

a. Dislodgeable Foliar Residue (DFR(0))

DFR(0) is the proportion of a pesticide on the treated surface available
to be dislodged immediately after application.  This proportion would be
used to estimate the maximum expected surface residue concentration. 
Climate change is unlikely to have a significant impact on this
parameter.  It may be that higher temperatures would lead to increased
volatility, reducing deposition, but this suggests that exposure
assessments might be conservative over future conditions.

b. Dissipation (D)  

With warmer and moister climates, degradation and elimination of
residues may be hastened.  Conversely, a drier climate may prolong the
degradation, although some chemicals may photodegrade faster if a drier
climate is associated with increase ultraviolet light penetration. 
Thus, the effects of climate change may depend on nature of changes in
various geographic areas along with the inherent physical and chemical
properties of the pesticide.  Given that the data used by OPP are
already from diverse climatic regions, climate change may not a
significant impact on OPP’s exposure assessments for occupational and,
by extension residential post-application scenarios. 

B. Ecological Case Study: PRZM/EXAMS

The screening review identified components of the PRZM/EXAMS model that
may be affected by climate change, which OPP places into a few
categories:  pesticide behavior in the environment, pesticide use and
other inputs.  As for most pesticide behavior, most of the physical and
chemical properties and the environmental fate properties that are used
in the model are likely to be affected, primarily due to the change in
temperatures.  Pesticide behavior in soils and water, however, is
determined by complex interactions and the models, which were developed
and parameterized, under certain meteorological conditions, may be less
accurate under substantially different conditions because parameter
values no longer represent physical and chemical processes.

Pesticide use and other inputs include pesticide use practices, of which
the application date was the only variable identified as likely to
change over time due to the indirect effects of changes in climate. 
Other factors embodied in the crop scenarios that OPP employs and these
crop parameters and typical agricultural cultural practices may change
with climatic conditions.  Similarly, the percent of crop area in
watersheds, which is an input into the model estimating pesticide
concentrations in drinking water, may change.  Clearly, the
meteorological variables are directly tied to climate change and the
data used today may not reflect conditions in the future.

1. Physical/Chemical and Environmental Fate Properties

Physical and chemical properties such as aqueous solubility, vapor
pressure, and Henry’ law constant, have considerable impact on the
rate and pathway of pesticide dissipation and degradation.  Data on
these properties come from studies that are often conducted at
temperatures ranging from 20 to 30oC.  EPA reports that the global
average surface temperature is likely to increase by 1 to 6 oC by the
end of the century, with temperatures in most of North America rising
more than average.  More of this increase will come in winter months
rather than in summer.  Thus the direct effect of higher temperature on
these properties is likely to be manifested slowly.

However, pesticide fate behavior in soils and water depends not only
temperature, but also largely on a complex interaction with
physico-chemical properties, along with abiotic and biotic processes in
water, soil/sediment, and foliage under certain temperatures and
moisture conditions.  There is a possibility that these properties will
be affected to some extent by climate change, but it would be difficult
to predict the overall impact on exposure assessment.  For example,
changes in temperature and precipitation due to climate change may alter
the soil partitioning behavior (Kd or Koc ) of pesticide.  Pesticide
adsorption to soil depends on both the chemical properties of a
pesticide (e.g. water solubility) and properties of the soil (e.g.,
organic matter and clay contents, pH, permeability).  For most
pesticides, organic matter is the most important soil constituent
controlling the degree of pesticide retention.  Temperature also induces
the metabolic degradation rate of soil organic matter that may affect
the retention capacity of applied pesticide.  The interaction between
partitioning, dissipation and degradation of pesticide in various media
is a very complex process, and it would be a challenge to isolate a
dominant pathway driven by climate change.  Bloomfield et al, (2006)
concluded that the effects of climate change on pesticide fate and
transport is likely to be very unpredictable and would difficult to
estimate.

Although various fate inputs are derived at temperatures ranging from 20
to 30oC, the model can adjust the degradation/dissipation rates based on
meteorological data.  PRZM has a routine to correct temperature
dependent degradation and dissipation rates, which is based on the Q/10
equation similar to Arrhenius’s equation.  Also, soil temperatures
could increase, and models can easily adapt to climate change provided
underlying data are representative of the observed changes.  However,
having quality data on soil temperature could be important.

OPP concludes that environmental fate characteristics are likely to
change as a result of climate change.  It is unclear if the changes
would be significant over a fifteen year period the maximum period
between reviews of any pesticide.  It is also unclear if there are
specific geographic areas that are more likely to be affected by climate
change.

2. Pesticide Use and Other Inputs

PRZM/EXAMS uses a suite of fields, orchard crop and non-agriculture
scenarios that are representative of areas of major agricultural crops
grown and non-agriculture activities.  Each scenario contains detailed
information on soil properties (chemical, physical and hydrological),
crop inputs and parameters, and agricultural cultural practices.  The
soil inputs selected are typical values, i.e., representative soil
properties as compiled from soil survey reports.  Soil hydrological
parameters and runoff curve numbers developed by USDA are   HYPERLINK
"http://en.wikipedia.org/wiki/Empirical" \o "Empirical"  empirical 
parameter imbedded into crop scenarios for predicting direct   HYPERLINK
"http://en.wikipedia.org/wiki/Surface_runoff" \o "Surface runoff" 
runoff  or   HYPERLINK
"http://en.wikipedia.org/wiki/Infiltration_(hydrology)" \o "Infiltration
(hydrology)"  infiltration  from   HYPERLINK
"http://en.wikipedia.org/wiki/Rain" \o "Rain"  rainfall .  The crop
inputs, including application dates, were obtained by contacting county
extension agent of crop interest.  Crop Scenarios are likely to change
over time.  It is unclear whether significant changes in cropping would
occur during a 15-year pesticide review cycle.  However, OPP currently
adds new scenarios and update existing ones on an ongoing basis to
ensure that the scenarios remain current with respect to current
agricultural practices and weather patterns in the U.S.

Percent of Crop Area (PCA) may change, but for most crops, OPP uses
conservative default values.  For the crops for which there is an
established PCA, the model can accommodate an alternative PCA,
consistent with data collected by USDA.

PRZM/EXAMS uses meteorological data files containing measurements taken
at 237 weather stations located throughout the United States for a
period extending from 1961 to 1990.  OPP plans to acquire updated
meteorological data files on an ongoing basis to ensure that the weather
files remain current with respect to current agricultural practices and
weather patterns in the United States.  Although updated meteorological
data will better reflect changing climatic conditions since 1990, it is
not clear how best to use the full dataset of historic data to predict
exposure in light of climate change.  The most profound changes
resulting from climate change may be in the distribution of certain
weather events.  

C. Summary and Conclusions 

1. Climate change is occurring and is projected to continue to occur. 
These changes have and will affect pest pressure.  The changes in pest
pressure are likely to affect how pesticide are used and will be used in
the future.  Changes in use patterns will affect the exposure to
pesticides. 

2. Provided up-to-date inputs, such as food residue and weather data,
the current models used for conventional pesticides can, to a large
extent, continue to provide reliable predictions of pesticide exposure
for proposed and existing pesticide uses, even though climate change
will alter the environmental conditions under which such uses occur in
the future.

3. Under the Registration Review Program, all pesticides are reevaluated
at least every 15 years.  It is not clear if the rate of change in
weather patterns will occur at a pace such that uncertainty in our
exposure assessments will significantly increase.

4. Climate Change will alter the kinds of pests and where they are found
and will alter agricultural practices (e.g.; where crops are grown) --
both of which could indirectly affect pesticide exposure and risks.  We
need to consider the pace of such changes -- whether significant changes
will occur at a rate faster than the maximum time between risk
assessments under EPA’s Registration Review program.  Such an
understanding would be useful in deciding how climate change might
affect the currency of EPA’s pesticide exposure assessments.

 

t

Ë

痿ĈᤀË

.

Ž

摧垅

☊଀᥆㜀$␸䠀$摧ȶÇ

ഀ׆Āː㜀$␸䠀$摧㾱3

gdß

hß

 hß

hß

 hß

 hß

 h2

h2

  h2

 h2

h2

 h2

h2

  h2

- h2

h2

 h2

h2

 h2

 h2

h²

&

D

E

h

i

h

ty of conditions that pesticide users might encounter.  We recognize
that as a consequence of climate change, environmental conditions at a
test site will not necessarily be the same as those which existed in the
past, raising the possibility that prior studies may not predict what
would happen in the future.  Therefore, we need to consider whether the
variability attributable to climate change would be greater than the
variability that exists across test sites throughout the country and
therefore the extent to which EPA can continue to rely on predictions
derived from existing data intentionally collected from across the
country.  

6. As OPP acquires updated weather data, it is unclear how these data
could be best used in light of climate change.  For example, should the
more recent weather data be weighted more heavily than older weather
data or might there be a need to consider incorporative a predictive
component to account for predicted changes in climate.  Given the 15
year maximum between reviews of individual chemicals, such modeling may
not be needed. 

7. Many model parameters and algorithms that could reasonably be
affected by climate change.  Some of the imbedded model parameters (e.g.
CN) were developed using data from studies conducting in the 1980s and
earlier.  It is unclear if more recent data would yield significantly
different parameters and if they would, to what extent, would they be
likely to affect the results of the exposure assessment.

8. Many inputs that might be affected by climate change may increase or
decrease exposure.

 http://www.usda.gov/oce/climate_change/sap_2007_FinalReport.htm

 http://www.whitehouse.gov/sites/default/files/microsites/ceq/Interagenc
y-Climate-Change-Adaptation-Progress-Report.pdf

 http://www.usda.gov/oce/climate_change/sap_2007_FinalReport.htm

 Volatilization was subject of a separate SAP meeting in December 2009.

 Centers for Disease Control and Prevention’s National Health and
Nutrition Examination Survey (NHANES)

   HYPERLINK "http://www.epa.gov/pesticides/science/models_db.htm" 
http://www.epa.gov/pesticides/science/models_db.htm 

 http://www.ams.usda.gov/AMSv1.0/getfile?dDocName=STELDEV3003972

   HYPERLINK
"http://www.epa.gov/climatechange/science/futuretc.html#projections" 
http://www.epa.gov/climatechange/science/futuretc.html#projections .

   HYPERLINK "http://websoilsurvey.nrcs.usda.gov/app/HomePage.htm" 
http://websoilsurvey.nrcs.usda.gov/app/HomePage.htm 

   HYPERLINK
"http://www.mi.nrcs.usda.gov/technical/engineering/neh.html" 
http://www.mi.nrcs.usda.gov/technical/engineering/neh.html 

 PAGE   2 

 PAGE   21 

