EPA 600/R-03/024
Research Plan: Effects of Chemical Herbicides
and Gene Flow on Non-target Plants
Pesticides Research Project
National Health and Environmental Effects Research Laboratory
Western Ecology Division
200 SW 35th St.
Corvallis, Oregon 97333
-------
To address the needs of OPP and OPPT, EPA's Office of Research and Development (ORD) is conducting research
and fostering the sound use of science and technology to provide scientific information to facilitate health and ecological risk
assessments. ORD has developed a multi-year plan (MYP) to establish long-term research goals and to coordinate research
among different research laboratories and centers concerning health and ecological effects from pesticides and genetically
engineered plants. Under the ORD MYP, the National Health and Environmental Effects Research Laboratory, Western
Ecology Division (WED) in Corvallis, will identify and understand the specific terrestrial effects of pesticides (focusing on
herbicides) and genetically engineered plants.
Within WED, the Pesticides Project has been established to develop tools to improve EPA ecological risk
assessments for use chemical herbicides and genetically engineered plants. With these tools, ecological risk assessments
will be better equipped to predict potential effects of chemical herbicides and engineered plants upon important agricultural
and ecological endpoints, i.e., for agricultural ecosystems, crop quality and yield; arid for non-agricultural or native plant
ecosystems, ecosystem structure and function, especially as they relate to wildlife habitat and the viability of threatened
and/or endangered species. The Pesticides Project has developed this Research Plan "Effects of Chemical Herbicides and
Gene Flow on Non-Target Plants" which identifies three research goals relating to terrestrial ecosystems to address ORD's
long-term goals and OPP and OPPTs ecological research needs. These goals are to: 1) determine ecological effects of gene
flow from transgenic crops, 2) develop regional analysis and interpretation tools, and 3) determine effects of chemical
herbicides on non-target crops and native plants. The research to address these goals is described in three strategic
components of the Research Plan: Regional Analysis and Interpretation, Effects of Chemical Herbicides on Terrestrial
Plants, and Ecological Effects of Gene Flow from Transgenic Crops.
The Regional Analysis and Interpretation Research will develop a system to collect, analyze and interpret data for
use in the Problem Formulation and Risk Characterization phases of assessing risks from chemical herbicides and GM crops.
Data and model components will be obtained through collaborative efforts with federal, state and local agencies, as well as
industry as feasible. The analysis will use Geographic Information System (GIS) to carry out the assessments on a regional
basis. The GIS research will provide tools for spatially locating plant species potentially at risk from use of a new product, as
well as phenology (e.g., timing of occurrence of developmental events during plant life-cycle, such as flowering) of non-target
plants relative to timing of pesticide application. The system will provide a basis for selecting appropriate test species and
response endpoints for risk assessments. The GIS platform will then be used to characterize risk, by combining exposure models
and relevant plant response data in a probabilistic framework. As part of this regional analysis effort, plant responses relevant
to ecosystems (e.g., species composition, productivity) will be recorded as possible input parameters for wildlife habitat models
that are being constructed by the WED Terrestrial Habitat Project.
The Effects of Chemical Herbicides on Terrestrial Plants Research will develop methodologies to test effects of
chemical herbicides on individual terrestrial plant species and communities. It will use outputs from the regional analysis
research to identify crop and native plant species, herbicides, and herbicide treatment (e.g., timing, concentrations) conditions
important for specific areas of the United States. Experimental protocols for terrestrial plant tests will be refined for application
to nontraditional species (e.g., perennials or woody species) and response endpoints (e.g., seed yield or other reproductive or
developmental parameters). Research will also be conducted to develop molecular or cellular tools to extrapolate responses to
non-tested species or to verify field exposures. The plant test research will focus on low-dose, high-potency herbicides, but also
use some high-volume compounds. The protocols developed for chemical herbicide testing will also be adapted for use with
other industrial chemicals.
The Ecological Effects of Gene Flow from Transgenic Crops Research will develop molecular methods to detect
the presence of transgenic or other marker genes, evaluation of gene flow from engineered to non-engineered plants,
measurement of potential ecological effects of gene flow on plant community structure and function, and definition of inputs
for a prototype model for gene flow. Experiments will be conducted at various scales, ranging from contained laboratory, to
growth chamber or greenhouse, to field. Current advances in genomics and proteomics will be evaluated for their ability to
identify potential adverse effects of gene flow in agronomic and non-agronomic ecosystems.
Overall, this research project will provide tools to assist EPA in its regulatory role in registration of chemical
herbicides and genetically engineered crops that produce chemical pesticides; thereby promoting sustained productivity of
agricultural crops while maintaining the ability of ecosystems to support wildlife and to carry out other essential services.
The tools also will aid post-registration monitoring to determine the success of registration restrictions in protecting non-
target crops or native plants. Furthermore, the will be useful for determining ecological effects of other chemical pesticides
and industrial chemicals.
-------
Disclaimer
The information in this document has been funded wholly or in part by the U.S.
Environmental Protection Agency. It has been subject to the agency's peer and
administrative review. It has been approved for publication as an EPA document.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
Cover Photos: Upper left- greenhouse experiments with crops for herbicide
toxicity test development, upper right- Geographic Information Systems analysis of
counties with risk from non-target herbicide drift, lower left- contig assembly showing
the automated sequencing chromatogram trace DNA data, lower right- wild mustard
growing adjacent to agricultural fields
-------
Executive Summary
This project supports EPA's mission to protect human health and to safeguard the natural
environment — air, water, and land — upon which life depends. Specifically, we address EPA's
responsibility to prevent pollution and reduce the impacts from pollution to communities and
ecosystems (Government Performance and Results Act (GPRA) Goal 4, "Safe Communities").
To achieve this goal, EPA's Office of Prevention, Pesticides, and Toxic Substances (OPPTS)
requires scientifically credible information and methods for use in assessing health and
ecological risks from products used in commerce, including chemical pesticides and genetically
engineered plants. OPPT regulates chemical and biological pesticides primarily under the
Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) administered through the Office of
Pesticide Programs (OPP). Other acts and programs, especially the Toxic Substances Control
Act (TSCA), and the Federal Food, Drug and Cosmetic Act (FFDCA) are administered by
OPPTS's Office of Pollution Prevention and Toxics (OPPT) to provide for protection of the
environment from chemicals and biological pesticides. In the past, protection of ecological
resources has received minimal attention under these regulations compared to concerns
regarding impacts on human health. Recently, however, awareness of adverse effects from drift
of new low-dose high-toxicity herbicides to non-target crops and native vegetation has
heightened awareness of the need to improve tests for effects of chemical herbicides to plants.
Similarly, public concern regarding the release of genetically engineered plants and the adoption
of the "Final Rules and Proposed Rules for Plant-Incorporated Protectants" (40CFR Parts 152
and 174) have increased the need for tools to evaluate the risks from engineered plants and gene
flow from engineered crops to other plant species. Thus, OPP and OPPT need tools to assess
ecological risks from transgenic crops, improved methods for spatially explicit ecological risk
assessments, new methods to provide for efficient and effective gathering and interpretation of
herbicide hazard identification and dose-response data, and investigations of the potential effects
of high priority hazards.
To address the needs of OPP and OPPT, EPA's Office of Research and Development (ORD)
is conducting research and fostering the sound use of science and technology to provide scientific
information to facilitate health and ecological risk assessments. ORD has developed a multi-year
i
-------
plan (MYP) to establish long-term research goals and to coordinate research among different
research laboratories and centers concerning health and ecological effects from pesticides and
genetically engineered plants. Under the ORD MYP, the National Health and Environmental Effects
Research Laboratory, Western Ecology Division (WED) in Corvallis, will identify and understand
the specific terrestrial effects of pesticides (focusing on herbicides) and genetically engineered
plants.
Within WED, the Pesticides Project has been established to develop tools to improve
EPA ecological risk assessments for use chemical herbicides and genetically engineered plants.
With these tools, ecological risk assessments will be better equipped to predict potential effects
of chemical herbicides and engineered plants upon important agricultural and ecological
endpoints, i.e., for agricultural ecosystems, crop quality and yield; and for non-agricultural or
native plant ecosystems, ecosystem structure and function, especially as they relate to wildlife
habitat and the viability of threatened and/or endangered species. The Pesticides Project has
developed this Research Plan "Effects of Chemical Herbicides and Gene Flow on Non-Target
Plants " which identifies three research goals relating to terrestrial ecosystems to address ORD's
long-term goals and OPP and OPPT's ecological research needs. These goals are to: 1) determine
ecological effects of gene flow from transgenic crops, 2) develop regional analysis and
interpretation tools, and 3) determine effects of chemical herbicides on non-target crops and
native plants. The research to address these goals is described in three strategic components of
the Research Plan: Regional Analysis and Interpretation, Effects of Chemical Herbicides on
Terrestrial Plants, and Ecological Effects of Gene Flow from Transgenic Crops.
The Regional Analysis and Interpretation Research will develop a system to collect,
analyze and interpret data for use in the Problem Formulation and Risk Characterization phases of
assessing risks from chemical herbicides and GM crops. Data and model components will be
obtained through collaborative efforts with federal, state and local agencies, as well as industry as
feasible. The analysis will use Geographic Information System (GIS) to carry out the assessments
on a regional basis. The GIS research will provide tools for spatially locating plant species
potentially at risk from use of a new product, as well as phenology (e.g., timing of occurrence of
developmental events during plant life-cycle, such as flowering) of non-target plants relative to
timing of pesticide application. The system will provide a basis for selecting appropriate test species
ii
-------
and response endpoints for risk assessments. The GIS platform will then be used to characterize
risk, by combining exposure models and relevant plant response data in a probabilistic framework.
As part of this regional analysis effort, plant responses relevant to ecosystems (e.g., species
composition, productivity) will be recorded as possible input parameters for wildlife habitat models
that are being constructed by the WED Terrestrial Habitat Project.
The Effects of Chemical Herbicides on Terrestrial Plants Research will develop
methodologies to test effects of chemical herbicides on individual terrestrial plant species and
communities. It will use outputs from the regional analysis research to identify crop and native plant
species, herbicides, and herbicide treatment (e.g., timing, concentrations) conditions important for
specific areas of the United States. Experimental protocols for terrestrial plant tests will be refined
for application to nontraditional species (e.g., perennials or woody species) and response endpoints
(e.g., seed yield or other reproductive or developmental parameters). Research will also be
conducted to develop molecular or cellular tools to extrapolate responses to non-tested species or to
verify field exposures. The plant test research will focus on low-dose, high-potency herbicides, but
also use some high-volume compounds. The protocols developed for chemical herbicide testing will
also be adapted for use with other industrial chemicals.
The Ecological Effects of Gene Flow from Transgenic Crops Research will develop
molecular methods to detect the presence of transgenic or other marker genes, evaluation of gene
flow from engineered to non-engineered plants, measurement of potential ecological effects of gene
flow on plant community structure and function, and definition of inputs for a prototype model for
gene flow. Experiments will be conducted at various scales, ranging from contained laboratory, to
growth chamber or greenhouse, to field. Current advances in genomics and proteomics will be
evaluated for their ability to identify potential adverse effects of gene flow in agronomic and non-
agronomic ecosystems.
Overall, this research project will provide tools to assist EPA in its regulatory role in
registration of chemical herbicides and genetically engineered crops that produce chemical
pesticides; thereby promoting sustained productivity of agricultural crops while maintaining the
ability of ecosystems to support wildlife and to carry out other essential services. The tools also
will aid post-registration monitoring to determine the success of registration restrictions in
iii
-------
protecting non-target crops or native plants. Furthermore, the will be useful for determining
ecological effects of other chemical pesticides and industrial chemicals.
IV
-------
Acknowledgments
The authors thank Don Hall, Karen Gundersen and Kay Wetz from the Senior Environmental
Employee program of the National Asian Pacific Center on Aging, for their assistance in preparing
this document.
v
-------
Table of Contents
EXECUTIVE SUMMARY I
ACKNOWLEDGMENTS V
LIST OF TABLES X
LIST OF FIGURES XI
LIST OF APPENDICES XII
1 PROJECT GOAL AND OBJECTIVES 1-1
2 GENERAL PROJECT OVERVIEW 2-1
2.1 Regulatory Authority and Responsibilities for Control of Chemical and
Biological Pesticides 2-1
2.2 Risk Assessment Needs: Exposure and Effects • 2-2
2.3 Research Plan Organization 2-5
3 REGIONAL ANALYSIS AND INTERPRETATION 3-1
3.1 Introduction 3-1
3.2 Objectives 3-2
3.3 Approach 3-4
A. Spatial Databases 3-4
B. GIS Platform and Analysis 3-5
C. Probabilistic Systems Modeling 3-6
vii
-------
3.4 Time Line 3-8
EFFECTS OF CHEMICAL HERBICIDES ON TERRESTRIAL
PLANTS 4-1
4.1 Introduction 4-1
A. Regulatory Basis 4-1
B. Scientific Rationale 4-3
4.2 Objectives 4-12
A. Improved process for selection of test specie. 4-13
B. Improved plant test guidelines for reproductive/developmental response
endpoints and nontraditional species 4-13
C. Input for ecosystem response tests 4-14
D. Develop mode of action studies / molecular biology tools 4-15
4.3 Approach 4-16
A. Plant species 4-16
B. Plant test guidelines 4-18
C. Future studies 4-23
4.4 Time line 4-27
EFFECTS OF GENE FLOW FROM TRANSGENIC CROPS 5-1
5.1 Introduction 5-1
A. Rationale 5-1
B. Background 5-4
viii
-------
5.2 Objectives 5-6
5.3 Approach 5-7
A. Develop molecular methods to assess gene flow . 5-8
B. Gene flow studies 5-9
C. Greenhouse/growth chamber/field studies to measure potential
ecological effects of gene flow. 5-10
D. Field studies of potential ecological effects of gene flow 5-12
E. Inputs for prototype model 5-13
F. Additional research to consider 5-14
G. Specific research proposed 5-15
5.4 Time Line 5-18
6 OUTPUTS AND PERFORMANCE MEASURES 6-1
7 PROJECT MANAGEMENT AND QUALITY ASSURANCE 7-1
7.1 Management Responsibilities 7-1
7.2 Communications 7-2
7.3 Quality Assurance (QA) 7-3
A. QA Responsibilities 7-4
B. Communications 7-5
C. Document Control 7-5
D. Special Health and Safety Considerations for Chemical Herbicide
Studies. 7-6
E. Special QA Considerations for Gene Flow Studies. 7-6
8 REFERENCES 8-1
IX
-------
List of Tables
Table 3.1 Time line for regional assessment and GIS platform. 3-10
Table 4.1 Examples of current criteria for performing vegetative vigor studies for terrestrial
plants and chemical herbicides. 4-29
Table 4.2 Summary of recent efforts to address research needs for tests of plant effects from
chemical pesticides, emphasizing WED contributions and concerns regarding new low-dose
high-toxicity herbicides. 4-30
Table 4.3 Examples of reports of non-seedling stage pesticide effects on non-target or simulated
non-target plants. 4-31
Table 4.4 Examples of limitations for current assessment indicators for non-target effects of
herbicides on plants. 4-32
Table 4.5 Examples of experiments on effects of chemicals on terrestrial plants 2002-2003. 4-33
Table 4.6 Time Line for Chemical Herbicides and Terrestrial Plants Research. 4-34
Table 5.1 Evaluation Criteria and Ranking of Northwest Crops/Traits of Interest. 5-19
Table 5.2 Timeline/Outputs Gene Flow Research. 5-20
Table 6.1 Annual Performance Measures and Goals for the Pesticide Research Project. 6-3
x
-------
1 Project Goal and Objectives
The Pesticides Research Project at the Western Ecology Division (WED) supports EPA
in its responsibility to prevent and reduce the impacts from pollution to communities and
ecosystems (Government Performance and Results Act (GPRA) Goal 4, "Safe Communities").
To accomplish this, we will meet critical research needs identified by two offices within EPA's
Office of Prevention, Pesticides and Toxic Substances (OPPTS), specifically the Office of
Pesticide Programs (OPP) and Office of Pollution Prevention and Toxics (OPPT), as necessary
to provide scientific support for ecological risk assessments for proposed or existing plant
pesticidal products. The four research needs as shown in the Project Critical Path (Figure 1.1)
are to:
1. assess ecological risks from transgenic crops,
2. improve methods for spatially explicit ecological risk assessments,
3. develop new methods to provide for efficient and effective gathering and
interpretation of herbicide hazard identification and dose-response data, and
4. investigate potential effects of high priority hazards.
Research to address these needs is important as there are limited methods or approaches
available for OPP and OPPT to assess ecological risks associated with movement and expression
of novel genetic material from genetically engineered crops. Furthermore, there is a need to
evaluate potential risks for herbicides and genetically engineered crops in spatially explicit and
probabilistic modeling frameworks that go beyond the more traditional deterministic framework
for risk assessments. Both OPP and OPPTs need targeted test development for improved hazard
identification and to obtain hazard dose and plant response data. Even though there are standard
methods available to OPP and OPPT to test the effects of most herbicides on crops, OPP and
OPPT need additional tests to evaluate herbicide effects on non-target, non-crop plants, and to
determine the effects of high-priority hazards to both non-target crops and non-crop plants such
as low-dose, high-potency herbicides.
1-1
-------
To meet OPP and OPPT's needs EPA's Office of Research and Development (ORD) has
prepared a multi-year plan (MYP) to address GPRA Goal 4. This plan identifies Long Term
Goals (LTG) for ORD research, and explains how research will be coordinated among ORDs
laboratories and centers. WED has developed this research plan to assist ORD in achieving two
LTG (Figure 1.1). We address LTG # 3, "To provide OPPTS with the scientific underpinnings
for guidance to prevent or reduce risks of human environments within communities, homes,
workplaces and to .assess risks of biotechnology to ecological systems," which will improve risk
assessments for transgenic crops (OPP and OPPT needs 1 and 2). We also address LTG 4 #, "To
provide OPPTS with strategic information and advice concerning novel or newly discovered
hazards," which will improve risk assessments for chemical herbicides and potentially other
chemicals (OPP and OPPT needs 2, 3 and 4).
WEDs Pesticides Research Project has developed this Research Plan "Effects of
Chemical Herbicides and Gene Flow on Non-Target Plants " which identifies three unique
research goals relating to terrestrial ecosystems which will address ORD's LTGs and OPP and
OPPTs ecological research needs (Figure 1.1). These goals are to:
1. determine the effects of gene flow from transgenic crops,
2 develop regional analysis and interpretation tools, and
3. determine effects of chemical herbicides on non-target crops and native plants.
For Goal 1, we will develop molecular biology and genetic methods to measure gene
flow and ecological consequences to determine risk in a spatially explicit landscape construct.
For Goal 2 we will improve models to determine ecological risks of herbicide use and gene
flow from transgeneic plants to other plants in a spatially explicit landscape. For Goal 3 we
will produce comprehensive and efficient in vivo assays to evaluate adverse effects of
chemical herbicides at critical plant life stages, and will develop new approaches for tier
testing, including methods for native plant species
WED Pesticides Project also will support EPAs GPRA Goal 4 to provide "Safe
Communities", not only by providing tools for initial ecological risk assessments for registration
1-2
-------
of chemical herbicides and transgenic crops, but also by providing tools useful in post-
registration monitoring. The pesticide and gene flow research will advance other research at
WED which supports GPRA Goal 8 "Sound Science," by increasing EPAs ability to assess,
improve, and restore the integrity and sustainability of ecosystems over time. At WED, the
Pesticides Project complements the Terrestrial Habitat Project, which is developing spatially
explicit models to evaluate the risks to wildlife populations resulting from changes in landscape
structure and habitat quality, including the use of herbicides.
1-3
-------
Intentionally Blank Page
-------
Pesticides Project Critical Path
WED
Research
Approach
Develop Molecular
Methods
Conduct Gene Flow
Studies
Conduct Greenhouse/
Growth Chamber/Field
Ecological Studies
Provide Inputs for —
Prototype Model
Hold International /
Workshop
Evaluate Spatial
Databases, GIS Platform
Develop Probabilistic
Systems Modeling
Melliods and
Protocols
Conduct case Studies
Develop Regionally
Based Species Selection
Process
Develop Tests for
Reproductive /
Developmental Endpoints
Conduct Greenhouse/
Field Studies
Design ecological, _
molecular effects studies
WED
Research
Objectives
Develop molecular
methods to asses gane
flow
Assess ecological
effects of exposure to
transgenic genes
Identify Uncertainty
and Knowledge Gaps.
Develop Spatially
Explicit Risk
Assessment Model
Provide web-based
tool for assessment
activities
. Improve plant selection
process-.-.; • i;I
Improve plant test
guidelines
Provide Input to Eco-
¦ system Response Tests
'Develop Mode of Action
WED
Research
Goals
> Determine
Ecological
Effects of
Gene Flow
from
Transgenic
Crops
Develop
Regional
Analysis and
Interpretation
Tools
ORD
Multi-Year
Plan
LTG 3 Develop
Scientific
Basis to
Reduce Risks
LTG 4 Develop
Information on
Novel New
Hazards-1
Determine
: Effects of
'• Chemical
Herbicides on
¦ Terrestrial
Plants «.•
Vr. l^,fe_lt
OPP, OPPT
Research
Needs
1. Assess
Risks From
Transgenic
Crops
2. Improve
Methods for
Spatially
Explicit Risk
Assessments
3, Develop
New Methods
I for Gathering
Herbicide
V- J -
r-
•. ;*/' Dose-
Response Data
4. Investigate
Potential
Effects of High
.Priority,
Figure 1-1 Critical path of Pesticides Project to meet research needs of Office of Pesticide Programs (OPP) and Office of Pollution Prevention and Toxics
(OPPT). It describes Office of Research and Development (ORD) multi-year plan goals, Western Ecology Division (WED) research goals, and the objectives and
approach for WED to address the goals and meet OPP and OPPT needs. The project has three components corresponding to WED goals.
1-5
-------
2 General Project Overview
2.1 Regulatory Authority and Responsibilities for Control of Chemical and Biological
Pesticides
EPA's mission to protect human health and to safeguard the natural environment — air,
water, and land — upon which life depends. EPA is responsible for protecting human health and
ecosystems, by enforcing government laws which limit the release of pollutants produced by human
activities into the environment. EPA's Office of Prevention, Pesticides, and Toxic Substances
(OPPTS) regulates chemical and biological pesticides under the authority of the Federal Insecticide,
Fungicide and Rodenticide Act (FEFRA), administered by the Office of Pesticide Programs (OPP).
Other chemicals and biological materials are regulated under the authority of the Toxic Substances
Control Act (TSCA), Federal Food, Drug and Cosmetic Act (FFDCA), Pollution Prevention Act
(PPA), and the Residential Lead-Based Paint Hazard Reduction Act administered by the Office of
Pollution Prevention and Toxics (OPPT). The OPPT also manages programs concerning new and
existing chemicals in the marketplace, asbestos, lead, PCB's, and other areas.
Within OPP, it is responsibility of the Ecological Fate and Effects Division (EFED) to
evaluate the potential ecological risks of pesticides (including those produced by genetically
engineered crops) during the product registration process, and the responsibility of the
Biopesticides and Pollution Prevention Division (BPPD) to evaluate ecological risks of
biological pesticides. Within OPPT, the Risk Assessment Division (RAD) evaluates the risks of
toxic substances.
In the past, OPP and OPPT had relatively fewer tools to protect ecological resources
compared with those available to protect human health. However, awareness of adverse effects
from drift of new low-dose high-toxicity herbicides to non-target crop and native vegetation has
heightened awareness of the need to improve tests for effects of chemical herbicides (and other
industrial chemicals) to plants. These tests would be useful not only in the pesticide or chemical
registration processes, but also for post -registration monitoring. At the same time, public
concern regarding the release of genetically engineered plants and the adoption of the "Final
Rules and Proposed Rules for Plant-Incorporated Protectants" (PIPs) (40CFR Parts 152 and
2-1
-------
174) have emphasized the need for new tools to evaluate the risks from chemical compounds and
genetically engineered plants in terrestrial ecosystems.
Development of the appropriate test methods and risk models for use by program offices
historically has been done by EPA's Office of Research and Development (ORD). The National
Health and Environmental Effects Research Laboratory (NHEERL) is ORD's focal point for
scientific research on the effects of chemical compounds and genetically engineered plants on
human health and ecosystems. NHEERL also is ORD's lead laboratory for development of a
multi-year plan under GPRA Goal 4 to formulate and conduct research to address key concerns
regarding chemical compounds and genetically engineered plants, including both ecological and
health effects research. Other components of ORD, such as the National Center for National
Exposure Research Laboratory (NERL), Environmental Assessment (NCEA) and National Risk
Management Research Laboratory (NRML) also contribute to the Goal 4 research in areas of
fate and transport of chemicals, environmental assessment and environmental remediation,
respectively. The Western Ecology Division (WED) in Corvallis is the component of NHEERL
charged with identifying and studying potential effects of chemical and gene flow from
engineered crops to other plant species, through its Pesticides Research Project.
2.2 Risk Assessment Needs
To determine potential environmental impacts of regulated pollutants (including
chemical and biological pesticides), EPA developed The Ecological Risk Assessment
Framework (US EPA, 1992,1998) (Figure 2.1). The framework has three main components:
Problem Formulation Phase. During this phase, EPA managers meet with interested
parties including risk assessors, risk managers, scientists, industry and the public to articulate the
problem, define the scope of the problem, and develop a plan to characterize and manage the
potential risk of the pollutant (or other environmental stress).
Analysis Phase. During this phase, aspects of the pollutant exposure and the resulting
effects on target organisms and ecosystems are evaluated. Both exposures and effects are
characterized quantitatively and the complex relationships between exposure and effects are
determined.
2-2
-------
Risk Characterization Phase. In this phase, the exposure and stressor response (effects)
profiles are integrated to estimate the risk from different levels of exposure. Models are used to
integrate exposure and effects data. The final product is a description of the likely risk along
with a description of key assumptions, scientific uncertainties, and strengths and limitations of
the analyses and characterization activities.
EPA's risk assessment process is iterative: as new data are acquired they are used to
strengthen the science to inform decisions, and as analysis and risk characterization occurs new
hypothesis and experiments become evident. The assessment results are communicated to risk
managers who develop a plan to manage the risk and communicate the results to interested
parties. A few examples of locations where the risk assessment approach was used to
understand potential impacts to ecosystems from a range of stressors such as sediments and
organic compounds include the Cordorus Creek Watershed in Pennsylvania (Obery and Landis,
2002), Darby Creek Watershed in Ohio (Cormier et al., 2000), Clinch River/Poplar Creek
System in Tennessee (Cook et al., 1999) and Fjord of Port Valdez in Alaska (Wiegers et al.,
1998).
There is a critical need for information with which to conduct similar risk assessments of
ecological effects from chemical pesticides and genetically engineered organisms (Fairbrother
and Kapustka, 2001; Taylor, 2001; Peterson et al., 2000; Landis et al., 2000). Through the
research described in this plan we will contribute information and develop tools to improve all
phases of the ecological risk assessment project (Figure 2.1):
Problem Formulation Phase. The project will develop tools for spatially explicit models
and methods (Geographic Information System or GIS-based) for determining which non-target
crops and native plants might be exposed to off-site drift of a proposed pesticide. The spatial
analysis will be in a regional context, so that species to be considered in the risk assessment are
pertinent to the geographic location where the crop/pesticide combination is likely to occur. It
also will develop the phenological relationships between timing of pesticide application(s) and
life-history patterns of non-target plants. These relationships will provide a rational basis for
requesting dose-response information on particular species and endpoints. Similar concepts
2-3
-------
apply to gene flow; i.e., geographic areas in which genetically engineered crops are likely to
outcross to compatible native plants need to be identified.
Analysis Phase. The project will determine whether existing plant test protocols are
suitable for non-target annual herbaceous and also for perennial or woody species; and non-
traditional endpoints such as tuber formation, fruit set, and yield also will be evaluated. Where
existing test protocols are inadequate, new protocols will be developed and tested to provide the
capacity to obtain dose-response information on regionally important non-target species. On a
broader level, chemical herbicide effects need to be extrapolated from the small number of crop
plants usually tested, to species that have not been tested. Given the large number of species in
the plant kingdom, it is not possible to test all directly. Species-to-species extrapolations can be
improved based on cellular and/or molecular mechanisms of action for representative species of
most plant groups. Similarly, few tools exist to measure, quantify and determine non-target
ecological effects of gene flow from crops to native plant species. We will conduct research to
assist in development of such tools to study ecological effects of gene flow.
Risk Characterization Phase. The spatially explicit, probabilistic modeling framework
also will be used to characterize risks to non-target plants from off-site drift of pesticides and
movement of genetically modified genes. This modeling framework will incorporate measures
of drift (e.g., AgDRIFT model for chemicals and new models for GM crops) with the dose-
response information developed in the Analysis Phase work. It will include stochasticity in
exposure parameters (including input variable such as wind speed) and variability in response
functions to develop probability bounds on risk outputs. Probability distribution functions will
be assigned to each source of uncertainty and variability in the data and knowledge bases in
order to fully assess the sensitivity of the output response to perturbations in the input data.
Similarly, potential movement of novel transgenes from genetically engineered crops to
adjacent crops or native vegetation depends upon close phylogenetic, geographic and
phenological relationships (e.g., timing of occurrence of developmental events during plant life-
cycle such as flowering) between crop and native plants. Gene flow can occur when pollen is
disseminated by wind or by pollinators to compatible recipient plant species. The GIS platform
will be useful for determination of co-location of related plant species and phenological
2-4
-------
relationships such as the timing of flowering of sympatric populations of weedy, native or crop
species that may be compatible with genetically engineered crops.
All of the tools developed in this project will help give risk managers the data they need
to determine the level of protection which they want to put in place for chemical herbicides or
genetically engineered plants.
2.3 Research Plan Organization
For each of the three WED Pesticide Research Project goals described in the Critical
Path, there is a corresponding component in this Research Plan: Regional Analysis and
Interpretation (Section 3.0), Effects of Chemical Herbicides on Terrestrial Plants (Section 4.0),
and Ecological Effects of Gene Flow from Transgenic Crops (Section 5.0). In each Section, we
provide an introduction including the regulatory and scientific rationale for the research,
objectives and scientific approach (Figure 1.1), and detailed time-line. The Critical Path and
specific details for each research component are intended to be a dynamic set of guiding
principles and not inflexible requirements for the direction of the Project. We will be in regular
contact with EPA staff and other scientists in the regional analysis, chemical pesticide and gene
flow research communities to further define the research objectives and approach. In addition,
the International Workshop in 2005 will not only aid in development of studies on ecological
effects of gene flow, but also be an opportunity to evaluate the progress and prospects for the
regional analysis and chemical herbicide research.
While each of the three research components in the plan is described independently, they
are intimately related. For example, the regional analysis and interpretation research provides
the basis for selecting species and exposure conditions to develop testable hypotheses for
research on effects of chemical herbicides on terrestrial plants. The regional analysis research
also provides information for selecting native and weedy compatible plant species which could
be affected by gene flow from genetically modified crops. The chemical herbicide and gene
flow research will contribute inputs for development of probabilistic risk assessment methods.
Applications of the molecular methods developed in the gene flow research maybe useful in the
development of new protocols for rapid screening of the sensitivity of a wide range of plant
2-5
-------
species to chemical herbicides and/or to be used as markers indicating that plants have been
affected by herbicides
In addition to the three current terrestrial ecosystem research components, an additional
area could be added to the Pesticides Research Project to address non-target effects of chemical
herbicides on plants in aquatic ecosystems. Any aquatic plant research would have to have 1)
specific objectives address OPP and/or OPPT research needs, 2) a well thought out series of
experiments using appropriate and well described methodology, and 3) defined outputs for use
by agency offices. Any plan for aquatic plant research would have to be peer reviewed, and
meet project management and QA requirements. An example of possible aquatic plant research
is shown in Appendix A.
Finally, specific outputs from the research will be the Annual Performance Measures
(APMs) which indicate the success of the project in meeting the EPA GPRA goals (Section 6.0
of this plan). Over time, these outputs will help provide the EPA with the broad outcome of
reducing non-target effects from chemical herbicides and gene flow to vegetation and improving
the health of ecosystems. In addition, a management plan is needed to assure that the project
follows the Critical Path and that the outputs produced by this project are reliable. Thus, Section
7.0 of this plan describes Project Management and Quality Assurance (QA) aspects of the
project, including the responsibilities of project participants, efforts to promote communications
within and to those outside the project, and QA requirements.
2-6
-------
Integrate Available Infonaatiea
Planning
(Risk Assessor/*
Interested Parties
Dialogue)
PROBLEM
FORMULATION
Assessment
Endpoints
Analysis
Plan
Measures
Ecological Response
Analysis
Exposure
Analysis
RISK
CHARACTERIZATIO N
Cornrnunrcmiion Results to the Risk Manger
Risk MsBgeneaf ami Camnnafeatiag
ResaMa to Interested Parties
Figure 2.1. Expanded risk assessment framework with expanded views of problem
formulation, analysis and risk characterization phases (U.S. EPA 1998). Rectangles
indicate inputs, hexagons indicate actions and circles represent outputs. This project will
make contributions to all components in the problem formulation and risk
characterizations phases, and the items to the right of the dashed line in the analysis phase.
2-7
-------
Intentionally Blank Page
2-8
-------
3 Regional Analysis and Interpretation
3.1 Introduction
Pesticide drift to unintended fields is inevitable and the magnitude of potential and the
effects to non-target plants are highly variable over time and space. The OPP is mandated under
FIFRA to evaluate the potential ecological risks of crop pesticide drift to non-target plants.
Specific research is needed to assess the potential impact of pesticide drift and to understand the
effects on individual plants and higher biological assemblages across the landscape. A
probabilistic systems modeling approach will be used to deal with variation in plant responses
(spatial, temporal), and to quantify uncertainty in modeled exposure/effect relationships for
individual plant species and communities. Probabilistic systems modeling offers the advantage
of incorporating the uncertainty and variability in the existing databases from multiple sources as
well as the uncertainty in the knowledge gap when no data are available. The existing databases
represent potentially large sources of variation because the data were not specifically designed
for regional risk assessment and, consequently, the data supports had low spatial and/or temporal
resolution. Though the basic probabilistic modeling approach used in this project will be
developed to assess the magnitude of the risk from chemical herbicides to non-target crops and
native species, it also will be applicable to evaluate risks from other classes of pesticides as well
as other chemicals. Aspects of the probabilistic modeling approach also will be applicable to
questions concerning impacts of gene flow from target crops to native species.
The rationale underlying the proposed research on herbicide drift effects is that non-
target crops and native plants in fields in close proximity to target crops are likely to be exposed,
and to respond to, chemical pesticides. The potential for exposure that can cause effects is
further determined by local wind conditions. In our regional analysis studies we will use
existing data and not conducting exposure research as such, i.e., we will not develop new ways
to determine, quantify or model exposure.
Herbicide drift generally occurs within 300 m of the crop field margin during and
shortly after application (SDTF, 1997b). Drift amount and location depends upon the
application rate and method (ground or aerial), environmental conditions, droplet spectrum,
3-1
-------
application height, and distance from the field boundary (Bird et al., 2002). Vegetation located
at the immediate boundaries of agricultural crop fields is at greatest risk from drift (Teske et al.,
2002). The magnitude of herbicide drift effects on plant productivity or community
composition depends upon the amount and type of herbicide and plant sensitivity at time of
exposure, which varies with plant species and its phenological stage.
3.2 Objectives
The overall goal of this component of the research plan is to develop tools to improve
regional analysis and interpretation aspects of ecological risk assessments (Figure 1.1). The
primary objective to address this goal is to provide a spatially explicit (using a GIS framework)
probabilistic risk assessment model to examine ecological risks associated with pesticide (i.e.,
herbicide in our studies) drift (Figure 1.1). A secondary objective is to examine the variability
and uncertainty in data on herbicide exposure and on the effects of exposure on non-target plant
species and population communities over time and space, and to identify the gaps in existing
knowledge. Uncertainties in modeling deposition of herbicide spray drift on unintended crop
and native plant species are distinctly different than those in modeling pesticide effects on non-
target vegetation than exist for target species. Different databases and models need to be
developed for estimating pesticide exposures and effects to non-target species. Another
objective is to provide a web-based tool for access and extraction of issue-specific databases and
maps for use by OPP and OPPT in their risk assessment activities. Though this research is on
effects of chemical herbicides, meeting our objectives will also provide a regionally-specific
framework with which to evaluate the effects of other pesticides, industrial chemicals and gene
flow from genetically modified crops to native terrestrial plant species
Depending upon availability of data sources and models, WED will develop:
• A spatial database of potential herbicide exposure to non-target plants: This will
require the linkage through a GIS platform of existing herbicide-specific data (e.g.,
toxicity, application rates and usage), climate (e.g., wind speed and direction,
temperature, relative humidity), general data (e.g., crop land cover, native vegetation
cover, soil type, hydrology, agricultural practices, field boundaries) with a spray drift
3-2
-------
model (Figure 3.1). The spatial database will be used to: 1) estimate potential crop
pesticide drift deposition to adjacent vegetation; 2) determine dominant crop, native
and endangered species at greatest risk from herbicide exposure for plant testing; and
3) select areas of highest risk of gene flow associated with wind-pollinated
genetically engineered crops which include PEPs. We will use existing data that are
national in scope. For particular case studies, more site specific data may be used if
available.
• A database of herbicide effects on plants: This will require the compilation of
scientific literature relating to the ecological effects of herbicides on non-target
terrestrial plants over wide geographic and taxonomic ranges, including stages of the
plant life-cycle that are not covered by standard phytotoxicity testing protocols.
Ecological effects may be direct, such as reduction in reproductive output or change
in plant community composition, structure or function. No similar information exists
for risks associated with gene flow. Non-target effects of low-dose, high-potency
herbicides (e.g., ALSase inhibitors) and broad-spectrum herbicides (i.e., glyphosate)
are a primary concern due to their increasing use and widespread distribution
(Maxwell and Weed, 2001). If feasible, manufacturer's databases on herbicide
effects will be obtained for our analysis.
• A database of crop planting dates, pesticide use dates and weed emergence dates:
This will reveal the time and location of greatest potential for substantial impact of
herbicide drift to non-target species. Due to the scarcity of herbicide use data,
scenarios for herbicide usage will be generated based on local knowledge of the
target crop and the presence of weeds.
• Several case studies: These will identify and prioritize potentially important
uncertainties, and identify the actions needed to address the gaps in the knowledge
base. Several regions will be selected for intensive study and development of the
probabilistic risk assessment of the impact of herbicide drift to individual species and
plant communities at the landscape level.
3-3
-------
• Web-based tool: This will provide on-line access to databases on herbicide exposure
and plant effects and generate issue-specific databases and maps for use by OPP and
OPPT in their risk assessment activities.
3.3 Approach
A. Spatial Databases
General aspects of the regional analysis and interpretation research approach are shown
in Figure 1.1. In terms of spatial databases, spatial information on land use, field boundaries and
ownership, crop and non-crop coverage, pesticide use, climate, soil, and hydrology will be
obtained and compiled in a GIS. For detailed, probabilistic analyses, determination of
agricultural field boundaries is especially critical for evaluation of vegetation at risk. Downwind
deposition decreases with distance from the edge of the field and approaches zero at 300 meters
in a typical aerial application (SDTF, 1997b). This defines the spatial resolution needed to
estimate pesticide exposure due to drift. Thus, databases for delimiting agricultural fields where
pesticides have been applied are the most important spatial data required to adequately assess the
potential drift impact on non-target species. The USDA National Agricultural Statistics Service
cropland data coverage based on Landsat thematic mapper (TM) scenes at 30 m2 resolution is
available for eight states (Arkansas, Illinois, Indiana, Iowa, Mississippi, Missouri, Nebraska and
North Dakota). This database will be critical in determining the spray drift loadings to
vegetation downwind for those states.
For California, cropland data for several counties at 30 m resolution are available from
the California Department of Water Resources. The California Department of Pesticide
Regulation (2000) has improved the resolution of the pesticide use data from 1 square-mile to an
actual field site for several counties (Neal, 2002). Field border databases and parcel boundaries
for reported agricultural field sites are currently available or are under development for counties
from the San Joaquin and Sacramento valleys, the coastal region, the Sierra foothills, and the
San Diego-Imperial area in California
Crop, native vegetation, pesticide use and effects data will be obtained through
collaborative efforts with federal, state and local agencies as well as industry as feasible.
3-4
-------
Existing databases including the USGS National Land Cover, GAP state land use coverage, the
California PUR and the USDA databases based on Landsat satellite imagery will be used to
identify the crops and native vegetation growing within and adjacent to the field boundaries,
especially for those states without detailed data. This will be used to identify the non-target
crops and native vegetation at risk from drift within a 300-m zone adjacent to the field
boundaries of specified target crop species. Pesticide deposition within the 300-m zone will be
determined based on pesticide use information reported to the field level as required by state law
in California or from state agricultural extension services for other states. Spatial data on wind
speed and direction and other factors will be used to refme the drift zone of herbicides on
unintended vegetation in later stages of model development.
B. GIS Platform and Analysis
Spatial information will be compiled in a GIS platform using ARCINFO (or other GIS
software). The GIS will be adapted and documented to provide maximum usefulness as a web-
based tool for OPP and OPPT staff and other interested individuals.
Figure 3.2 is a simplified example of how the GIS platform and databases might be used
to determine counties in the United States which are at risk for herbicide drift based on the
intensity and diversity of agriculture, the amount of herbicide usage, and wind speed data. It
includes representative steps for species selection and potential herbicide exposure. The
example used agricultural statistics (crop acreage per county, number of crops per county) were
from the 1997 Census of Agriculture (USDA-NASS, 1999). Herbicide use data were from the
National Center for Agricultural Policy (NCFAP), which has released summaries of agricultural
pesticide use for 1997. The example uses the total amount of all herbicides applied in a county.
Consideration of acres sprayed with all herbicides, or a particular herbicide, could be considered
in the future, when selecting species to be tested for specific purposes. Wind speed data were
from the National Climate Data Center (NCDC) at the National Oceanic and Atmospheric
Administration (NOAA). Hourly wind speed based on NCDC's TD-3280 and TD-3281
databases were obtained from Earthlnfo's Surface Airways database.
3-5
-------
The geographic areas with greatest risk for non-target herbicide effects as identified by
the type of analysis shown in Figure 3.2, will be candidates for more detailed probabilistic
analysis of herbicide impacts. This type of GIS-based analysis also will be used to determine
areas to be evaluated for selection of candidate crop and non-crop test species for plant tests to
determine the risks from herbicides to non-target vegetation on a regional basis in the US (see
Section 4.3 A below). Crop and non-crop-related information could be used in such an analysis
to determine areas of the US of interest for more intensive gene flow research (see Section 5.4
below).
C. Probabilistic Systems Modeling
The proposed probabilistic approach will define the probability and magnitude of the
risk, and uncertainty that spray drift effects will occur, on non-target species using a GIS-based
framework. Following the ECOFRAM approach (US EPA 1992, 1998), a tiered approach for
regional assessment will be used. A simple deterministic model will be developed first, followed
by a probabilistic systems approach to deal explicitly with spatial and temporal variation in
exposure, plant response and sensitivity, and uncertainty in input parameters. Level 1 is a
screening step and is based on existing data required to identify the regions at greatest risk to
pesticide drift. In case studies at WED over the next 2-3 years, published generic exposure-
response functions based on controlled experiments will be used to infer pesticide effects on crop
and non-crop species. If feasible, efficacy (dose-response) data also will be obtained from
herbicide manufacturers.
The risk assessment-related research is concerned with aerial applications of particular
pesticides because, in a typical ground hydraulic application, more than 99.9% of the applied
active ingredient stays on the field and less than 0.1% drifts, whereas about 8% of the applied
active ingredient drifts off field in a typical aerial application (Spray Drift Task Force, 1997a,b).
Scenarios of pray drift loading to non-target areas will be developed using AgDRIFT 2.0.05, a
spray drift model developed under a cooperative research agreement with EPA, USDA and the
Spray Drift Task Force (Teske et al., 2002; Bird et al., 2002). The AgDRIFT model will be used
to estimate the fraction of pesticide drift and downwind deposition based on climate information
and default settings for method of application. For each pesticide, information on the date,
3-6
-------
location of application, maximum application rate, and timing of application in relation to non-
target plant phenology will be used in conjunction with the AgDRJFT output to estimate
pesticide exposures to non-target areas.
An initial screening (Level 1) will be used to identify the pesticides, geographic locations
and non-target species of greatest concern as case studies to illustrate the probabilistic approach
for assessing risk. A "worst-case" scenario of potential non-target herbicide effects will be used
to identify those pesticides and regions of greatest concern for more intensive study.
Level 2 and Level 3 assessments will be stochastic in nature, using progressively finer
level data to identify areas with low, medium and high impacts from chemical herbicides. Level
2 assessments will refine the earlier calculations using more detailed GIS layers on land use,
pesticide use, field boundaries and climate as well as more specific crop profiles and exposure-
response functions. Level 2 estimates will still rely on point estimates for most input parameters
for estimating pesticide exposure and effect whenever published data are available. For other
parameters, expert judgment will be used to set the parameter values or establish hypothetical
probability distributions. A sensitivity analysis will be performed at this level to identify those
parameters that contribute the highest variability to the risk assessment. For Level 2
assessments, a spatial database of expected pesticide use will be developed assuming
recommended application rate, timing and method for each chemical and crop type. Any more
specific herbicide exposure-crop (or native plant) response functions available from the plant
testing component of this project (see Section 4.3,A,B) will be incorporated into these
assessments.
Level 3 assessments will use the best available data on the potential hazards of herbicides
to non-target plant species to address the uncertainty and variability in the impact of pesticide
drift on crop and non-crop species in adjoining fields. Level 3 will focus on exposure and effects
parameters identified as important contributors to risk in the Level 2 assessment, as well as on
specific case studies of the pesticides and species of greatest concern. This level will address
highly specific pesticide use scenarios and incorporate additional data to establish the temporal
and spatial pattern of exposure and effect on individual populations and communities, including
estimates of uncertainty in the data. Given the geographically focused nature of databases for
-------
location of plant species and identification of cropping practices, it is possible to estimate
pesticide deposition to unintended fields using a distribution of potential deposition rates. There
is considerably less information on the time of pesticide application in relation to the
developmental stage and sensitivity of the non-target species at the time of exposure and the
ensuing plant response for each plant species and chemical. Dose-response distributions (with
uncertainty bounds) will be retrieved from the literature, or PHYTOTOX database, or developed
by WED researchers for input to the model. Due to limited information of pesticide effects on
reproductive endpoints (fruit, seed, tuber production), it is likely that the uncertainty in pesticide
effect will contribute as much, or more, to the probabilistic impact of spray drift to non-target
crop and non-crop species.
3.4 Time Line
The research on regional assessment of pesticide drift effects on non-target plants has
three distinct phases: 1) development of a GIS database for estimating pesticide exposure to non-
target plants across time and space; 2) development of a phytotoxicity database for estimating
pesticide effects on crops and native vegetation; 3) synthesis and integration whereby
information on pesticide exposure and effects are used to develop a probabilistic risk assessment
model (Table 3.1). During FY2002-2003 the focus has been on development of the GIS
databases and procedures necessary for risk assessment, identifying crop and native plant species
suitable for tests on regional bases, and gene flow research. The GIS databases will be used for
initial screening to select crop and native species that are most likely to be exposed to pesticide
drift or gene flow.
The time line for development of the GIS platform is pursuant to the availability of high
spatial and temporal resolution data for crop use and native vegetation, pesticide use, and wind
speed and direction. Currently, cropland data coverage at 30 m2 resolution is available for eight
states and several counties in California. Over the next six years, high resolution crop land
coverage based on Landsat images are expected to become available for more counties in
California and other states. In California, all agricultural pesticide use must be reported to the
Department of Pesticide Regulation via the county agricultural commissioners; the reports must
include information on the date and location of application, kind and amount of pesticide, and
3-8
-------
type of commodity if applied to a crop. Oregon is the second state that has a law that requires
detailed reporting on pesticide use, similar to that for California, but uncertainties in funding and
the extent of pesticide use reporting have hampered progress towards implementing this
legislation.
From FY 2004 onward, the research will focus on the states for which crop land data
coverage is available at sufficient resolution to determine crop field boundaries in order to
estimate pesticide drift. The initial case study will be California where pesticide use data have
been used to identify the agricultural field boundaries and to provide detail needed to estimate
the amount and timing of pesticide drift to non-target species from aerial and ground
applications. Data from California will be used to determine whether pesticide exposures can be
generalized to other states where crop use data at 30 m2 resolution and county-level pesticide use
data are available. During FY2005-2007 research will extend the modeling efforts to infer
pesticide drift impacts on non-target species for states with low spatial and temporal resolution
data on crop use and pesticide use. Research is needed to understand how uncertainty of risk
predictions increases with decreased knowledge of field boundaries and pesticide use.
The research will link the pesticide exposure data with the plant effects database using a
probabilistic risk assessment model to estimate the potential impact.of pesticide drift on crop and
native plant species. The products of the five-year research plan are probabilistic risk
assessment tools for evaluating potential ecological risks from pesticide products based on data
at different spatial and temporal scales. These products will range from databases, to models, to
a web-based tool for on-line access to databases on herbicide exposure and plant effects for use
by OPP and OPPT in their risk assessment activities.
3-9
-------
Table 3.1 Time line for regional assessment and GIS platform
FY 2003 FY 2004 FY 2005 FY 2006 FY 2007
Level 1 - screening
Level 1 - screening
Level 2 - case studies
Level 3 - probabilistic
Level 3
¦ Obtain GIS databases
¦ Pesticide exposure -
¦ Refine pesticide
risk assessment
¦ Develop regional
¦ Crop coverage
estimate drift
exposure estimates-
¦ Refine pesticide
assessment tool
¦ Noncrop coverage
deposition to non-
more detailed info
exposure and
based on GIS
¦ Pesticide use
target areas
on land use,
effects estimates
framework and
* Climate
¦ Pesticide effect -
pesticide use, field
" Focus on
probabilistic risk
¦ Soil properties
estimate plant
boundaries, climate,
parameters
assessment
¦ Hydrology
response to
etc.
identified by
¦ Regional case
¦ AgDRlFT model
pesticide exposure
¦ Refine pesticide
sensitivity analysis
studies
¦ AGDISP model
on regional scale
effect estimates -
¦ Determine
¦ Field boundaries
¦ Select region for
crop-specific
probability
¦ Weed management
case studies based
exposure-response
distributions for
practices
on potential
functions, spatially
model inputs and
¦ Weed profiles
exposure and effect
explicit coverage of
parameters
¦ Crop profiles
target and non-
¦ GIS to develop
¦ Exposure-response
Contribute to APM on
target areas
high-resolution
functions
evaluation of risk
¦ Sensitivity analysis
maps of target and
¦ Literature review
assessment methods for
(iterative process
non-target species
herbicides
continuing in 2006
¦ Time of pesticide
Contribute to APM on
and 2007)
application in
Strategy for Updated Test
Contribute to APM on
relation to
Guidelines: Finalized
developmental stage
Research Plan
regional approach to risk
assessment
of non-target
species
3-10
-------
u
Climate
AgDRIFT Model
GIS Cell Processing
Pesticide
Exposure
Cropland
Database
Crop
Profiles
Soil
Properties
Exposure-
Response
Data
Noncrop
Land Cover
Database
Wind speed and
direction
Pesticide Use
Database
Probable Pesticide Effect on
Nontarget Species and
Communities
Figure 3-1 Databases for Probabilistic Pesticide Risk Assessment. The focus in this project will be
risks from herbicides.
3-11
-------
Intentionally Blank Page
3-12
-------
600 -
Mean =41
Standard deviation = 28
60
100
Number of Days
¦ o to is
I 17 to 32
D 33 » 48
~ 48 10 04
~ 85 K> 81
¦ 82 to 127
kilometers
Days With Avg Dally Wind Spaed > 10 MPH With > = 50% Agriculture,
Herbicide Rate > = 0.67 lbs per acre, and > = 12 Reported Crops
SHOMevn«t. j f. mm. iphmioi
Figure 3-2. Results from preliminary analysis of U.S. counties with vegetation at highest risk from
drift of agricultural herbicides. Counties depicted here in different colors had variable numbers of
days with wind speeds averaging >10 mph. These counties also had 250% of total acreage in
agriculture, 2t0.67 pounds/acre of herbicides applied to the agricultural acreage, and 2J2 different
crops grow.
3-13
-------
4 Effects of Chemical Herbicides on Terrestrial Plants
4.1 Introduction
A. Regulatory Basis
Since the end of World War II, American agriculture has become dependent on synthetic
chemical pesticides. According to the most recent EPA estimates (1997, EPA OPP website),
approximately 778 million pounds of conventional pesticide active ingredients (herbicides,
insecticides, fungicides) are used in the United States each year. Of this amount approximately 73%,
or 568 million pounds, are herbicides, 83% of which are used in agriculture. Greater than 90% of all
corn and soybean acreage in the U.S. are treated with one or more herbicides annually. Because
herbicides are designed to kill certain plant species (i.e., weeds), they have a high potential for
impacting individual non-target plants, plant communities, and function and structure of ecosystems
if they migrate off the intended use area. Depending on the mode of action and spectrum of pest
control, herbicides can cause visible damage to plants within hours, days, or weeks following
exposure. Furthermore, persistent herbicides can remain in plants, sediments and/or soil and affect
plants in subsequent growing seasons.
Chemical compounds are regulated under the Federal Insecticide, Fungicide and Rodenticide
Act (FIFRA) and the Toxic Substances Control Act (TSCA) by EPA's Office of Pesticides
Prevention and Toxic Substances (OPPTS) to protect human health and the environment. However,
in the past the protection of plant resources has received minimal attention under these regulations
(OPP Environmental Fate and Effects Division), and there have been questions whether non-target
plants have been affected by herbicides. Some states have restricted the use of certain pesticides
(e.g., 2,4-D) for the protection of non-target plants since the 1950s and 1960s. The registration in
1986 of clomazone for use on soybeans changed the regulatory picture because movement of
clomazone resulted in bleached non-target vegetation and started the first serious actions by OPP to
review plant test data. Incidences of clomazone impacts were followed by cases of off-target plant
damage from the low-dose, high-potency herbicides, rising public concerns that could not be
satisfactorily addressed with the scientific data available in the early 1990s. These concerns
suggested that plant resources outside the bounds of treated areas are at risk from the movement of
4-1
-------
herbicides from targeted lands. At potential risk are non-targeted crops, rare and endangered plant
species, native plant communities and organisms that are dependent on natural plant communities
for food and shelter.
EPA required tests for plant effects are the Pesticide Assessment Guidelines Subdivision J,
which have been refined as the Series 850, Ecological Effects Test Guidelines. Guideline 850.4000
provides general information on conducting plant tests with pesticides and industrial chemicals for
both OPP and OPPT. In addition there are OPP-specific methods (850.4025 Target Area
Phytotoxicity, 850.4100 Seedling Emergence, 850.4150 Vegetative Vigor, and 850.4300 Terrestrial
Plants Field Study), and OPPT-specific methods (850.4230 Early Seedling Growth Toxicity Test,
850.4600 Rhizobium-Legame Toxicity, and 4800 Plant Uptake and Translocation). These tests are
usually required in a "Tier" sequence, i.e., for "Tier I" or maximum challenge tests, a single
concentration of a pesticide is required to determine the general phytotoxicity of a chemical. In
"Tier II" tests, multiple concentrations of a pesticide are required to establish pesticide dose-
response functions when the chemical is known to have phytotoxic effects. The Tier II chemical
dose (concentration)-plant response data are used to establish EC25 (effective pesticide concentration
for a 25% reduction in plant response) values for the different species. In a "Tier III" test (OPPTS
850.4300) plants are grown under conditions similar to those where they would be exposed to a
pesticide in the field. The field test is a long-term test where plants are grown and evaluated for
pesticide responses over their entire life cycle. This enables determination of reproductive, biomass,
richness of species, population density, and other parameters as indicated at time by Agency. Tier
III studies are requested on case by case basis with protocols determined for a particular situation of
decreasing risk uncertainty to non-target plants. If there is no standard EPA test for a particular
pesticide or chemical application, other standard plant tests can be used, such as those approved by
ASTM International or the Organization for Economic Co-operation and Development (OECD).
Table 4.1 summarizes the general protocol for growing conditions for EPA's Tier I and II
vegetative vigor test and other compatible tests of other agencies and organizations for plant effects
from pesticides (Stavely 2002a). In brief, plants are grown under standardized, controlled-
environment climatic conditions (or possibly small field plots). The conditions (e.g., 25/20 C
day/night temperature, 350 |omol/m2/s light intensity in wavelengths between 400-700 nm) can
produce generally healthy vegetative plant growth in most greenhouses and growth chambers. A
4-2
-------
natural mineral soil with <3% organic matter is used with bottom watering preferred. Only a small
number (3) of the experimental units (pots) are required for the test, with data averaged for several
plants (10) per pot. Pesticides are applied either to the leaves or incorporated into the soil. Ten
species of plants normally are suggested for the OPP Tier I vegetative vigor test. Corn and soybean
are required, and a root crop (often carrot), plus tomato, cucumber, lettuce, cabbage, oat, ryegrass
and onion usually are used. Plant response measurement endpoints generally are injury, height and
shoot dry biomass. The vegetative vigor test involves pesticide application at an early growth stage
(14 days after seedling emergence) and harvest for analysis a short time later (generally 7 to 14
days).
However, since the establishment of the OPP Subdivision J test protocols in the 1980's and
OPPTS Series 850 tests in the 1990's, EPA has sought to improve and provide scientific justification
for these tests to better evaluate the toxicity of chemicals to terrestrial plants and plant communities.
Similar efforts have occurred in Canada (Boutin and Rogers, 2000), and in OECD countries.
Recently, there has been a movement to harmonize all regulatory test protocols including plant tests
across agencies and countries. Table 4.2 summaries a series of meetings and events over the past 12
years which have helped to crystallize the need for a new program focusing on revising tests for
plant effects from chemical pesticides. Recommendations regarding those tests have been stated in
the report from the 1991 workshop on non-target plant testing (Fletcher and Ratsch, 1991), expanded
upon by subsequent meetings and documents, culminating in findings from the International
Workshop on Plant Tests held in Alexandria, Virginia in January 2002. Some important findings
from the Alexandria meeting are summarized as follows (Stavely, 2002b):
"Some of the most important research needs were: tests for terrestrial plant
development and reproduction; ...field tests, including multi-species approaches;
monitoring tools; ...research on alternative test species, on relative sensitivity, and
• on approaches to selecting test species; methods to evaluate recovery; and
greenhouse-to-field extrapolations."
B. Scientific Rationale
4-3
-------
Non-target Plant Test Protocols. Plant research addressing ecological effects of herbicides
traditionally has fallen under two headings: target and non-target. Target herbicide research is
confined to studying the influence of herbicides on approximately 100 crop and weed species
(Fletcher, 1997) that are intended recipients of particular herbicides applied within the borders of
cultivated fields. Non-target herbicide research deals with the influence of herbicides on all plant
species growing outside the borders of fields treated with herbicides.
The goals of target plant herbicide research are to identify suitable herbicides, application
concentrations, and methods which will eradicate weeds in cultivated fields without damaging
and/or reducing the yield of crop plants of interest. These herbicides must not have human health
effects, must be compatible with other cultural practices for crops of interest, and must be
economically viable to develop, produce and market.
Target plant herbicide research has been primarily interested in managing the development of
herbicide resistance in weeds because resistance reduces the usefulness of the specific herbicide for
future use. Unfortunately, this research is not useful for addressing the risks of herbicides to non-
target plants, risks that include species and exposure conditions (herbicide concentrations, time
during plant life cycle, environmental condition) not typical of target plants.
The existing Tier I, II and III test protocols used to determine non-target plant effects were
established using the best available consensus scientific information at the time, but do not reflect
subsequent methodological questions or advances in scientific methodology. For example, concerns
have arisen regarding both aquatic and terrestrial test species used by EPA to collect preregistration
data for protecting plants and other photosynthetic organisms. EPA-required tests for terrestrial
plants currently use ten angiosperm species, six dicotyledonous species from four families including
soybean and a root crop, and four monocotyledonous species from at least two families including
corn. All other species are only recommended, and substitution of other test species (other crops,
weeds controlled native plants, perennials, woody species) especially is encouraged when species
sensitivity to the test compound is known ahead of time. However, in practice, all ten species are
annual agricultural species as indicated earlier.
4-4
-------
In any event, the ten species are considered to be surrogates for all potential non-target native
plant species (16,000 in the U.S., Heywood, 1978), in addition to non-target crops. However, the
narrow taxonomic and life form range of the ten required test species raises the question of whether
the majority of plant species will be protected, including all the rest of the angiosperms (both
herbaceous and woody), gymnosperms, and ferns. Species also vary greatly by ecological regions
and taxonomy yet there is no provision for including this diversity in non-target plant risk
assessments.
In addition to species concerns, the time frame for current tests may not be appropriate for
non-target plants. The Tier I and II vegetative vigor tests last a maximum of 28 days. This time
period is insufficient to capture the reproductive phase of the plant's life cycle. Not only is this
important for the individual plant's ability to pass along its traits, but reproductive yield is one of the
most important economic aspects of agriculture. Furthermore, many wildlife species depend upon
seed production of noncrop plants for their food source. The limited data available suggest that
exposure during the vegetative versus reproductive phases may not have equal influences on
reproduction and crop yield (Fletcher et al., 1996). Current Tier I and Tier II tests include only
measurements of injury, height and biomass that do not correlate well with yield. There is some
argument that early seedling growth parameters are protective of reduction in yield responses, but
this may lead to over regulation of some herbicides. Table 4.4 indicates the limitations of different
current and possible assessment indicators for non-target effects of herbicides (Maxwell and Weed,
2001; Obrigawitch et al., 1998).
Additional uncertainty about current non-target tests concern the extrapolation of greenhouse
tests to field conditions where plants exist in complex relationships with other organisms, typically
are in competition for water, nutrients, space and light, and where they are threatened by herbivores
and pathogens. Fletcher et al. (1990) compared results from studies included in the PHYTOTOX
(Fletcher et al., 1985) database that were conducted on similar plant species with the same chemical
in both greenhouse and field. Out of 20 combinations, 55% showed plants in the field to be more
sensitive, 30% were more sensitive under greenhouse conditions, and 15% were equal in sensitivity
in the field and greenhouse. Sensitivity differences were less than 2-fold in all cases. In contrast,
McKelvey et al. (2002) reported that the crop species currently used in vegetative vigor tests usually
were more sensitive than non-crop plants tested. However, the results from McKelvey et al. (2002)
4-5
-------
only considered "weed" species (primarily annuals) as non-target plants and only used high (several
x the field application rate) herbicide concentrations.
Current plant tests do not fully address risk assessment needs, as they were developed before
the risk assessment paradigm came into common use, and before development of GIS technology for
spatial and temporal analysis of data sets. EPA's Science Advisory Panel (SAP, 2001) endorsed an
approach to incorporate both these technologies into the risk assessment process for herbicides.
Specifically, an initial problem formulation process should be conducted to determine the plant
species growing in proposed areas of herbicide use, and their stage of development at time(s) of
application. This information provides a rational basis for selection of species and endpoints to test
in the Tier I and II assessments. GIS systems also can be used to collate post-registration incident
monitoring data, which will provide further information about types of plants and adverse effects
most frequently associated with herbicide use.
ALS Inhibitors. Much of the recent interest in herbicide testing has been associated with
low dose, high potency herbicides such as acetolatacte synthase (ALSase) inhibitor herbicides which
where initially introduced into U.S. agriculture in the mid '80s (Fairbrother and Kapustka, 2001).
Use of these chemicals potentially addressed major environmental concerns regarding herbicide
toxicity in that they have a relatively narrow spectrum of susceptible organisms, are relatively short-
lived in the environment, nonbioaccumulative and used in low volume. The first class of these
herbicides used was the sulfonylureas (SU). They were quickly followed by the imidazolinones and
more recently by the triazolopyrimidine sulfonanilides and pyrimidinyl oxybenzoates. The primary
mode of action of these herbicides is by disruption of the synthesis of the branched chain amino
acids leucine, isoleucine and valine. However, there may be secondary modes of action within
plants leading to the accumulation of toxic metabolites, disruption of assimilate transport and
inhibition of reproduction (Fairbrother and Kapustka, 2001; Taylor, 2001). These herbicides are
generally not considered to be toxic to animal systems due to animals' inability to synthesize
branched chain amino acids. However, the ALS inhibitors can affect bacteria and fungi which play
key ecological roles in nutrient cycling, soil fertility, and plant nutrition and health.
One of the most striking features of the ALSase herbicides is their exceptionally low field
application rates (g Ha"1 or oz Ac"1). Such low rates make chemical detection of these herbicides on
4-6
-------
plant material impossible with present technology. Thus, millions of pounds (est. 2.1 million pounds
in 1996, Fairbrother and Kapustka, 2001) of these toxic materials are released into the environment
without practical means of tracking their fate and influence in the environment. However, from past
experience with other pesticides, from information presented at a recent workshop on low-dosage-
herbicides in December, 1999 (Ferenc, 2001), and based on findings from the FIFRA Scientific
Advisory Panel Meeting in June, 2001, there is general agreement that they are moving off site in
water, on soil particles and as spray drift. For example, the use of Oust® (sulfometuron) on
rangeland may have resulted in damage to potatoes at least 8 km distant (personal communication,
Dr. Pamela Hutchinson, University of Idaho, Aberdeen Research and Extension Center).
Maxwell and Weed (2001), Obrigawitch et al., (1998) and Ratsch and Fletcher (1991)
summarized recent reports of non-target impacts from herbicides, many of which were ALSase
herbicides. Obrigawitch et al. (1998) specifically assessed the effects of sulfonylurea herbicides
using field approaches. They concluded that the risks to non-target plants from sulfonylureas were
similar to those from other herbicides used at higher application rates. Obrigawitch et al. (1998) also
stressed the need for standardized protocols to assess the effects of herbicides, in general, on non-
target plants. The review by Ratsch and Fletcher (1991) and other papers in the report by Fletcher
and Ratsch (1991) indicated pesticide effects occurred at various stages of plant development,
including reproduction.
Based on the reviews by Maxwell and Weed (2001), Obrigawitch et al., (1998) and Ratsch
and Fletcher (1991), and additional literature; Table 4.3 summarizes some of the reproductive or
other developmental effects of herbicides, especially ALSase inhibitors. These effects were found
primarily in controlled herbicide exposures. There is additional literature on non-target effects of
herbicides on leaf injury (e.g., AL-Khatib et al., 1993). However, the relationship between leaf
injury and reproductive effects is not clear. Some of these studies showed reproductive effects at
herbicide concentrations that did not produce visible leaf injury (Fletcher et al., 1996; Fletcher et al.,
1993). In contrast, other studies showed that while reproductive effects from herbicides were always
associated with leaf injury (Al-Khatib and Peterson, 1999), the amount of leaf injury was directly
related to the amount of yield loss. The nature of the injury-yield loss relationship may depend on
when plants come in contact with the herbicide during their life cycle (Fletcher et al., 1996).
Especially important may be periods when rapidly growing sinks such as seeds or tubers are at risk
4-7
-------
due to the effects of ALSase inhibitors on cell division, meristematic activity, phloem loading and
photosynthate transport (Taylor et al., 2001).
Overall, there are a number of reasons why low-dose, high-potency herbicides such as the
ALSase inhibitors may be of greater interest than conventional herbicides. As summarized by
Maxwell and Weed (2001), compared with conventional herbicides, the low-dose, high-potency
herbicides may: 1) have increased total amount applied on crops if current use trends continue, 2)
have greater aerial drift because of their application methods, 3) be used more at the boundaries of
agricultural regions as they become used more for roadside maintenance, 4) be used more to
suppress forest understory plants in forest ecosystems, and 5) have more potential for reproductive
effects due to exposure amounts and timing for different crops. In his preliminary ecological risk
assessment and characterization for ALSase inhibitors, Taylor (2001) concluded that "...the
uncertainty of the data is significant in terms of breadth and depth..." He recommended areas of
needed research to provide peer-reviewed literature addressing the uncertainties including...
".. .analytical methodologies to quantify the concentration of low-dose, high
potency herbicides in multiple media..." (including biosphere, e.g. plants)
".. .effects.. .on plant structure and function..."
"...role of temperature, light, precipitation, pH, etc....with respect to
behavior as well as effects on non-target species."
"The development of a system-level model to predict the behavior of low-
dose, high-potency herbicides is needed, with a special interest in simulating
exposure of at-risk processes in the biosphere."
Ecosystem Responses to Herbicides. The movement of herbicides from targeted land has the
potential to adversely affect both agricultural and natural ecosystems. The ecological effects may
be direct, such as the elimination or reduced reproductive output of certain plant species in a
community, which leads to the alteration of the community's species composition, structure and
function. Effects may also be indirect, such as changes in microbial communities, controlling plant
pathogens, or diminished insect populations causing wildlife populations to increase. These changes
4-8
-------
can lead to numerous negative impacts on wildlife habitat, nutrient cycling, control of soil erosion,
recreation, timber or pulp production, livestock grazing, control of noxious plant species, and
aesthetics (Obrigawitch et al., 1998). In 1991 the SAP recommended that:
"Community response measures need to be developed that have the potential
to identify significant structure and functional changes in exposed
communities. Many more invaded natural communities will be targeted for
herbicide use, given the increased recognition of invasive plant problems.
Community response metrics are available in the ecological literature.
However, the specific value of these responses with regard to characterizing
responses due to chemical exposures need to be determined and possible
modifications of designs identified."
Direct effects. Observational data suggests that plant assemblages at field margins
experienced substantial change in species frequency and distribution due to differential susceptibility
to herbicides (Kleijn and Snoeijing, 1997). Others (Jobin et al., 1997) have found lower species
diversity in the herbaceous layers of hedgerows and woodland edges of cultivated fields with a
history of herbicide use as compared with those near fields without herbicide use. In controlled
experiments with plant communities, Pfleeger and Zobel (1995) demonstrated that variable species
responses to herbicide exposure alter the competitive interactions within a community.
The high selectivity of the low dose, high potency herbicides could accentuate the
differential stresses and subsequent shifts in dominance in a plant community. Such shifts in a
community can result in changes in frequency and production and even extinction of desired species
(Tillman, 1988). In addition, Boutin and Jobin (1998) demonstrated that herbicides can contribute to
shifts in plant communities adjacent to intensively cropped fields from native species toward more
weedy species, and, thus, these adjacent communities can promote the spread of weed species.
Additionally, crops are being genetically engineered to be tolerant to the highly active herbicides,
which will stimulate more widespread use and subsequent potential for non-target effects (Maxwell
and Weed, 2001).
4-9
-------
Besides undesirable changes in plant communities, threatened and endangered plant species
are at risk. The federal government has listed over 500 plant species and the Nature Conservancy
considers 5000 of the 16,000 native species in the U.S. to be at risk. Almost 50% of these species
are annuals that are dependant on seed production or the seed bank for survival. The highest
percentage of these plants is located in Southeast wetlands and Southwest deserts.
Because the data available are very limited and highly speculative, there is a need for
controlled experiments in different regions of the country to determine the effects of herbicide
exposure on plant community dynamics. Considerable work will be required to identify an effective
methodology for determining meaningful endpoints.
Indirect effects. The direct effects of herbicide drift to plants and plant communities is
straightforward compared with the complexity of food web dynamics and habitat alteration effects
on wildlife populations. The vast majority of the reproductive output of plants is used by animal
species as sources of energy. Therefore, changes in plant community dynamics will affect wildlife
populations. For example, mammal populations in eastern deciduous forests are controlled by the
abundance of acoms, years with high acom production mouse populations increase (Ostfeld et al.,
1996). Increased mouse populations lead to more effective gypsy moth control through increased
mouse predation of moth larvae. Abundant acorn crops also provide deer with sufficient food, and
as a result, they browse less on tree samplings. More saplings grow into the tree canopy, thus
determining future forest composition. In another example, the reproductive output of the tree
Casearia corymbosa in Costa Rica is responsible for the survival of at least seven bird species
during the dry season (Howe, 1977). Many granivorous invertebrates depend on the reproductive
output of plants for survival and do not have the luxury that most birds have of moving to a new
habitat when food resources are scarce. Besides performing ecological functions, many
phytophageous invertebrates are the basis of many webs that support vertebrate species that are
popular with the public (Greig-Smith, 1991).
Changes in habitat quality caused by herbicides can affect wildlife populations. For
example, populations of the gray partridge in the United Kingdom have been affected by herbicide
use (Greig-Smith, 1991). Plant species composition within hedgerows between herbicide sprayed
fields was altered, resulting in a 50% loss in populations of arthropods, which were a high-protein
4-10
-------
food source for partridge chicks. Fewer arthropods resulted in more frequent partridge brood
movements leading to greater predation of chicks. Other studies also have shown effects of
agricultural pesticides on wildlife (Freemark and Boutin, 1995, 1994; O'Conner, 1992; Mineau et
al., 1987; Sheehan et al., 1987; Hill, 1985).
Various effects of herbicides on metabolic activities and on overall growth of a host and/or a
pathogen can cause an increase in soil borne diseases, resulting in greater pathogen damage to plants
from a variety of organisms including fungi and nematodes (Altman and Rovira, 1989). Though
there is limited understanding of effects of herbicides on plant pathogens, even less is known about
mycorrhizal associations (Altman and Campbell, 1977). Most plant species require some form of
symbiotic relationship with mycorrhizal fungi, and herbicide effects to fungi may have a significant
effect on plant health and possibly on ecosystem structure.
Herbicide application can lead to changes in insect herbivory. Increases in herbivory have
been attributed to higher concentrations of nitrogenous compounds including amino acids and
proteins in exposed plants (Chaboussou 1986). Stanley and Hardy (1984) suggested that bare land
after nonselective herbicide treatment provides a suitable environment for invasive species of plants
and insects to colonize. Bare soil is receptive to numerous and widely distributed seeds from weed
plants and the monoculture crop is an easy target for plant feeding insects such as aphids. Invading
insects are preyed upon by similarly invasive predatory species of insects such as ants. The non-crop
plant species and invasive insects have several features in common including rapid multiplication
and dispersal mechanisms that allow rapid colonization. Herbicide use in and around cultivated
fields has also resulted in a decline in the abundance of certain plant species, which in turn has
resulted in the decline of certain insect populations (Hume 1987).
Freemark and Boutin (1994) summarized the impacts of herbicides on biotic communities
by stating:
"Different taxonomic groups.. .play important roles in agroecoystems in soil
fertilization and aeration, the recycling of organic material and nutrients and
the degradation of contaminants... .The use of agricultural herbicides (and
other pesticides) can interfere with these functions by altering plant
4-11
-------
biochemistry, developmental processes and morphology, changing
population dynamics, species composition and diversity, interrupting energy
and nutrient flows, degrading water quality and changing the composition,
heterogeneity and interspersion of habitats for wildlife."
4.2 Objectives
The overall goal of WED's chemical pesticide research is to assist OPP and OPPT in
development of methodologies to determine the effects of chemical herbicides on non-target
terrestrial crops and native plants ecological risk assessments (Figure 1.1). Current plant tests,
while adequate for determination of basic toxicity of pesticides or other chemicals to plants (i.e.,
by measuring occurrence of death for young plants) cannot supply OPP and OPPT with the
necessary information needed for ecological risk assessments. Test species currently used
generally are not the non-target crop species at risk from herbicide exposure, but rather are
common agricultural species. In terms of regulatory and scientific needs, recommendations for
research made at various meetings over the past 12 years (Table 4.2) illustrate the breadth of new
information needed for improved plant tests. Tests are needed at a range of scales of responses
from molecular, to individual plant seedling life-cycle responses, to multispecies ecological
responses. Despite these needs, there presently is very limited methodology available to
determine the risks of herbicides to terrestrial plants, both native and cultivated, plant
communities, and the organisms associated with those plants. Methodology must be developed
to allow realistic ecological risk assessments to be made prior to the registration of new
chemicals or reregistration of existing chemicals
The types of tests needed for risk assessments range from specific, designed for assessing the
risk of particular chemicals on particular plants, to general designed to supplement the information
required for vegetative vigor and ecosystem response tests in the current tiered approach. There are
four specific objectives for research at WED to improve plant tests to evaluate the effects of
chemical herbicides (and potentially other chemicals): (A) improve the process for selection of test
species, (B) improve species test guidelines, (C) provide input for ecosystem response tests, and (D)
develop mode of action studies / molecular biology tools (Figures 1.1,4.1). In Figure 4.1, the two
most important immediate objectives, (A) and (B), are indicated in bold; while objectives (C) and
4-12
-------
(D), though very important over the longer term, are indicated with dashed lines to indicate that they
are highly dependent on resources in the future.
A. Improved process for selection of test species
Improving the process for selection of test species is our highest priority because of questions
regarding the scientific justification for test species selected by EPA for use in collecting registration
and re-registration data for protecting green plants (both terrestrial and aquatic test species), (A in
Figure 4.1). The development of an improved GIS-based protocol to select plant species for plant
testing purposes is central to all other aspects of our plant test research. A GIS-based methodology
is being developed as part of the regional analysis and interpretation research described in Section
3.0 above. It will be used to select a range of crop and native plant species for improved vegetative
vigor tests to provide the initial screening information indicating whether or not a chemical is toxic
to terrestrial plants (current Tier I and II tests). The GIS would be used to select species from
regions of greatest herbicide use, and, thus where potential non-target drift problems from the
chemical would be expected to be greatest. A larger range of species would allow consideration of
diversity in possible plant responses. Inclusion of information on species phenology would allow
EPA to better suggest which species should be subject to reproductive or developmental tests as well
(current Tier III test). The methodology would also be used to select species assemblages for
ecological tests and species of interest for molecular tests.
B. Improved plant test guidelines
Our second-highest priority is development of improved test guidelines that will include
endpoints reflecting the entire life-cycle of plants (i.e., developmental responses especially
reproduction), and include nontraditional test species (B in Figure 4.1). For life-cycle endpoints,
there are limitations in terms of relating currently used growth assessment endpoints, such as shoot
height and shoot dry weight, to crop yield (Table 4.4) (Maxwell and Weed, 2001; Obrigawitch et al.,
1998). Early growth responses measured two weeks after exposure to an herbicide or chemical may
not correlate well with responses when plants are exposed to herbicides at critical developmental
stages later in their life-cycle (e.g., at flowering or fruit set). Figure 4.2 illustrates different times of
possible herbicide application compared with the life cycle of a plant. Life-cycle responses
4-13
-------
especially may be important for perennial crops and native plants. For native plants, life-cycle
responses such as reproduction, in terms of seed production, will be critical for determining
ecosystem-level responses to chemicals such as changes in plant interspecies competitiveness or
changes in suitability as wildlife habitat.
Both traditional seedling-oriented (e.g. vegetative vigor) tests and proposed life-cycle
(reproductive/developmental) tests also will be evaluated to verify applicability to nonstandard test
species. These may include biennial or perennial crops (herbaceous or woody), native plants (both
annual and perennial) and specific threatened or endangered species. The test methodology
considered will include standardized cultural procedures for plants and specifics on herbicide
application procedures. In terms of cultural procedures, current plant tests are conducted under a
minimal set of climatic conditions geared towards production of uniform, vegetative, seedling plants.
These likely are not adequate to provide the resources (e.g., soil volume, fertilizer, water, space per
pot per plant) or conditions (e.g., photoperiod, air temperature) to carry different species through
their full life cycles as required for determination of reproductive and developmental responses.
Current plant growth protocols also likely are not adequate for native plants growing under a variety
of different conditions. For example, there as been little emphasis in the protocols in terms of
growing media other than the requirement for a sandy loam soil. Soil types and soil properties (pH,
% organic matter, and cation exchange capacity) differ widely, and affect both plant health and
herbicide chemistry. In terms of herbicide exposures, the exposures protocol must represent field
application methods, concentrations and timing to realistically assess the impacts of the herbicide on
the endpoints of greatest interest.
C. Input for ecosystem response tests
The primary focus of the plant test species selection (A) and test guideline (B) objectives
above is to provide information to improve the plant testing protocols where single species are
grown independently, either in pots or small plots. However, in reality, ecological risk assessments
need ecosystem-based tests to determine the effects of chemicals on multi-species ecosystems and
not just individual species. Thus, we will provide input to develop protocols for ecosystem
response tests (C in Figure 4.1). The protocols will be based on information from the plant species
and response endpoint objectives, and they will take into account the modeling needs for risk
4-14
-------
assessments. Such tests will provide guidelines more realistic than those in the current Tier III Field
Effects test. The tests will be field-based and include the measures of plant productivity and
community structure that are vital for defining quality of wildlife habitat and the persistence of plant
communities. An important emphasis of any ecosystem response protocols is the provision for links
to wildlife population models, because such models are being developed in other projects such as the
WED Terrestrial Habitat Project.
D. Develop mode of action studies and molecular biology tools.
For the long term, research on herbicide modes of action and molecular biology-level
responses to herbicides is a critical objective for the plant effects studies (D in Figure 4.1). Research
into the application of genomics and proteomics for the prediction of the effects of herbicides on
plants was strongly recommended by the SAP in 2001.
Both the current and proposed plant and ecosystem tests (A-C above) focus on whole plant
growth and biomass or yield responses of plant species. Such tests are based on known mechanisms
of action of herbicides. These tests require growth of whole plants for full life cycles, which takes
considerable time, space, and money. If the physiological, biochemical mechanisms, and molecular
modes of action were better known for some herbicides, especially in terms of the molecular basis
for how herbicides cause reproductive effects, extrapolation and prediction of responses across
species may be facilitated. If so, studies with multiple plant species for long periods of time would
not be required. Thus, molecular biology offers the promise to develop tests to screen plants
susceptibility to herbicides in vitro to possibly supplement or replace whole-plant tests in some
cases.
Molecular biology tools may also have the potential to the effects of detect low doses of
herbicides in field ecosystem studies. Development of molecular biology tools based on chemical
herbicide effects needs would be intimately related to the development of molecular biology tools
for gene flow studies (genomics and proteomics) as they would use similar molecular biology
laboratory methodology, equipment and scientific expertise (See Figure 5-5).
4-15
-------
4.3 Approach
Research to address objectives 4.2.1 through 4.2.2 will be through controlled experiments
based on the issues and recommendations from scientists and agency policy makers (Table 4.2).
Every effort will be made to design the experiments to gain the information needed to meet several
objectives with the same study (e.g., by including new possible non-target species, reproductive
responses, and different environmental conditions). The research conducted to meet each objective
will include measures of the variability in the data in order to more fully characterize the risk and
facilitate probabilistic modeling. The major characteristics of our approach to address the four
objectives are shown in Figure 1.1 and described below.
A. Plant species
Information on both crop and native plant species to chemical herbicides and other chemicals
will be obtained for different types of test endpoints (vegetative vigor, reproductive/ developmental,
ecological), and to characterize candidate species for mode of action and molecular research. We
will use a regional GIS approach to species selection (see Section 3.3). For major agricultural
regions of the U.S., a list of plant species and plant communities expected to be exposed will be
generated. The list will include crops ranked by the number of acres occupied, dominant and sub-
dominant native plants, important plant species for wildlife habitats or forage and threatened and
endangered plants. Plant species of interest will be obtained and cultivated to determine their
suitability as test species. Specific steps in the species selection process are:
1. Select species on a regional basis, defined in terms of US agroecoregions of interest. The
agroecoregions will be based on:
a). Intensity of agriculture, by identifying relative amount of farmland (or other type of
land use) vs. total land area in a given spatial unit (e.g., county)
b). Herbicide usage, using statistics to identify crops receiving the most herbicides in an
area (pounds/acre, total pounds, total pounds of active ingredient, acres treated)
4-16
-------
c). Crop diversity, using statistics to identify the number and relative amount of different
crops in an area (more diverse agriculture will indicate a greater likelihood of
herbicide sensitive crops growing in proximity to herbicide target crops)
d). Wind direction and intensity data, to indicate likelihood of drift problems due to aerial
application.
Figure 3.2 illustrates the results of preliminary analysis of agroecoregions of interest in
the US based on potential for herbicide drift impacts to non-target crops and native plants.
The agroecoregions are comprised of counties with the highest percentage of acres in
agricultural production, the highest percentage of crop acreage receiving herbicides, the
greatest diversity of crops grown, and wind speeds >10 mph.
2. Develop list of most important terrestrial plant test species for each agroecoregion.
Begin with the current EPA, ASTM, OECD, Canadian lists of test species. Ten species
have been used for testing based on the current EPA Tier I and II vegetative growth tests,
but the number of species selected could vary depending on the diversity of plants
growing in an agroecoregion. The addition or subtraction of species to be selected will
be based on the following considerations:
a) Dominant crop species
Based on yield, area or economic value
b) Dominant native vegetation based on
i) Potential native vegetation (e.g., historical based)
ii) Current vegetation surveys
iii) Satellite imagery or other form of remote sensing
iv) Wildlife habitat (food and shelter)
v) Ecological importance of species (keystone species for productivity, nutrient
cycling)
4-17
-------
c) Threatened and endangered plant species (or surrogate species in same genera)
d) Consider phylogenetic relationships to insure that a diversity of plant genotypes which
may vary considerably in responses, are considered as test species.
3. For single-species test studies, conduct experiments in pots or small field plots for the
most important endpoints and with experimental conditions representing the climate of
interest, cultural practices and herbicide exposure characteristics typical for the crop (B,
below).
4. For ecological studies and field tests, key assemblages of species for major regions will
be identified.
5. For mechanistic and molecular biology studies, the species selected initially will reflect
model systems currently in use for plant genomics / proteomics research. Later, the list
will be broadened to include those identified for different agroecosystems.
The GIS analysis will identify specific regions of the U.S. for intensive species analysis.
Based on preliminary research, the first area of interest will be the Midwestern com belt (due to
intensity of agricultural production and herbicide use), followed by California (in part because of
the intensity of pesticide use data), Pacific Northwest (in part because of ability to conduct initial
field work), northern plains states (location of crops of interest for both herbicide and gene flow
studies) or other areas.
B. Plant test guidelines
Response Endpoints. Current short-term, vegetative growth endpoints (biomass, height),
while useful to indicate general lethality of herbicides, may not provide data regarding longer-term
effects on plants such as sublethal effects, the ability of plants to recover from stress, or changes in
the competitiveness of plants. It also is clear that early plant growth effects may not predict latent
adverse reproductive effects. The current plant tests last a maximum of 28 days. This short time
period is insufficient to capture the critical reproductive phase of most plant life cycles. Not only is
this aspect important for the individual plant's ability to pass along its traits, but reproductive yield is
one of the most important economic aspects of agriculture. The very limited data available suggests
4-18
-------
exposure during vegetative versus reproductive phases of growth may not have equal influences on
reproduction and crop yield. Thus, we will conduct research to determine the effects of herbicides
on endpoints most important for the risk assessment process.
Experiments will be conducted to determine the following:
1. What is the relationship between the time of exposure during a plant's life cycle and
resultant developmental effects, both in terms of altered seed production and damage to
other storage organs such as tubers or roots?
2. What are reproductive and development endpoints of particular usefulness when
evaluating responses of native plants, and how best can these be quantified?
3. How do different classes of chemicals affect various specific reproductive or
developmental responses?
4. Do different families of plants respond differently in terms of mechanisms of uptake,
transport and degradation of different chemicals? Can we predict the specificity of this
mode of action for different species and different stages of plant development?
5. What physiological processes (biochemical and molecular endpoints) are most sensitive
to different chemicals?
Annual Plants. The relationship between vegetative and developmental endpoints will be
determined for important annual crops from key agroecoregions by:
1. A limited number of crops will be used to determine reproductive (seed production)
endpoints, building on the results from previous studies. For example, soybeans will be
used a) to gain additional information on the risks to this major crop from the central U.S.
corn-belt agroecoregion, b) to verify and build upon results with soybean from previous
studies at WED and elsewhere, and c) to develop experimental protocol methodologies
for a crop growing areas with hot summers. Peas will be used to obtain similar
information, with respect to a crop growing under cooler environmental conditions.
4-19
-------
2. Several crops will be used to determine developmental endpoints such as tuber or storage
root formation and growth. Potatoes will be used due to their known susceptibility to
ALSase herbicides along with other candidate crops such as sugar beets and carrots.
3. Crop plants will be grown in pots under controlled greenhouse conditions and exposed to
a variety of different herbicides at different growth stages. Vegetative endpoints will be
measured at 14 days after treatment to provide data that will correspond to the results
from current vegetative vigor tests. Reproductive and developmental endpoints will be
measured at the time of target organ maturity, i.e., mature seed or seed for reproductive
endpoints and tuber maturity for a developmental endpoint. Results will be given in
terms of various parameters.
4. Regression equations will be calculated to relate response to over a range of herbicide
concentrations to obtain a range of EC values (Effective Concentration for a certain
response), and to obtain coefficient of variability around the regression line. Such
measures of variability are necessary for probabilistic risk assessments. They also
indicate the possibility of sublethal plant responses at lower concentrations of the
herbicides. For compatibility with the current vegetative vigor tests, EC25 values (EC for
a 25% reduction in a response) will be calculated from the regression equations for
different response parameters.
5. After basic vegetative and developmental relationships are established for annual crops,
experiments will be conducted with annual native plants from key agroecoregions to
determine if similar relationships occur for those species.
6. Annual plant studies will be conducted first with potted plants under greenhouse
conditions to develop test protocols that can be used in multiple locations throughout the
year. Experiments will then be conducted with plants in pots placed outside, and plants
in soil in small field plots to develop test protocols that can effectively be used for
specific crops under particular exposure scenarios.
4-20
-------
7. As feasible, studies will also consider the use of annual native plants as test species
considering reproductive and developmental endpoints in addition to general growth
responses.
Perennial Plants. Reproductive and developmental responses are of particular interest for
perennial crops due to the long-time frame elapsing from field planting to crop production. For
native plants, responses of perennial species to stressors are key aspects of ecosystem responses.
Perennial plants of interest include both herbaceous plants such as grasses where the belowground
portion persists each year, as well as woody shrubs and trees. Basic aspects of the studies will be as
follows:
1. Key perennial crops from agroecoregions of interest will be evaluated both to verify
reproductive responses and to identify relevant experimental conditions for testing.
Examples are strawberries from California and grapes from California and other regions.
2. Perennial crop plants will be initially grown in pots under controlled greenhouse conditions
and exposed to a variety of different herbicides at different developmental stages.
Vegetative endpoints will be measured to correspond to the results from current vegetative
vigor tests and developmental endpoints will be measured at the time of target organ
maturity, as feasible, i.e. mature seed or seed for reproductive endpoints from species such as
strawberries and belowground sinks for other species. Results will be given in terms of EC25
for different response parameters at different harvest times.
3. The feasibility of using perennial crop plants in the field will be evaluated.
4. Experiments will be conducted with perennial annual native plants from key agroecoregions
to evaluate the feasibility of their use as herbicide test species and to determine if native
plant responses are similar to those of perennial crop plants. For example, perennial grasses,
forbs (herbaceous plants other than grasses) and seedlings from woody plants characteristic
of the central U.S. corn-belt agroecoregion will be obtained, propagated and evaluated as
possible test species in pots. Our initial research will focus on developing the experimental
conditions necessary for the successful use of these species as possible test plants for
herbicide studies. Of special importance will be vernalization requirements, soil, and water
4-21
-------
regimes. The phenology of species will be determined to optimize the time of herbicide
exposure. Initially perennial grasses and forbs may be studied because their short life-cycle
means that it takes a relatively short time for them to produce seed. In addition, their
extensive root systems provide a belowground sink which may be affected by herbicide
treatment.
5. The most advanced studies will focus on the responses of native perennial herbaceous and
woody plants species to herbicides in situ. These would be long-term studies (optimally five
years or more), primarily to measure reproductive and developmental effects. The studies
likely would be conducted in primarily Oregon, but they may be conducted in other areas of
interest. This work will build on the advances in research that result from the studies with
annual and perennial crops and annual and perennial native plants grown in pots.
Specific Experimental Objectives and Conditions. We will determine the cultural
conditions required to grow healthy test plants and herbicide treatments which produce the most
field-relevant results. Examples of key experimental conditions that will be considered are:
1. plant growth containers (size of pots for plant life-cycle tests)
2. media for potted plants (mineral soil vs. artificial soil mix, mineral soil type especially
critical for native plants)
3. watering protocols (amount, top vs. bottom)
4. environmental conditions (greenhouse vs. field)
5. herbicide concentrations (based on modeled exposures based on GIS analysis)
6. herbicide mode of exposure (foliar spray, soil application, timing during life-cycle of
plant)
7. number of herbicides in formulation (single vs. multiple chemicals).
Our experimental objectives and conditions initially will be based on the methodologies used
in current test guidelines. This will ensure that we build on past efforts while providing a path to
4-22
-------
future test designs. We will refine the methodology based on outputs (e.g., soil type) from GIS
analyses for specific agroecoregions of the U.S. Each individual experiment will contribute to
address several of the research objectives. For specific experiments, details will depend on plant
species chosen for the study and the response endpoints of interest. Examples of specific potential
experiments which may be conducted in 2002-2003 are given in Table 4.5. An example of the
research protocol used for a preliminary study to determine effects of the herbicide "Oust " on
soybeans is shown in Appendix B.
The particular experimental design (including treatments, replicates, and observations) and
statistical analysis protocol for a study will depend on the objectives and hypotheses to be tested.
We will consider other types of statistics to describe responses besides the commonly used EC25.
For example, we will consider the nonlinear curve fitting techniques such as described by
Stephenson et al. (2000) which can provide a measure of the uncertainty associated with the
response.
C. Future studies
Ecological Studies. Ecological studies will be designed to address the greatest areas of
greatest uncertainty in the risk assessment process for chemical herbicides and other chemicals.
Even though ecological studies likely will be highly herbicide- and ecosystem-specific, we hope to
establish ecological study protocols that will provide data that are broadly applicable to a variety of
risk assessments. Ecological studies could be conducted at different scales depending on the
scientific question being asked and available resources, and could range in scale from the simple to
complex. These studies will require detailed protocols including questions asked, experimental
design and endpoints needed to address those questions, and types of statistical analyses and power
of those statistical analyses to determine to evaluate results. Examples of the types of possible
studies and a sense of how they might be staged are given below:
1. Initially, ecological information could be obtained under controlled conditions. The simplest
experiments could evaluate plant competition (measured as numbers of species, biomass)
with herbicide exposure under controlled conditions, possibly in large pots. These could be
replacement series experiments with different densities of two to several species. Such
4-23
-------
studies with herbicides could be similar to those with elevated CO2 and described by Olszyk
and Ranasinghe (1994).
2. Next, small plots could be used to simulate multi-species responses to chemical herbicides
under controlled soil conditions (Pfleeger and Zobel, 1995). Test scenarios would include
sampling of existing vegetation surrounding test plots such as field margins and more
structured designs where mixtures of plants are seeded or planted at various densities and
proportions prior to intentional herbicide application. End points will include production,
reproduction and species interactions. The effects of herbicides on susceptibility of plants to
other stresses (e.g., disease, insects, and climate) could also be studied.
3. Large field-plot studies then would be conducted to refine and further develop the
methodology for determining herbicide effects at the ecological level (Taylor, 2001). The
studies would be conducted, initially at sites in the Pacific Northwest, and later in various
geographical locations throughout the country, targeted to answer specific herbicide risk
questions of interest. The sites will be selected by the previously detailed GIS approach.
The studies would be conducted adjacent to agriculture fields by a multi-disciplinary team of
scientists to determine ecological effects. The studies would be based on models of the
potential responses of a variety of plants to herbicides in ecosystems. Herbicide exposures
would be as modeled for those systems.
4. Finally, field surveys for ecosystem effects of pesticides will be conducted by comparing
different farming systems. For example, organic farming systems are now reaching
sufficient maturity so that their ecological characteristics can be compared to those of
conventional (pesticide using) farming systems. Studies of organic farming systems to date
have focused on the effects on crops including yield and profitability, and of the general
environmental effects such as changes in soil quality and nutrient runoff (Reganold et -al.
2001). However, these systems could be used to test the hypothesis that low levels of
herbicides can, indeed, affect native plant populations, and, hence, their suitability as habitat
for wildlife populations. Other wildlife population studies could also be extended to study
herbicide effects. For example, OPP's Biopesticides and Pollution Prevention Division
(BPPD) currently is looking at the relationship between insecticide use and breeding bird
4-24
-------
data, and will be exploring further linkages to understand relationships with avian
populations (Brandt, E., Personal Communication, OPP).
Mechanism of Action for ALS Herbicides. Most past research on low-dose, high-toxicity
herbicides has identified a single mode of action, impairment of the ability to produce branched
chain amino acids. However, there is reason to believe that other process such as cross membrane
transport may be impaired, especially in the species where reproductive effects have been identified
(Taylor et al., 2001). Studies will be developed to evaluate alternative modes of action for ALS
inhibitors, and the effects of ALS modes of action on plant structure and function. For example, in
terms of transport processes within the plant, radioactive tracer studies could be performed to
determine the tissues containing the herbicide or its breakdown products periodically after
application. Identification of the labeled compounds will be done to verify their composition and to
determine if the herbicide is directly causing the impairment or if a physiological system has been
disrupted. While ALS herbicides will serve as model systems for the proposed mode of action
research, we will address herbicides with other novel modes of action which may be developed
during the course of this project. Studies on modes of action will be based on literature review and,
if feasible, on information available from herbicide manufactures (Beyer et al., 1988).
Rapid screening genomic tests to detecting potential effects of pesticides. As currently
conducted, the pre-registration evaluation of chemical herbicides or other chemicals considers a very
narrow range in the genetic, economic and ecological breadth of organisms present in the
highly-diverse ecosystems found in the United States. We need to develop methods that can serve as
biomarkers of ecological effects of environmental stresses (McCarty et al., 2002; McCarthy, 1990).
They need to be specific for herbicide effects in plants, as such as cholinesterase inhibition as a
biomarker of insecticide exposure in animals (Chambers et al., 2002). Such biomarkers would
enhance the ability of EPA to better understand the effects of herbicides and other chemicals on
target and non-target plant species at the molecular level. Information on genes affected by
herbicides and other chemicals could be assembled into a single database for structure-activity
modeling of pesticide behavior and anticipated effects. Since the first detectable response of an
organism to a toxin is typically a change in the expression of its genes (discernible as altered
abundance of messenger RNA) it is logical to evaluate impacts of toxins initially by examining
mRNA through differential display methodologies or standard microarray assays to characterize the
4-25
-------
effects of classes of pesticides. However, in the longer-term, it would also be useful to study
pesticide effects through genomic analysis of notable instances of resistance or susceptibility to
pesticides. The range of methodologies involved in this latter work would allow for the collection of
data from a diverse set of organisms whose full genomic sequences have not yet been completely
determined, are not yet available in the public domain, or are not anticipated to be determined. Once
the genetic context for responses to pesticides by a broader range of organisms is characterized,
rapid screening methodologies can readily be developed and applied. From this overall approach,
EPA would gain the necessary flexibility to analyze new classes of effects of pesticides. The
Agency could then move from database construction to the development of microarrays specific for
assessing the effects of new pesticides with similar modes of action.
Molecular Biology Tools. Molecular biology tools may allow us to identify sites of action
and to elucidate modes of action for different chemical classes of herbicides. For example, while it is
generally recognized that sulfonylurea herbicides are inhibitors of the enzyme acetolactate synthase
and that phenoxy herbicides inhibit acetyl coA carboxylase, less is known about how other
herbicides may affect the expression of genes. Weeds and crops which have spontaneously
developed resistance to herbicides, and crop plants engineered to be resistant to specific types of
herbicides, can each serve as means to identify genes and proteins that may be useful in mode of
action studies. For example, by comparing the time course of gene expression at different
developmental stages in herbicide resistant and susceptible plants prior to and after exposure to
herbicides or other test chemicals or environmental stressors, one may be able to identify gene or
protein sequences of interest for subsequent biochemical characterization and for use in
physiological and ecological studies.
Genomics information such as DNA sequences for entire plant genomes are becoming
increasingly available in the public domain (e.g., the Arabidopsis and Oryza genomes that have been
sequenced and others still being worked out) and may become useful to evaluate effects of chemical
herbicides. Currently, the only commercially available gene chip for plant genes is for Arabidopsis.
However, within the next few years, it is anticipated that commercial gene chips for microarray
analyses will be available for corn, wheat and soybeans. In addition to commercial gene chips, we
have tools available locally at WED and through the OSU Center for Gene Research and
Biotechnology to isolate, characterize and sequence nucleic acids and proteins of interest as markers
4-26
-------
of exposure to specific kinds of agricultural chemicals. Use and critical analysis of microarray
approaches (see Fig. 5-5) are expected to play a key role in these future studies (Dahlquist et al.,
2002; Pan, 2002; Ramoni et al., 2002; Simon et al., 2002; Yang and Speed, 2002).
WED's ability to conduct this molecular biology research will require additional resources.
The first step in the establishment of the program will be the recruitment and hiring of an EPA
Postdoc familiar in current physiological/biochemical/molecular methodologies and research
questions regarding herbicides and terrestrial plants.
4.4 Time line
The timing of research on plant effects from herbicides is in three phases (Table 4.6). The
first in the sequence of research events was the problem formulation phase in FY2002. Then a five-
year (FY 2003-FY 2007) effects research program will be launched to address the highest priority
research objectives. Finally, there will be the synthesis and integration phase (from FY 2007
onward), during which time information will be analyzed and summarized to develop new tests.
Further plans will be made for intensive research on new areas needed for complete risk assessments
which may be beyond the scope of the current project, e.g. ecosystem response tests and
genomics/proteomics based mechanistic tests.
During FY2002 the focus was on preparation of this research plan and the development of
basic resources (i.e., plant growing facilities, herbicide treatment equipment, and experienced staff)
necessary for plant effects research at WED. Preliminary scoping research has been conducted to
develop the initial GIS procedures for species selection; to update the greenhouse, growth chamber,
and nursery area infrastructure for growing plants; to gain familiarity with phenology, morphology,
and productivity of a variety of possible test species; to evaluate performance of the pesticide track
sprayer; to establish health and safety and quality assurance protocols to insure scientifically
credible studies in the future; and to initiate preliminary experimental studies-
Over the following five years (FY2003-2007) the research plan will be implemented and the
research will follow a general sequence of experiments moving from more toxicologically-oriented
to ecologically-oriented work, to provide a range of information and tools available for ecological
risk assessments. In general research will progress over time from:
4-27
-------
Toxicological Responses-
•>Ecological Responses
Annual Plants-
•>Perennial Plants
Crop Plants
¦>Native Plants
Individual Species
•>Multiple Species
Vegetative Responses —
¦>Reproductive Responses
Plants in Pots
—> Plants in Field Soil
Greenhouse Studies
>Field Studies
Each individual experiment will address several research objectives based on questions
being asked and availability of staff. In addition to whole-plant test oriented research,
preliminary studies will be conducted to develop methodology needed for to address ecological
questions regarding chemical herbicides and to develop mode of action and molecular tools.
During the third year of the project (2005) a peer review workshop will be held to assess
progress in research to that date and to finalize the most critical research to be conducted during
the last two years of this plan.
At the end of the five years of the present research plan, single test species protocols will be
dev eloped and a course of action will be prepared for further work on ecological tests for chemical
herbicide effects and the development of molecular tools based on preliminary research and
evaluation. Should additional funds become available, the ecological and mode of action and
molecular studies can be initiated sooner.
4-28
-------
Table 4.1 Examples of current criteria for performing vegetative vigor studies for terrestrial plants and chemical herbicides (Stavely,
2002).
Subdivision
J
OPPTS Series
850
ASTM 1963-98
OECD 208
(existing)
OECD 208
(proposed)
Environment
Canada
Temperature °C
NS
25 ± 3 (Day)
20 ± 3 (Night)
20-30
NA
NS
NS
re: Humidity %
NS
70 ± 5 (Day)
90 (Night)
above 30, 50
recommended
NA
NS
NS
Light Intensity
NS
350 ± 50
A
-------
Table 4.2 Summary of recent efforts to address research needs for tests of plant effects from
chemical pesticides, emphasizing WED contributions and concerns regarding new low-dose high-
toxicity herbicides.
1. 1990 Workshop in Corvallis, OR to evaluate non-target plant testing in subdivision J of the pesticide
guidelines and define research needs. This workshop included participants from government
(OPP, ORD, Region 10, state agencies, Canada, academia (state universities from Colorado,
Idaho, Washington and Oregon (state universities, and industry (Du Pont, ICI„ ABC
Laboratories, Kodak, Springborne Laboratories). EPA Report EPA/600/9-91/041. A summary of
issues shown in Appendix 4.D.
2. 1993, 1994, 1996. WED staff published papers in support of OPP, confirming the inadequacy of
present test guidelines to determine reproductive effects of low-dose herbicides. This research
was supported and partially funded by OPP and Region 10.
3. 1999 WED Staff (T. Pfleeger and J. Fletcher) on the steering committee developing the non-target
plants workshop for ILSA Risk Science Institute sponsored by OPP EFED. The workshop was
held in 1999 in Washington, DC and was attended by government, industry and academics. One
of the major goals of the workshop was to identify OPP research needs. Following the workshop
staff from OPP, WED and Region 10 met for an additional day to discuss the findings of the
workshop as they pertained to the research needs of OPP EFED. Background information
published as SET AC Publication on Impacts of Low-Dose High Potency Herbicides on Non-
target and Unintended Plant Species (Ferenc, 2001), with specific research recommendations
from that publication given in Appendix 4.E.
4. Spring, 2001 WED staff became involved in the development of the NHEERL implementation plan
for Goal 4 (safe communities, i.e., pesticides). The goal of this committee is to take the research
needs of OPP and develop a comprehensive research strategy to fulfill those needs. OPP is a
participant at these meetings.
5. June, 2001 OPP EFED convened a meeting of the a Scientific Advisory Panel in Washington, DC in
conjunction with the NAFTA US-Canada Workshop on Impacts of Low-Dose High Potency
Herbicides on Unintended Plant Species to discuss the Non-target test guidelines and the need for
change. Staff from OPP and WED made presentations at this meeting.
6. July, 2001. Reports of non-target effects from low-dose, high-toxicity herbicide Oust in Idaho. Staff
from OPP, WED and Region 10 with Idaho State officials and local growers visited sites.
7. December, 2001. United Kingdom issues draft efficacy guidelines for Effects of Non-Target Crops of
Highly Active Herbicides - Including Mixtures and Sequences. This document recognizes the
need for highlighting the effects of newer herbicides, but no new specifics are given regarding
those tests.
8. January, 2002. Final Report of the FIFRA Scientific Advisory Panel (SAP) Meeting of June 27-29,
2001.
9. March, 2002. Memo by I. Suzenaurer (OPP, EFED) stating critical chemical pesticide effects research
needs.
10. May 2002. Report from the International Workshop on Non-Target Plant Risk Assessment. January
15-17, 2002, Alexandria VA.
11. June 2002. WED staff present invited seminar demonstrating an example of a GIS -based analysis to
identify crop and native plant species useful for a agro-ecoregion based plant testing protocol.
4-30
-------
Table 4.3 Examples of reports of non-seedling stage pesticide effects on non-target or simulated
non-target plants (modified from Maxwell and Weed, 2001; Ratsch and Fletcher, 1991). This list
only includes direct crop effects and not carryover effects due to residual herbicide from the
previous year.. The non-seedling effects may, or may not have been associated with leaf injury.
Chemical Non-target Response Reference
Plant
2,4-D
Fieldbeans
Flower and pod
Lyon and Wilson, 1986
chlorsulfuron
Barley
Reduced seed/fruit yield
Lemerle et al., 1993
chlorsulfuron
Cherry
Reduced fruit yield,
quality
Bhatti et al., 1995
chlorsulfuron
Cherry
Reduced fruit yield
Fletcher et al., 1993
chlorsulfuron
Canola
Reduced seed yield
Fletcher et al., 1996
chlorsulfuron
Pea
Reduced seed yield
Fletcher et al., 1995
chlorsulfuron
Smartweed
Reduced seed yield
Fletcher et al., 1996
chlorsulfuron
Soybean
Reduced seed yield
Fletcher et al., 1996
chlorsulfuron
Sunflower
Reduced seed yield
Fletcher et al., 1996
dicamba
Soybean
Pre- and post-bloom
Weidhamer et al., 1989
dicamba
Soybean
Pre- and post-bloom
Al-Khatib, K. and D.
Peterson, 1999
Maleic hydrazide
Soybean
Reproductive effects
Helsel et al., 1987
metsulfuron methyl
Soybean
Reduced seed/fruit yield
Boutin et al., 1999
metsulfuron methyl
Soybean
Reduced seed/fruit yield
Boutin et al., 1999
metsulfuron methyl
Soybean
Reduced seed/fruit yield
Boutin et al., 1999
metsulfuron methyl
Soybean
Reduced seed/fruit yield
Boutin et al., 1999
primisulfuron
Soybean
Pre- and post-bloom
Al-Khatib, K. and D.
Peterson, 1999
sulfometuron
Potato
Reduced tuber size and
quality
Westra et al., 1991
imazamethabenz
Potato
Reduced tuber quality
Westra et al., 1991
thifensulfuron /
tribenuron
Potato
Reduced tuber quality
Westra et al., 1991
4-31
-------
Table 4.4 Examples of limitations for current assessment indicators for non-target effects of
herbicides on plants (adapted from Maxwell and Reed, 2001; as based on Obrigawitch et al., 1998).
Assessment Indicator Limitations
Visual Conflicting data on correlation with
Chlorosis subsequent crop yield
Anthocyanin formation
Height Variable correlation with yield, not well suited
for mature dicotyledonous plants with multi-
stem or prostrate forms
Plant Biomass Suitable when vegetative portions are
harvested, but not necessarily when
reproductive parts are harvested. Can be
affected by increased branching. Varies in
sensitivity with time during plant life-cycle
4-32
-------
Table 4.5 Examples of experiments on effects of chemicals on terrestrial plants 2002-2003.
Time Frame Objectives
Species Herbicide Conditions
Summer-Fall Evaluate possible test plant species- annual crops
2002 Corn belt, Pacific Northwest
Compare vegetative and reproductive responses
Determine cultural conditions
Evaluate herbicide treatment conditions
Soybean Oust
Potato
Pots
Outside Nursery,
Greenhouse
Summer-Fall Evaluate possible test plant species- annual crops, corn belt
2002 Compare reproductive response with different herbicides
Determine cultural conditions
Evaluate herbicide treatment conditions
Soybean
Multiple
Pots
Greenhouse
Summer-Fall Evaluate possible test plant species- annual crops
2002 Pacific Northwest
Compare vegetative and reproductive responses
Determine cultural conditions - cool season crop
Evaluate herbicide treatment conditions
Pea
Oust
Pots
Greenhouse
Winter 2002- Evaluate possible test plant species- native plants, corn belt
Spring 2003 Compare vegttative and reproductive responses
Determine cultural conditions
Evaluate herbicide treatment conditions
Multiple
TBD
Pots
Greenhouse
Spring- Evaluate possible test plant species- crops
Summer 2003 corn belt, Pacific Northwest
Evaluate belowground developmental responses
Determine cultural conditions
Evaluate herbicide treatment conditions
Potato TBD
Soybean
Pots
Greenhouse
Small Field Plots
Fall 2003 Evaluate possible test plant species-native plants, corn belt
Compare vegetative and reproductive responses
Determine cultural conditions, soil type
Evaluate hprhiriHr trpatment rr.nrlitir.nq
Multiple
TBD
Pots
Outside Nursery
* TBD = To be determined
4-33
-------
Table 4.6 Time Line for Chemical Herbicides and Terrestrial Plants Research
FY2002
Produce Research Plan
OPP Workshop
Regulatory Needs
Literature Search
Produce QA Plan
Track Sprayer
Protocols
Health and Safety
Protocols
Preliminary Plant
Experiments
Soil Type
Environment
Vegetative vs.
Reproductive/
Development
Endpoints
FY2003
Contribute to APM on
Strategy for Updated
Test Guidelines:
Finalized Research Plan
Use GIS to Identify
Crops and Native
Plants at Risk from
Herbicides: Case
Studies
Corn belt
California
Pacific Northwest
Initiate
Greenhouse/Nursery
experiments on
Vegetative and
Developmental
(Seed/tuber) Responses
of Annual Crops and
Native Plants
Initiate Field
Experiments on
Developmental
(seed/root/tuber)
Responses of Annual
Crops
Test Experimental
Procedures for Plant
Growth and Herbicide
Exposure
Pots in
Greenhouse/Field
Mineral vs. artificial
soil
FY2004
Use GIS to Identify
Crops and Native
Plants at Risk from
Herbicides: Other Case
Studies (example)
Northern Plains
Continue Greenhouse
Experiments on
Developmental
Responses of Annual
Crops and Native
Plants Other Case
Studies
Field Experiments on
Developmental
Responses in Annual
crops and Native Plants
Plan Greenhouse and
Field Experiments for
Reproductive
Responses of Perennial
Crops
Test Cultural
Procedures for
Optimum Plant Growth
Watering
Hire PostDoc for Mode
of Action / Molecular
Biology Studies
Contribute to APM on
evaluation of risk
assessment methods for
herbicides
FY2005
Use GIS to Identify
Crops and Native
Plants at Risk from
Herbicides
Other Case Studies
Continue Greenhouse
and Field Experiments
on Developmental
Responses in Annual
crops and Native Plants
(other case studies)
Initiate Greenhouse and
Field Experiments for
Reproductive an
Developmental
Responses of Perennial
Crops and/or Native
Plants
Initiate Studies of
Mode of Action for
Developmental Effects
of ALS herbicides
Peer Review Workshop
on Plant Effects
Research
Contribute to APM on
regional approach to
risk assessment
FY2006
Produce Improved
Vegetative and
Reproductive /
Developmental Test
Methodology for
Crop and Native Plant
Species in Major US
Agroecoregions
Cultural procedures
(propagation, soil,
watering, field vs.
greenhouse)
Continue Greenhouse
and Field Experiments
for Reproductive and
Developmental
Responses of Perennial
Crops and/or Native
Plants
Continue Studies of
Mode of Action for
Developmental Effects
of ALS herbicides
Initiate Evaluation of
Molecular Basis for
Developmental Signals
and possible Molecular
Detection of Effects
Contribute to APM on
draft new protocol /
guidelines for
vegetative vigor test
FY2007
Produce Developmental
Test Methodology
Seed and belowground
sink herbaceous species
Cultural procedures
Measurement
Endpoints
Continue Field
experiments for
Reproductive
Responses of Perennial
crops
Develop Protocols for
Field Ecological
Research
Develop Protocols for
Genomics Research for
Factors
Controlling/Detecting
Herbicide Effects
Contribute to APM on
draft new protocol /
guidelines for
reproductive
/developmental test
4-34
-------
B. Improve Species
Test Guidelines
Reproductive/Developmental
Endpoints
Nontraditional Species
Experimental Conditions
A. Improve Test
Species Selection
Process
CIS, Potential Chemical
Exposure Based
D. Develop Mode of
Action Studies and
Molecular Biology
Tools
Extrapolation across species,
Field detection for Chemical
and Gene Flow Effects
C pr0vlde Input for
Ecosystem Response
Tests
Base on Model Needs, Multi-
species, System Endpoints
Inputs for Ecological Risk
Assessments, Wildlife Habitat
Models
Figure 4-1 Research Objectives for Effects of Chemical Herbicides on Terrestrial Plants
4-35
-------
VEGETATIVE GROWTH
leaves
roots
stems
40%
&
A
-------
5 Effects of Gene Flow from Transgenic Crops
5.1 Introduction
A. Rationale
Statutory authority for Agency research on the ecological effects of chemical and biological
pesticides is provided by the Federal Fungicide, Insecticide and Insecticide Act (FIFRA) and the
Federal Food Drug and Cosmetic Act (FFDCA). Evaluations for compliance to FIFRA and FFDCA
are determined in coordinated reviews of registrant applications with USDA and FDA respectively.
In July, 2001, Final Rules and Proposed Rules for Plant-Incorporated Protectants (PIPs, i. e.,
engineered traits for crop protection expressed in transgenic plants), were published in the Federal
Register (40CFR Parts 152 and 174, Appendix C). Regulated PIPs include nucleic acids and the
proteins they encode to confer to weed, disease or insect pests of plants. Because of the pesticidal
nature of these products, research to develop improved methods for assessing potential ecological
risk is covered under the Implementation Plan for Government Performance Results Act (GPRA)
Goal 4, Safe Communities.
Specific objectives of the Gene Flow Effects Project will be to address Agency needs for
registration standards for transgenic plants containing PIPs that will help ensure their safety to the
environment, and for human and animal food consumption. The Gene Flow Effects Project will
develop methods to assess and predict the ecological effects of gene flow on plant community
structure and succession. Crop protection PIPs to be evaluated may include traits for herbicide,
disease or insect resistance. Methods will be developed to measure, assess and predict the effects of
gene flow on selected individual species of herbaceous and woody plants and on plant community
structure and succession in crop and non-crop ecosystems. Recognizing that specific plant species,
plant communities and ecosystems may differ in different geographic regions of the United States,
specific crops, traits and geographies will be selected and prioritized. Anticipated Project outputs for
the Agency will include scientifically defensible methods that measure, assess and predict changes
in plant community structure of crop and non-crop plants in terrestrial ecosystems. While gene flow
may have broad potential national and international impacts, our initial emphasis at WED will be on
identifying suitable test plants and engineered crop protection traits for specific geographic regions
5-1
-------
of the United States, e. g., for herbicide, disease or insect resistant canola or grasses in the Pacific
Northwest. It is possible that similar work may be carried out at other NHEERL Ecology Divisions,
which would focus on gene flow, transgenic pesticidal pollen or feral transgenic plant effects in
aquatic, rather than in terrestrial ecosystems.
Research carried out under this plan is expected to contribute to recommendations for short-
term data requirements to be submitted by registrants for specific types of transgenic traits in
specific types of crops in specific geographies. It may also result in criteria for post-registration,
longer-term adverse effects plant monitoring requirements.
Gene flow effects research is anticipated to result in recommendations for measuring,
monitoring and mitigating the extent and effects of gene flow from crop plants to other crop and
non-crop plants. Effects of gene flow on plant community structure and succession, and on rare and
endangered plant species, will be examined in both crop and non-crop ecosystems. Evaluation of the
effects of gene flow from transgenic plants will focus on traits that have been developed for crop
protection purposes such as herbicide, disease or insect resistance. Plant parts of special interest are
pollen and seeds, since they typically are the primary means of dispersal via wind, insects, water or
animals. The germination of pollen and seed respectively can bring about hybridization between
compatible plant species, and establishment of a new generation of hybrid plants, which may in turn
shed and disseminate hybrid pollen and result in the establishment and spread of subsequent hybrid
seed progeny.
Initially gene flow effects will be evaluated empirically on a local or regional bases; longer
term, they should be considered domestically and internationally on landscape levels, given the
potential for movement of both pollen and seeds by biological and abiotic natural and anthropogenic
means. Adverse effects of gene flow on plant associated symbionts, saprophytes, and pathogens,
should also be considered, since those integral plant community components impact the sustain
ability of plant ecosystems. In the longer term, probabilistic risk assessment modeling approaches
will be developed, based on parameters identified in our initial empirical studies. While the
potential for horizontal gene transfer between plants and microbes or plants and invertebrate or
vertebrate herbivores and pollinators of plants is generally assumed to be low, such events could
have adverse ecological and health effects. Accordingly, it is recommended that in the long-term, as
-------
resources may permit, the ecological and health impacts of horizontal gene flow should be
considered between plants and other types of organisms as well.
Currently, engineered crops are planted on tens of millions of acres in the US alone. Pollen
from transgenic crops may hybridize with other related crops or weeds, potentially conferring
resistance to crop-crop or crop-weed hybrids. If fertile, the resultant hybrids and their subsequent
progeny can produce seed which may result in continued persistence and spread beyond the confines
of the original intended agronomic fields. Multiple resistances to PIPs can occur simultaneously.
For example, resistance (engineered and/or spontaneous in origin), to two or three herbicides
belonging to different chemical classes is already showing up in canola in the UK and Canada
(Orson, 2002). At a recent OPP Biological Pesticides Division and OPTS workshop in Washington,
DC, high concerns about these and similar issues were expressed by attendees from EPA Regional
Offices regarding potential impacts of transgenic gene flow. Due in part, to the relative scarcity of
ecological (Appendix D), or human health data on effects of genetically engineered plants in the
peer reviewed literature, concerns about the potential contamination of crops, feed and food with
genetically modified materials have been raised by the media, activist environmental groups,
commercial food processing companies and producers and consumers of organic crops. Demands
have been made to label foods or ingredients derived from transgenic crops, and "GMO-free" has
become a marketing tool to address the concerns of food processors, food retailers and domestic and
international consumers. Legal actions against the Agency by activist groups and the high media
attention paid to potential adverse ecological effects of transgenic pollen on monarch butterflies
(Danausplexippus) (Sears et al., 2001), allergenic effects of Starlink™ corn (Segarra and Rawson,
2001), and contamination of Mexican land races of maize with transgenic genes from cultivated corn
(Quist and Chapela, 2001), further suggest the need to support Agency policies through research in
assessing potential ecological effects of gene flow from transgenic crops.
Research to be done at NHEERL-WED initially will have a regional focus; e.g., in the Pacific
Northwest, assessing effects of gene flow from canola to mustard and other compatible weedy or
crop species or from grass crops to compatible weedy and native species. Two widely grown
engineered crops, corn and cotton containing herbicide or insect resistance genes, were excluded
from our proposed studies on gene flow. Our reasoning for that is based on the absence of known
compatible weedy and native corn species in the continental US, and the presence of compatible
5-3
-------
cotton relatives only in the Florida Keys, the U.S. Virgin Islands, Puerto Rico and the Caribbean,
and Hawaii (Wozniak, 2002). Potential impacts of competitive ability, spread and persistence of
crop-weed hybrids on plant community composition, structure, and wildlife habitat quality will first
be examined by empirical approaches; later, these effects may also be examined by modeling
approaches. The project is expected to have synergies with research at NHEERL-WED in the areas
of effects of chemical pesticides on plants (Plant Effects) and modeling to determine the effects of
multiple stressors on wildlife habitat. If areas beyond the Pacific Northwest are to be considered,
collaborative research efforts with other NHEERL ecological and health effects Divisions also may
ensue, e.g., to study potential impacts of the engineered genes, pollen, feral transgenic plants, and
plant-derived products in aquatic habitats and on human health. Currently, literature reviews,
discussions with plant breeders, molecular biologists and plant ecologists and participation in
transgenic plant workshops are helping us identify species and genes of most interest and thereby
refine our research strategies. Inputs for our proposed research are also being actively sought from
Agency colleagues in the Office of Pesticide Programs, the National Center for Environmental
Assessment and in EPA Regional offices. The research proposed herein would constitute part of an
ORD-wide Five Year Initiative in Biotechnology.
B. Background
Genetic engineering permits the introduction of genes from diverse plant, microbial and
animal sources into agronomic plant species. Once inserted into plants, those genes potentially may
spread or flow to other compatible crops, weeds and native species. The resultant F1 hybrids,
created by hybridization with transgenic pollen transported via wind or insects, may in turn self, out-
cross to other compatible species, or backcross to the transgenic or non-transgenic parent. In
addition to hybridization assisted scenarios, the transgenic genes may move via feral transgenic
plants or seeds, i. e., over-wintering transgenic plants or seeds which escape cultivation, or via seeds
that have fallen from planters, combines, trucks, or railroad cars during routine planting, harvesting,
and shipping activities. Incidental transport of seeds via birds, other animals, and humans can
further contribute to the unintended spread of transgenic genes beyond their intended areas of
cultivation. While it is commonly argued that cultivated crops would not persist well outside of
agronomic situations due to their need for high levels of fertilizer, limited information is available on
5-4
-------
the survival, fertility and out-crossing potential of hybrids formed between crops and compatible
weedy or native species. Many species in each of the two latter categories (weedy and native
species), commonly thrive in low fertility soils. It may thus be anticipated that soil fertility
requirements of hybrids between crops and weeds or native species may be lower than that of their
transgenic crop parents. Available hybridization information, based largely on information gained
with non-transgenic crops, tends to document that it can occur between specific combinations of
crops and weeds. Less information is available on hybridizations between transgenic crops and
weeds, or between crops or crop/weed hybrids and native species. Downstream impacts of such
hybridizations on plant community composition, function and habitat quality remain hypothetical
and largely unknown. The studies proposed below in the Objectives and Approach portions of this
Research Plan (Sections 5.2 and 5.3 respectively) address these data gaps, developing molecular
methods to track gene flow, and beginning to identify potential non-target ecological effects of gene
flow.
In recent years, since the advent of commercial transgenic crops, such as cotton and corn
expressing insecticidal (Bacillus thuringiensis subsp. kurstaki) genes, data are beginning to be
available on potential impacts of toxin production on populations of beneficial insects, insect pests
and birds. Much effort is being expended to develop strategies to minimize development of
resistance in target pest populations to pesticidal toxins and to promote the useful commercial life of
both plant and microbial delivery systems for pesticidal genes. These strategies include use of
buffer rows of conventional crops, planting refugia, using multiple or alternative insecticidal genes,
targeting tissues and times of gene expression. Some data also are available on potential non-target
effects of transgenic plants on invertebrate and vertebrate herbivore and pollinator populations, or on
plant-associated microbial pathogens, saprophytes or symbionts. However, the impact to the
invertebrate community of transfer of pesticidal genes to non-target plant populations has not been
well studied, particularly with regard to determining potential impacts on plant community structure
and function. Expression of pesticidal genes in transgenic crops and unintended transfer of those
genes to other crop and non-crop plant species, might each potentially result in changes in the
population sizes of beneficial insects such as pollinators, as well as reducing targeted populations of
insect pests or plant pathogens.
5-5
-------
Researchers at WED are experienced in the development and use of molecular methods
(Fischhoff and Watrud, 1988; Porteous, etal., 1994,1997; Watrud etal., 1985; Watrudetal., 1987a,
1987b; Watrud et al., 1996; Watrud et al., 1998; Watrud, 2000; Widmer et al., 1996 a, 1996b;
Widmer et al., 1998; Widmer et al., 1999; Winton et al., 2001). WED researchers also have prior
experience with risk assessment of genetically engineered microbes and plants (Pfender et al., 1995;
Seidler et al., 1998; Donegan et al., 1999; Di Giovanni et al., 1999a, 1999b). Additional key insights
for the research we propose below have been provided by National Academy of Sciences reports on
pesticides use (2000a) and on environmental effects of transgenic plants (2002); the Scientific
Methods Workshop: Ecological and Agronomic Consequences of Gene Flow (2002) (Appendix E),
meetings on biosafety and risk assessment of engineered plants (Appendix F), biotechnology
references recommended by the US EPA Biotechnology Steering Committee (Appendix G), and
numerous journal articles including (Bergelson et al., 1998; Dale et al.,1996; Duggan et al., 2000;
Purrington and Bergelson, 1995; Quist and Chapela, 2001; Snow, 1997; Vierhelig et al., 1995;
Siciliano et al., 1998). Numerous reports and publications [including those cited above as well as the
National Academy of Sciences reports (2000a, 2000b and 2002) and the Scientific Methods
Workshop: Ecological and Agronomic Consequences of Gene Flow from Transgenic Crops to Wild
Relatives (2002)], demonstrate a clear need for more risk assessment research on products of
agricultural biotechnology. The specific research questions and scientific approaches that we
propose below are focused on (a) development of molecular methods to detect gene flow and (b)
assessment of potential ecological effects of gene flow on the structure and functions of plant
communities (Tilman, 1988) in agronomic and non-agronomic ecosystems.
5.2 Objectives
To address the overall goal to determine the effects of gene flow from transgenic crops
(Figure 1.1), the proposed research will develop methods to assess the ecological effects of gene
flow from transgenic plants and also the effects of feral transgenic plants, i.e., escapes from
cultivation. The two major objectives of the research are (Figure 1.1):
1. to develop molecular methods to assess gene flow potential and exposure to
transgenic genes in compatible weed, native and crop plant species, and
5-6
-------
2. to assess ecological effects of exposure to transgenic genes in non-target plant
species and communities, i.e., plants and associated biota (pests, pathogens,
pollinators, herbivores and symbionts).
The research will focus on transgenic plants developed for resistance to herbicides, plant
disease or insect pests. Pending availability, transgenic plants designed for specialty purposes such
as chemical or pharmaceutical production, also would be of interest to examine in our studies. The
specific scientific objectives include the ones indicated below. The studies would be designed to
provide methods, protocols and data to track gene flow and its ecological consequences. Longer
term, the research is envisioned to provide inputs, i.e., define parameters for probabilistic models to
be used in the risk assessment of gene flow from transgenic plants.
5.4 Experimental Approach
An overview of the proposed ecological research and its fit with health effects research
within NHEERL is shown in Figure 5.1. The exposure and ecological effects components ofthe risk
assessment research for GM crops are shown in Figures 5.2 and 5.3 respectively. Figure 5.3
additionally illustrates the types of non-target potential ecological effects on plants that would be
examined in plants at individual, population and community levels. The research proposed herein is
focused on terrestrial habitats, primarily in the Pacific Northwest. The evaluation criteria considered
in the selection and prioritization of plant species of interest as potential donors and recipients of
transgenic genes in the gene flow exposure and ecological effects studies proposed below is shown
in Table 5.1. Depending on potential collaborations with public and private sector researchers, the
geographic scope of the research could expand both domestically and internationally. Regardless of
geographic location, the research will provide methods and proof-of-concept for approaches to
assess potential ecological effects of gene flow. Figure 5.4 conceptually illustrates the formation of
patches of transgenic plants, which may arise as a result of gene flow resulting from wind or insect
pollination of compatible species, or from the incidental transport of feral seeds. Progeny from the
transgenic patches in turn, may serve as additional sources of transgenic gene flow. Figure 5.5
illustrates the types of molecular methods based largely on PCR and genomic technologies that
would be used to estimate exposure to non-target plants such as weedy or native species, resulting
5-7
-------
from gene flow from transgenic plants. Some key aspects of the gene flow research are
highlighted in Figure 1.1
A. Develop molecular methods to assess gene flow
In FY2002, initial emphasis is being given to reviewing the literature and consulting with
experts in the academic, government and private sectors to identify and prioritize crops and
transgenic traits of greatest interest to the Agency. For example, is gene flow from herbicide
resistant wheat, corn or potatoes in the northwest of greater interest than herbicide resistant gene
flow from cotton in the southeast, or sugar beets in the southwest, or from corn, soybeans and
sunflowers in the Midwest? In addition to regulatory needs, the technical feasibility and availability
of tools will also be considered. For example, are plant materials and nucleic acid sequences of
genes of interest publicly available? If not, can we access them from the private sector? How and
where can we gain access to fields where the crops are/have been grown, so that we sample plant
and soil materials for the presence of genes of interest? Once a selection has been made, e.g., to
study the ecological effects of gene flow of a given trait such as herbicide resistance in a particular
crop, research will be conducted to develop and use a quantitative polymerase chain reaction (PCR)
methodology (Sambrook and Russel, 2001), to detect, monitor, and quantify the presence,
persistence and spread of a targeted gene. When information on the mode or site of action of a
herbicide or other pesticide is known, sequences of nucleic acids known as primers, can be selected
or designed to essentially "bait" for a specific type of DNA in an environmental sample. Due to the
presence of unique markers, promoters and coding sequences used in the engineering process,
primers can be designed to detect the targeted transgenic DNA sequence. DNA extracted from
environmental samples can then be restricted or cut with restriction enzymes that recognize specific
sequences of nucleotide bases. The fragments are then "amplified'' or replicated using a PCR
reaction mix containing the appropriate primer, nucleotide bases, and Taq polymerase enzyme.
Using ultra-violet light, the presence of the targeted DNA fragments in environmental samples are
visualized and photographed following electrophoresis of the fragments on agarose gels. The gels
also can be scanned to graphically illustrate the presence of bands of interest. Using fluorescently
labeled primers, and appropriate instrumentation (such as the PE 7700 Gene Detection System
5-8
-------
available at WED), we can determine the presence/absence of DNAs of interest, and also determine
the number of copies of a given gene in a sample.
Specifically, the following approaches will be taken to develop required methods:
1. Identify and access or design primers and/or probes for hybridization for qualitative PCR
or quantitative real-time PCR method to detect engineered gene trait (e. g,. herbicide
resistance to glyphosate or disease resistance to a plant pathogen). Primers to be selective
for gene in above and below ground plant parts of target crop (e. g., transgenic canola) and
non-target weedy plant (e. g., bird's rape mustard), in crop and non-crop plant ecosystems.
2. Develop and use qualitative and quantitative molecular methods such as RT-PCR to detect
genes of interest and to assess gene flow and its ecological effects on above ground and
below ground plant community composition and functions.
3. Create F1 hybrids between selected donor crop and compatible recipient weedy, native or
crop species and determine inheritance and expression of the transgenic gene in F1 and F2
populations.
4. Utilize Southern hybridizations, microarray and other genomic and PCR methods such as
RFLPs (restriction fragment length polymorphism), SNPs (single nucleotide
polymorphisms), ESTs (enhanced sequence tags), AFLPs (amplified fragment length
polymorphisms), RAPDs (random amplified polymorphic DNA), SSLP (single sequence
length polymorphism), microsatellite, SSR (simple sequence repeat) markers, STS
(sequence tagged site) etc., to identify the presence and transmission oftransgenic genes in
hybrids of transgenic crops with non-target plant species. Specific examples of these and
other molecular methods can be seen in Sambrook and Russel (2001), Cevera et al (2000),
Denise et al. (2002), Dionsis et al. (2002), Hardegger et al. (1999), Templin et al. (2002)
and Webster et al. ( 2002).
B. Gene flow studies
The general approach that will be used in greenhouse and growth chamber studies and to a
lesser extent in field studies will be to create constructed communities of potential donors and
5-9
-------
compatible recipients of transgenic genes. Donor plants may consist of transgenic crops and F1 or
subsequent hybrid or backcross generations between the initial transgenic crop and compatible non-
crop species. The non-crop species may consist of weedy or native plants; they may also be
represented by other cultivars of the primary donor transgenic crop, or of other cultivated crops with
which the donor crop may be compatible e. g., canola (Brassica napus), is compatible with other
Brassica species and also with wild radish (Raphanus raphanistrum) arid with cultivated radish
{Raphanus sativa)\ creeping bentgrass (Agrostis stolonifera) is compatible with other Agrostis
species (Wipff, 2002), and with rabbitfoot polypogon (Polypogon monspeliensis ). It thus may be
necessary to do analyses in a number of potential recipient plant species to determine the presence
and stability of transgenic genes that have originated from the original transgenic crop source or
from subsequent transfer of that gene to other crop and non-crop plants.
Specific questions to address gene flow are:
1. What is the potential for gene flow from crop to other crop and non-crop plants?
2. What is the geographic proximity of the crop to compatible wild and cultivated relatives
and their respective times of flowering?
3. Gene flow occurs between transgenic crops and non-transgenic crop or non-crop plants,
weeds, native, rare and endangered species?
4. Can F1 hybrids and back-cross progeny of F1 progeny x each parent, formed with crop,
or non-crop species and transgenic plant escapes persist, spread, outcross and produce
viable, fertile seed?
5. How long will DNA from the transgenic crops or their hybrids with weedy or native
species persist in the environment and remain biologically active?
C. Greenhouse/growth chamber/field studies to measure potential ecological effects of
gene flow.
The methods we will use at WED to assess the ecological effects of gene flow will be
focused primarily on effects on plant fitness and on plant community structure. Depending on the
5-10
-------
specific nature of the transgenic genes, which are being evaluated, additional specific types of
measurements and assays will need to be carried out. For example, via extra-mural collaborative
studies with academic and other federal or state agency collaborators, data could be obtained to
determine effects of gene flow from crop to weedy or native species, on insect community structure,
on beneficial pollinators and on predators of targeted insect pests. Similarly, if the transgenic gene
being studied has been developed for disease resistance to a specific pathogen, it would be of
interest to examine potential effects of presence of the transgenic gene to responses to other plant
pathogens or to symbionts.
Specific approaches/questions to be addressed in the greenhouse, growth chamber and field
studies are:
1. Identify, obtain and/or create hybrids between crops and weedy or native species.
2. Assess changes in herbaceous and woody plant fitness characteristics in response to
gene flow, specifically effects on developmental endpoints, vegetative biomass and seed
yield, viability, fertility and dormancy.
3. Sample transgenic donor and recipient compatible plants from plant growth facilities
on-site and from field plots at locations of public and private sector collaborators, for
the presence of the transgenic genes and for effects on fitness of the recipient plant
species. Plants and soil within and beyond donor source plots may also be of interest to
sample to determine the presence and persistence and transport of transgenic DNA.
4. Detect and quantify the presence of transgenic genes in transgenic plant sources and
potential compatible crop, weed and native plant species recipients. Analyze
greenhouse, environmental chamber or field samples using Sybr Green or other
quantitative PCR method (Heid et al., 1996). Compare results obtained with Sybr green
to those obtained using commercially available engineered trait testing materials or
information available in the literature, for detection of e. g., herbicide or disease
resistant traits of interest.
5-11
-------
5. Using model constructed plant communities in open top chambers and in controlled
environment chambers and in field tests, test the effects of population density and
spatial arrangements of donors of transgenic genes on gene transfer to compatible hosts
(see Fig. 5.4). Donors may be the primary transgenic crops or F1-F4 hybrids or back
cross (BC1-BC4) progeny containing the transgene of interest. Additional test
conditions may include optimal and sub-optimal moisture, soil fertility, and temperature
regimes, as well as application of selective pressures provided by chemical herbicide,
disease inoculum or insect populations.
D. Field studies of potential ecological effects of gene flow
1. Using constructed plant communities in contained environments (growth chambers and
greenhouses or fine-screened field plots), and containing Fl, F2 and/or back-cross
progeny between the designated transgenic crop/trait(s) of interest and non-target
weedy, native or compatible crops, assess transgenic gene flow between and among
potential transgenic crop donors and crop and non-crop recipients. Study ecological
parameters of interest, such as those indicated below, over a multi-year period.
2. Determine effects of gene flow on plant biomass, numbers, biomass and germination of
seeds, fertility of hybrids and changes to plant community composition, re: frequency
and abundance of given species, hybrids, etc.
3. Depending on the engineered trait(s) chosen for specific studies (i.e., whether
insecticidal for above or below ground insect or disease pests, or resistance to an
herbicide), apply specific biological pest or pesticide (e.g., herbicide) selective
pressure(s). Next examine potential ecological responses for above ground and below
ground biota such as beneficial invertebrates and microbes. Ecological parameters to
consider in examining those non-target populations include abundance, diversity and
metabolic functioning.
4. Given the expected interactions between genotype and environment, long-term efforts
should be made to measure gene flow and its potential impacts under a variety of
environmental conditions in the field. Factors to be evaluated may include different
5-12
-------
levels of soil fertility, moisture, climates and soil pollutants such as persistent
agricultural chemicals, e.g., herbicides, fungicides, insecticides which may confer
selective pressures over an extended period of time.
5. Evaluate the effects of cultural practices, including crop rotation (particularly to a
crop(s) containing the same transgenic gene), tillage methods, and pesticide use, on the
short-term and long-term ecological effects of gene flow from transgenic crops
6. Evaluate the effects of soil fertility on crop-weed hybrid establishment, survival seed
production, out-crossing potential and non-target ecological effects by (a) creating a
series of compatible transgenic crop-weed or crop-crop hybrids; plant under optimal
agronomic and low fertility conditions; assay for establishment, biomass, seed
production, seed germination, fertility, over-wintering survivorship, and seed dispersal
distances; assay for effects on non-target beneficial organisms, pests and wildlife.
E. Inputs for prototype model
Development of a probabilistic, regionally based, ecological risk assessment model is of
interest to provide predictive information on the likelihood of adverse effects related to gene flow
from transgenic crops. Based on empirical data obtained from greenhouse and field studies on
potential impacts of hybridization with transgenic crops, key parameters to include in model will be
identified. The model will be run for one key crop/trait combination. For example, effects of gene
flow from a herbicide, insect or disease resistant canola or grass crop on the diversity, functioning
and biomass of plant, insect or microbial communities. Parameters to consider in developing the
model include the probabilities of finding compatible crop and weeds in proximity or within
hybridization range (via pollinating insects or wind); overlaps of flowering times and other
similarities in phenology; seed number, biomass, and germination; fertility and out crossing potential
of resultant hybrids and the abundance (population density) of compatible donor and recipient
populations. Impacts of cultural practices (types of tillage, pesticide use and crop rotations), soil
fertility and climate on survival of crop-weed hybrids and feral transgenic plants also will need to be
considered.
5-13
-------
1. Generate empirical data from contained (growth chamber or greenhouse) and field tests
with model system(s), e.g., one crop, one or more compatible weeds or native species,
one or more transgenic pesticidal traits.
2. Identify parameters to include in model, e. g., rates of out crossing from crop and from
compatible weedy or native species; plant population density, seed number, plant and
seed biomass, viability; fertility, over-wintering survival, seed dormancy and lateral
spread of crop-weed or crop-native hybrid plants (F1-F4) and of backcross generations.
3. Obtain effects data with and without relevant selective pressures of disease, insects,
agricultural chemicals (herbicide, insecticide, fungicide).
4. Obtain field data on gene flow (exposure) and ecological effects from one or more
typical crop rotations with selected crop, e. g., alfalfa, wheat, oats or barley, after GMO
canola.
5. Run model and compare predictions to empirical data.
F. Additional research to consider
Examine the influence of the following parameters on rates of gene transfer and
ecological effects:
• nuclear vs. organelle (chloroplast or mitochondrial) inserts
• different sites of insertion, e.g., within different genomes of allopolyploids, minimizing
outcrossing to compatible wild relatives
• use of different promoters with different times, places and levels of expression
• inducible vs. constitutive expression of genes
• single vs. multiple engineered traits; linkages; interactions
Expand Crops/Traits/Weeds of interest beyond Northwest/ex-U.S. (e. g., Canada,
Mexico, western Europe):
5-14
-------
¦ crop protection (herbicide, insect, disease resistance)
¦ crop quality (nutrient) traits
¦ newer generation traits: specialty chemicals, pharmaceuticals
Use genomic and proteomic analytical methods to identify molecular markers of plant
fitness and development.
Evaluate effects of transgenic genes, crop pests, environmental factors and agricultural
chemicals on plant gene expression.
Aspects of plant development that may be impacted by biotic and abiotic factors include
pollen and seed formation and viability, shoot and root biomass, ability to overwinter and responses
to symbiotic and pathogenic organisms. By using microarray approaches (Figure 5.5), identify
changes in gene expression between conventional crops or weedy or native species with transgenic
crops and with crop-weed hybrids that contain the gene of interest. Examination of responses of the
conventional, transgenic and crop-weed hybrids at various developmental stages and under a variety
of environmental conditions, including exposure to conventional agricultural chemicals, is of interest
to identify potential plant molecular markers which may be diagnostic indicators of the presence of
the transgenic gene or of exposure to certain environmental conditions, including exposure to
chemical pesticides. Additional molecular markers that may be of interest include those which
signify potential changes in the diversity and functions of microbial communities in soil that are
involved in N cycling, i.e., using primer sets diagnostic for structural genes involved in nitrification
(amoA), denitrification (nir), or nitrogen fixation (nifH). Information on potential changes in soil
microbial communities involved in nutrient cycling are of interest, given the key roles of soil
organisms in decomposition, soil fertility and plant nutrient availability
G. Specific research proposed
Sections A through F above describe the breath of research that should be addressed in a
comprehensive gene flow project. Based on available resources the gene flow research will be
limited at least initially, to research on two crops. Bent grass (Agrostis stolonifera) which is wind
pollinated, and canola (Brassica napus), which can be both wind and insect pollinated, have been
5-15
-------
selected as potential sources of transgenic genes to use in our gene flow studies. Other factors (see
Table 5.1), in our choice of these crops include:
¦ Availability of essentially completely sequenced genomes for closely related plants, e. g.,
rice and Arabidopsis, which are in the same families Gramineae and Brassicaceae, as
Agrostis and Brassica respectively;
¦ Presence of compatible crop, weedy and/or native relatives with which each can hybridize,
both locally here in the Pacific northwest as well as nationally and internationally;
¦ Each of the selected transgenic crops will soon be or have been approved for commercial
release.
The transgenic trait which currently is/will be commercially available is herbicide tolerance,
a trait which is of agronomic interest and is an easily selectable marker to detect progeny and feral
plants containing the transgenic gene. Other transgenic traits which may be available for research,
but which are not yet commercially available for either of these crops, are ones that confer tolerance
to bacterial or fungal diseases or to insect pests. Our proposed research strategy with each of these
crops can be summarized as follows:
1. Determination of Gene Transfer Rates/Obtain /Produce Hybrids
a. Select specific compatible crop/weed or crop/native species of interest in particular
geographies of interest using GIS based crop, weed, chemical use and plant systematics information
(Fig 5.1).
b. Determine gene detection methods needs (Fig. 5.2). Are marker genes available? Is that
information proprietary/can it be accessed? Will we need to develop our own markers, based on
genomics research (Fig. 5.3)?
c. Obtain, find or make hybrids between the crop and weedy or native species of interest that
are most likely to be found within the same geographic area and which would flower at
approximately the same time (Fig 5.2).
5-16
-------
d. Use the transgenic parents, their hybrid (F1-F4) progeny with selected species and backcross
progeny (BC1-BC4), as sources of transgenic genes in controlled environment studies in growth
chambers or greenhouses or screen houses in the field. Determine and compare rates of gene
flow in reciprocal crosses, in the presence and absence of selective pressures appropriate to the
selected transgenic herbicide, disease or insect tolerant trait (Figs. 5.2, 5.3 and 5.4).
e. Compare rates of transfer from transgenic sources in which the trait of interest is located in
different locations on nuclear or organelle genes. If multiple nuclear genomes exist, assess
rates of transfer from different genomes and chromosomes within those genomes (Figs. 5.2 and
Fig. 5.3).
2. Evaluation of Hybrid Fitness and Ecological Effects
a. Compare ecological effects of gene flow on fitness of progeny at the plant community,
population and individual species levels. Examples of effects are, vegetative and seed biomass
production, germination, fertility of progeny, over-wintering survival, survival in the soil seed bank;
effects on plant community composition.
b. Determine ecological effects of gene flow and related fitness changes on beneficial
insects and crop pests, invertebrate herbivores; soil food web biota; and vertebrate herbivores and
invertebrate and vertebrate seed predators (Fig. 5.3).
3. Develop a Probabilistic Risk Assessment Model of the Ecological Effects of Gene Flow
a. Use exposure and ecological effects information to provide inputs for a probabilistic risk
assessment model (Fig 5.1), based on above-ground effects on plants, beneficial insects, insect pests
and vertebrate herbivores at community, population and species levels.
b. Incorporate inpu ts (as available) on effects of changes below ground, on beneficial and
pathogenic soil foodweb organisms.
5-17
-------
5.5 Time Line
Table 5.2 describes identification of technical resources (plants, traits, and contacts),
formulation of the research plan, hiring of molecular biology post-doc and definition of pilot studies
in year one (FY 2002). In FY 2003, the research and QA plans are approved and laboratory,
greenhouse and growth chamber studies are to be initiated. Pending availability of funding,
cooperative agreements and inter-agency agreements will also be implemented in FY 2003. In
addition to continuing intramural and extramural research with domestic collaborators, international
collaborations may be implemented in FY 2004. An all investigator meeting will be convened in FY
2005 to review research progress, e.g. on gene tracking methods and on ecological effects on plant
community composition in non-agronomic ecosystems. An international workshop may also be
convened in the latter part of FY 2005 or early FY 2006, to call for the development of an
international data collection network and potential risk assessment guidelines for evaluation of non-
target effects of gene flow from transgenic plants. Based on results obtained through FY 2005, in
FY 2006 it is anticipated that inputs for a probabilistic risk assessment model of ecological effects of
gene flow on non-target plant populations will have been identified, so that a prototype model can be
run.
5-18
-------
Table 5-1 Evaluation Criteria and Ranking of Northwest Crops/Traits of Interest
Crop Candidates
Compatible Weedy,
Native, Crop
Species in Pacific
Northwest
Engineered Traits
Commercially
Available1 or in
Development2
Technology Access
and Genomics
Information
Canola
Grass Seed/Wheat
Raspberries
Numerous Weedy
and Crop Mustard
and Radish species,
Brassica Vegetable
Crops
numerous; jointed
goat grass
Himalayan
Blackberry
Herbicide Tolerance1
Fungal Disease
Resistance2
Insect Resistance
Herbicide Tolerance
Virus resistance
Herbicide tolerant
canola commercially
available in Canada,
parts of northwest,
upper midwest;
Arabidopsis genome
published; canola
(Brassica) is also a
cruciferous plant
species; commercial
microarrays
available for
Arabidopsis
Grower concerns
with GMO traits
limiting sales to
Japan and western
Europe are causing
delay in commercial
introduction; recent
publication of
genomes of two rice
species could
enhance genomics
studies with wheat
and other grass
species (bent grass,
rye grass) being
considered for use in
Pacific Northwest
research at early
stage; gene
expression not yet at
commercial level
5-19
-------
Table 5-2 Timeline/Outputs Gene Flow Research.
FY 2002
FY 2003
FY 2004
FY 2005
FY 2006-2007
Review
Contribute to
Continue R &D
Continue R&D
Complete:
Literature
APM on
-Intramural
-Intramural
Short-Term
Strategy for
Studies in:
-Extramural
R&D
Identify
Updated Test
Laboratory
Resources:
Guidelines:
Continue:
-Plants, Traits,
Finalized
Chambers
Long-Term
People,
Research Plan
R&D
Organizations
Field
Approved QA
Model Inputs
Convene
Produce:
Attend
Plan in Place
Meeting of
Protocols
Workshop:
Initiate extra-
Investigators to
Publications
- Identify Data
Initiate Lab and
Mural R&D
Review:
Test Model
Gaps
Chamber
via:
-Findings
-Select
Studies With
-Coops
-Methods
Crops/Traits,
Transgenic and
-IAGs:
-Problems
Agency Reports:
& Geographies
Parental Plants
USDA-ARS
-Define Model
-Findings
DOI-NPS
Parameters
-Methods
Draft Research
Initiate DNA
DOI-BLM
-White Paper
Plan
Analyses of
-Contracts
on Strategies to
Plant and Soil
-CRADAs with
Minimize
Pilot On-Site
Materials from
Private Sector
Convene
Gene
Studies:
Field Sites in
Workshop to
Flow Effects
-DNA
US region(s) in
Identify
Develop
Characteristics,
which selected
Collaborators for
National and
Persistence,
crop(s) are
International
International
Expression,
grown
Ecological
Data Collection
Transforming
Effects and
Network
Ability
Hire NRC
Post-doc
Molecular
Tracking
Hire NHEERL
Collaborations
Post-Doc
Formulate RFA
Issue RFA
in Multi-Year
Field Studies
Identify
With a Wind and
Potential IPA
With An Insect
(academic and
Identify US and
Pollinated Crop
federal agency)
International
and GSF
Collaborators
collaborators
5-20
-------
Allergenicity
Probabilistic Risk
Assessment Models
J
Exposure
Gene Tracking Methods
—*
Detection
—
Quantification
Compatible Species
Pollen/Seed
GM vs Conventional
m
Effects
L
Plants
Communities
Populations
Individuals
Figure 5-1 Overview of exposure and effects research needed for risk assessment of genetically modified (GM) crops
5-21
-------
Gene Detection
Quantification
Exposure
GM vs Conventional
Molecular Markers
Molecular Methods
Gene Tracking Methods
Compatible Crops
and Weeds
Pollen and Seed of Crops,
Crop/Weed Hybrids,
Feral Crops/Crop Spills
Choose Agroecosystem
Select Model Plants/Genes
Design Lab/Field Studies
Gene Transfer Rates and Mechanisms
With/Without Selective Pressures
Pollen vs Seed Dispersal Distances/Mechanisms
Multiple/Stacked Gene Sources (Crops/Hybrids)
Persistent Seed Banks
Figure 5-2 Criticalpath for exposure component of gene flow risk assessment.
5-22
-------
Plants
Individuals
Populations
Effects
Communities
Changes in Gene Expression
Changes in Phenology
Changes in Fecundity
Changes in Community Composition:
Species Richness, Diversity, Abundance
Percent/Type Cover
Habitat Type/Quality
Changes in Stress Tolerance (Abiotic/Biotic)
Changes in Plant/Seed Biomass, Nutrient Quality
Changes in Invertebrate/Vertebrate Pollinators
Changes in Invertebrate/Vertebrate Herbivores
Changes in Below Ground Biota/Biomass
Figure 5-3 Critical path for ecological effects component of gene flow risk assessment.
5-23
-------
© Transgenic
O Non-Transgenic
Figure 5-4 Illustration of production of transgenic patches of varying population density which may result from gene flow or
from incidental transport of feral transgenic seed.
5-24
-------
Microarray Approach
Range of Microarray Applications
Treated Sample
Control
Isolate
mRNA
Label with fluorescent dyes
using reverse transcriptase
and labeled nucleotides
cDNA
Combine in equal amounts
I
Scan and Analyze Microarray
-1 tj -...»: V --
11 • ¦ y. vfir.~> t-
, **^l "r7 ¦ r **¦ *. * .• i * <.* *.v
-: -v.- fe'.t: h>—>
;v^:;r,
¦F*
Control sample only
Treated sample only
RNA - Snapshot of short term
condition of organism (hours).
PROTEIN — Longer term
phenotypic response (days).
DNA — Detection of genotype
changes in species and populations
Hybridize to Microarray of nucleotide
sequences on a glass slide
Both (no difference)
Color in the microarray indicates which treatments
affect expression of specific genes
Figure 5-5. Examples of microarray methods including approaches and
applications.
5-25
-------
6 Outputs and Performance Measures
The major outputs of this project will be tools to carry out ecological risk assessments
(Figure 2.2).
¦ The Regional Analysis and Interpretation component will develop databases useful for
both chemical herbicide and gene flow studies including a spatial database of potential
pesticide exposure to non-target plants, a database of pesticide effects on plants, and a
database of crop planting dates, pesticide use dates and weed emergence; as well as
several case studies to identify and prioritize potentially important uncertainties.
¦ The Effects of Chemical Herbicides on Terrestrial Plants component will contribute
refined plant testing methodologies for herbicide exposures, experimental conditions, and
response endpoints for risk assessments. The methodology can be used to develop new
individual plant tests and multispecies ecological tests for herbicide responses. The
chemical herbicides and terrestrial plants area will also include mode of action and
molecular effects tools for future development.
¦ The Ecological Effects of Gene Flow from Transgenic Crops component will
contribute gene tracking methodology which will be used to develop metrics of potential
ecological effects of gene flow from GM-crops to compatible native, weedy and crop
species in agronomic and non-agronomic ecosystems.
Project outputs will be in the form of scientific papers, protocols, data sets, and a GIS
platform. These outputs will meet the objectives of the Government Performance and Results Act
(GPRA) by contributing to a series of Annual Performance Goals (APGs) which are major
achievements across ORD laboratories and Annual Performance Measures (APMs) which are
specific milestones at the Division Level and contribute towards the APGs. Table 6.1 indicates the
APGs and APMs for this project for 2003 through 2008. These are also shown in the project critical
path (Figure 2.2) and individual time lines for the regional analysis (Table 3.1), chemical herbicide
(Table 4.5), and gene flow components of this project (Table 5.2). The 2003 APM will be met
primarily by the publication and implementation of this research plan. All three areas of research,
6-1
-------
regional analysis, chemical herbicides, and gene flow will contribute to this APM. The APMs for
2004 and 2005 will concern risk assessment and regional approaches to species selection and risk
assessment for chemical herbicides. They will include contributions from the regional analysis and
chemical herbicide areas of this project and, and will be based on peer reviewed manuscripts,
databases, presentations and other outputs. The APMs for 2006 and 2007 concern protocols for
improved tests of plant effects from chemical herbicides. They will be based on the chemical
herbicide research. A possible APM for 2008 (beyond the five year timeline for this project) would
integrate outputs from all three areas of research, and take the form of a variety of regional
ecological risk assessment tools including models and a GIS platform.
6-2
-------
Table 6.1 Annual Performance Measures and Goals for the Pesticide Research Project
APG 19
Develop improved tools and models to assess and predict human health and ecological risk from
exposure to commercial chemicals and microorganisms. Attainment of this goal will include the
successful completion of the suite of annual performance measures.
Reporting. ALD Contact: Tim Gleason
APM 208
Strategy for research to update herbicide testing guidelines and to evaluate gene flow
Contact: David Olszyk (WED)
2003 NHEERL/WED/RCB
APG 97
Provide an improved capability to assess the ecological risks associated with high potency
herbicides.
Reporting. ALD Contact: Jack Fowle
APM 167
Evaluation of Risk Assessment Methods for Herbicides
Contact: E. Henry Lee (WED)
2004 NHEERL/ WED/RCB
APM PROPOSED
Guidelines for regional approach to selection of plant species for herbicide risk assessment
(WED)
2005 NHEERL/WED/RCB
APM PROPOSED
Draft revised protocol / guidelines for vegetative vigor test with crops and selected native plants
2006 NHEERL/ WED/ RCB
APM PROPOSED
Draft new protocol / guidelines for reproductive/developmental endpoints with annual species
2007 NHEERL/ WED/RCB
APM PROPOSED
Refined regional assessment tools for assessing risk to plants from herbicides and gene flow based on
GIS framework and probabilistic risk assessments
2008 NHEERL/ WED/ RCB
6-3
-------
Intentionally Blank Page
6-4
-------
7 Project Management and Quality Assurance
7.1 Management Responsibilities
This project is coordinated with other projects within WED that address other EPA Goals
(e.g,. Goal 8, Sound Science), and with projects in other NHEERL Divisions which address other
aspects of Goal 4 research (e.g. health related questions). Coordination is necessary not only to
make sure that key agency needs are addressed without repetition among Projects and Divisions and
that resources (staff, funds) are used effectively, but to make sure that scientific information and
ideas can be "cross-fertilized" among scientists working in different areas. Line management to
make sure the project achieves its goals is as follows:
A. NHEERL/Associate Laboratory Director. The WED Pesticides Project is part of a
coordinated NHEERL effort to characterize human health and ecological effects of pesticides.
Within NHEERL the multi-year plan is being developed to elucidate the rationale and overall
objectives for this research. The Associate Laboratory Director for pesticides (Goal 4) related
research is Dr. Jack Fowle, located at the RTP headquarters of NHEERL.
B. Branch Chief. At WED the Project is managed within the Risk Characterization
Branch and comes under the general responsibilities of the Ecosystem Characterization Branch
Chief. The Branch Chief is responsible for ensuring that the Project research meets the objectives of
Goal 4 and that all technical outputs meet the quality requirements of the Division, Laboratory and
Agency. The Branch Chief also is the direct line manager to the Project Leader, and can apply
Branch resources to resolve project issues. The Branch Chief is responsible for coordinating the
WED Goal 4 research with the research of other NHEERL Divisions through the ALD and the
Resource Characterization Team (RCT). Dr. Anne Fairbrother is ECB Branch Chief.
C. Project Leader. The Project Leader is management's principle contact with the
Project and is responsible on a day-to-day basis for the performance of, and coordination of, research
within the Project. The Project Leader works with the Branch Chief and Principal Investigators on a
collaborative basis to accomplish those goals. The Project Leader also is a member of the WED
Science Council. Dr. David Olszyk is Pesticide Project Leader.
7-1
-------
D. Principal Investigators. The Principal Investigators (Pis) are responsible for the
scientific questions being addressed by the three main areas within the project (Figure 1), and for
managing day-to-day decisions concerning the research within their area. They also are responsible
for managing resources (e.g. EPA and Senior Environmental Employee/National Asian Pacific
Center on Aging (SEES), Level of Effort (LOE) contract, purchase orders) to accomplish their area's
research goals. They serve as Work Assignment Mangers for the WED LOE contract, advisors for
postdoctoral associates, mentors for interns and other scientific and technical staff. Pis and their
areas of research within the project are as follows (note that all PI scientists may collaborate in all
three areas of the project on a limited basis):
¦ Effects of Chemical Pesticides on Terrestrial Plants, Dr. Thomas Pfleeger, Dr.
John Fletcher, Dr. David Olszyk
¦ Ecological Effects of Gene Flow from Transgenic Crops, Dr. Lidia Watrud
¦ Regional Analysis and Integration, Dr. Henry Lee, Dr. Thomas Pfleeger
Based on resources we will add staff and/or work collaboratively with other ORD and EPA
organizations to develop the risk assessment tools.
E. Project Scientists. Under the direction of the Pis, the project Scientists work directly
on the research in a specific area and may be from a variety of organizations including the EPA, the
SEES staff, LOE Contract, guest workers and others. The Project Scientists' assist in the
development and implementation of experimental protocols, conduct the research, process
experimental data, and assist in the production of scientific documents (journal papers,
presentations, briefings). The Project Scientists follow Project Quality Assurance and Health and
Safety protocols.
7.2 Communications
The Pesticides Project is a sophisticated, very complex, multi-discipline research endeavor.
Completing the Project successfully will require continual communication among project
participants, at all levels. Communication will be fostered by regular weekly meetings of the Pi's to:
(1) coordinate sampling and various experimental activities, (2) exchange data and information
7-2
-------
about the various tasks, (3) share scientific information, and (4) refine and/or modify the Research
Plan, the QAPP and the SOPs. The Pis and other Project participants will meet as necessary to: (1)
coordinate experimental activities (e.g., availability of equipment, harvests, etc), (2) exchange
information, and (3) resolve possible problems. For work conducted through the on-site Level of
Effort Contract (LOE), technical direction will be from the EPA Work Assignment Manager to the
Contractor Work Plan Manager.
The Project Leader will meet regularly with the Branch Chief to keep her informed on the
status of the research. The Project Leader and Branch Chief will participate in WED Science
Council meetings to insure that the Pesticides Project research is coordinated with other WED
research efforts. Contact will be made on a regular basis through appropriate levels (e.g. P.I.,
Project Leader, Branch Chief, Associate Director for Science, Division Director) to insure that the
research responds to agency Goal 4 needs and is coordinated with other NHEERL and ORD research
efforts. Finally, the Pis and other project scientists will be active participants in scientific meetings
relating to non-target effects of chemical pesticides and GM crops, including hosting a scientific
workshop in those areas, tentatively in year three of the project.
If the full complement of resources is not available, this plan will be implemented to the
extent feasible with available resources to achieve the most critical of all the important goals of the
project. For example, the plant effects from chemicals research may be limited to developing an
improved vegetative vigor test for annual plants if funds are severely restricted.
7.3 Quality Assurance (QA)
In order to produce reliable data of known quality and to meet EPA and WED QA
requirements this project has a QA Project Plan. In developing the Project's QA organizational
structure to meet the QA goals, five essential QA/QC elements were addressed: (A) QA/QC
responsibilities and research responsibilities, (B) communications, (C) document control, including
the importance of standard protocols for the experiment especially Standard Operating Procedures
(SOPs) for experimental data collection. In addition, we have separate QA issues to consider
specific to the chemical herbicide (D) and gene flow research (E).
7-3
-------
A. QA and Research Responsibilities
WED management and research staff share responsibility for implementing the Laboratory's
QA policies, and they are accountable for those aspects of QA/QC associated with their work areas.
The QA Responsibilities in this Quality Assurance Project Plan (QAPP) were derived from Section
1.0 of the US EPA, NHEERL, Western Ecology Division Plan (U.S. EPA 1995). The general
Project QA/QC organizational structure can be described as follows:
Division Director. The Division Director has ultimate responsibility for all research
conducted, funded, or managed within the division. They must approve the division Quality
Management Plan . Within the Office of the Division Director, the WED Quality Assurance
Manager is responsible for ensuring that all WED QA activities are in compliance with agency
QA policy and guidance. He/she reports to the Associate Division Director for Science.
Branch Chief. The RCB Branch Chief is responsible for the quality of all research
conducted, funded, or managed within the Branch and for ensuring that all technical outputs meet
the quality requirements of the Laboratory and Agency. The Branch Chief also is the direct line
manager to the Project Leaders, and can apply Branch resources to resolve QA issues. The Branch
Chiefs key QA responsibilities include: a) review and evaluation of work on QA implementation
and progress, b) review the quality of outputs generated by each project, and c )review and evaluate
audit and performance evaluation reports.
Project Leader. The Project Leader is responsible for production of the QA Project Plan
(QAPP) and oversees all QA management aspects of the project. The Project Leader determines
quality criteria based on the intended use of the results to be generated, and communicates these
criteria to the Project participants. The Project leader conducts periodic reviews of QA procedures
and data gathered with them within the project and writes a report and implements QA procedure
changes if necessary based on those reviews.
Principal Investigators. The Pis are responsible for carrying out specific research areas
within the project and for insuring the quality of the results generated by those areas. They approve
specific SOPs and other QA documents relative to their areas.
7-4
-------
Project Scientists. Project Scientists work directly on the Project's research and QA/QC
procedures. The Project Scientists' key QC/QA responsibilities include: a) assist in writing SOPs as
necessary, b) implementation of the SOPs, c) evaluation and documentation of QA methods used for
measurements.
B. Communications:
The periodic project meetings will be used to refine and/or modify the Research Plan, the
QAPP and the SOPs. The Pis and other Project participants also will meet as necessary to: (1)
coordinate experimental activities (e.g, availability of equipment, harvests, etc), (2) exchange
information, and (3) to resolve possible problems.
C. Document Control
The QAPP and experimental protocols define the key aspects of the Pesticides Research
Project QA program, consequently, it is important that all Project participants have access to these
documents. The Project Leader will be responsible for maintaining the original signed copies of the
QAPP and approved SOPs and any other QA documents. These will be kept in room 203 of the
Main Building. Paper copies of the Research Plan and QAPP will be available for Principle
Investigators. In order to minimize paper copies, electronic versions of the Research Plan, QAPP
and all other QA documents will be available to the Principal Investigators and all other Project
participants on the server: NABU/Pesticides/EPA/QA. It is the responsibility of individual Project
participants to print out paper copies of the documents required for their work. The Project Leader
will ensure that version numbers of the approved QAPP and each approved SOP, EP or OP are
correct and changed as necessary. If the QAPP or a SOP is revised, the Project Leader will ensure
that relevant members of the project are notified. At that time any previous version of the document
is to be discarded by the user. However, copies of older versions of SOPs and other QA documents
shall be retained by the Project Leader in the files in room 203 of the Main Building in order to
document collection procedures prior to the current versions of SOPs
Although the QA/QC elements described above are highly interdependent to successfully
execute the Project's research and QA programs, SOPs have an especially critical role in these
7-5
-------
programs. A SOP is the keystone element upon which experimental procedures will be based, and
Management and Project personnel will interact in the other four elements of the Project QA
program. Because the Project's QA organizational structure depends heavily on SOPs, their format
was structured to serve as guidelines for all Project personnel to accomplish both the scientific and
QA/QC procedures of the. SOPs should follow the same format as the QAPP, modified as needed.
In addition to describing methodology for collection of data, SOPs include information as to
processing of data and location of files and/or databases. Statistical analysis procedures will be
documented in descriptions for individual experiments and/or in methods sections for manuscripts.
The SOPs will be reviewed periodically, and re-authorized as necessary
D. Special Health and Safety Considerations for Chemical Herbicide Studies.
Since all active ingredients used in the chemical herbicide studies are considered to be toxic
materials, their purchase, storage, use, and fate must be accounted for in the Project Health and
Safety Plan. This plan also describes use of the track sprayer for plants in pots grown in
greenhouses, growth chamber, or an outside nursery area. It will be modified to include new
herbicides or field plot herbicide treatments as specific experiments are designed.
E. Special QA Considerations for Gene Flow Studies.
The gene flow research is developmental, involving evaluation and testing of new procedures
for which there may be no existing protocols. Nevertheless, all researchers shall be required to
accomplish the Quality Assurance procedures described in existing and new SOP's in order to
comply with EPA's Quality Assurance Program. Researchers will comply with all relevant current
SOP's and provide new SOP's as required for the execution of specific tasks. QA support will be
provided for plant and microbial growth facilities (e. g., incubators, environmental chambers),
coolers, freezers, ovens, autoclaves, pipettors, balances, reverse osmosis (R. O.) water supplies. A
waiver for exploratory and preliminary research used in the development of skills in the WED
Pesticides Project was obtained as described in the Quality Assurance Request for Exemption memo
of November 7,2002.
7-6
-------
8 REFERENCES
Al-Khatib, K., G.I. Mink, G. Reisenaurer, R. Parker, H. Westberg, and B. Lamb. 1993.
Development of a biologically-based system for detection and tracking of airborne herbicides.
Weed Technol. 7:404-410.
Al-Khatib, K., and D. Peterson. 1999. Soybean (Glycine max) response to simulated drift from
selected sulfonylurea herbicides, dicamba, glyphosate, and glufosinate. Weed Technol. 13:264-
270.
Altman, J., and A.D. Rovira. 1989. Herbicide-pathogen interactions in soil-borne root diseases.
Can. J. Plant Pathol. 11:166-172.
Altman, J., and C.L. Campbell. 1977. Effects of herbicides on plant diseases. Ann. Rev.
Phytopathol. 15:361-385.
Bergelson, J., C.B. Purrington, and G.Wichmann. 1998. Promiscuity in transgenic plants. Nature
395:25.
Beyer, E.M., M.F. Duffy, J.V Hay, and D.D. Schlueter. 1988. In: Herbicides-Chemistry,
Degradation and Mode of Action. Kemy, P.C., and D.D. Kaufman, Eds. Marcel Dekker. New
York, pp.117-189.
Bhatti, M.A., K. Al-Khatib, A.S. Felsot, R. Parker, and S. Kadir. 1995. Effects of simulated
chlorsulfuron drift on fruit yield and quality of sweet cherries. Environ. Toxicol. Chem.
15:537-544.
Bird, S.L., S.G. Perry, S.L. Ray, and M.E. Teske. 2002. Evaluation of the AgDISP aerial spray
algorithms in the AgDRIFT. Environ. Toxicol. Chem. 21:672-681.
Boutin, C., and B. Jobin. 1998. Intensity of agricultural practices and effects on adjacent habitats.
Ecol. Appl. 8:544-557.
8-1
-------
Boutin, C., H.B.Lee, T.E. Peart, S.P. Batchelor, and R.J. Mcguire. 1999. Response of five plant
species sprayed with sublethal doses of metsulfuron methyl. British Crop Protection Council
Brighton Conference Proceedings; Farnham, United Kingdom.
Boutin, C., and C.A. Rogers. 2000. Pattern of sensitivity of plant species to various herbicides- an
analysis with 2 databases. Ecotoxicol. 9:255-271.
California Department of Pesticide Regulation. 2000. Pesticide Use Reporting: An Overview of
California's Unique Full Reporting System, Sacramento, California.
www.cdpr.ca.gov/docs/pur/purovrvw/tabofcon.htm.
Cervera, M. T., D. Remington, J.-M. Frigerio, V. Storme, B. Ivens, W. Boeijanand, and C. Plomion.
2000. Improved AFLP analysis of tree species. Can. J. For. Res. 30:1608-1616.
Chaboussou, F. 1986. How pesticides increase pests. The Ecol. 16:29-35.
Chambers, J.E., J.S. Boone, R.L. Carr, H.W. Chambers, and D.L. Strauss. 2002. Bioindicators as
predictors in health and ecological risk assessment. Human Ecol. Risk. Assess. 8:165-176.
Cook, R. B., G.W. Suter II, and E.R. Sain. 1999. Ecological risk assessment in a large river-
reservoir: 1. Introduction and background. Environ. Tox. Chem. 18:581-88.
Cormier, S.M., M. Smith, S. Norton, and T. Neiheisel. 2000. Assessing ecological risk in
watersheds: a case study of problem formulation in the big Darby Creek Watershed, Ohio,
USA. Environ. Tox. Chem. 4:1082-96.
Dale, P., and J.A. Schefifler. 1996. Gene dispersal from transgenic crops. Transgenic organisms and
biosafety. In Transgenic organisms and biosafety: horizontal gene transfer, stability of DNA,
and expression of transgenes, eds. E.R. Schmidt, and T. Hankeln, pp. 85-93. Springer-Verlag.
Berlin.
Dahlquist, K. D., N. Salomonis, K. Vranizan, S.C. Lawlor, and B.R. Conklin. 2002. GenMAPP, a
new tool for viewing and analyzing microarray data on biological pathways. Nat. Gen.
31:19-20.
8-2
-------
Denise, D., M. Dalma-Weiszhausz, E. Chicurel, and T. R Gingeras. 2002. Microarrays and genetic
epidemiology: A multipurpose tool for a multifaceted field. Gen. Epidemiol. 23:4-20.
Di Giovanni, G.D., L.S. Watrud, R.J. Seidler, and F. Widmer. 1999a. Comparison of parental and
transgenic alfalfa rhizosphere communities using Biolog GN metabolic fingerprinting and
enterobacterial repetitive intergeneric consensus sequence-PCR (ERIC-PCR). Microbiol. Ecol.
37:129-139.
Di Giovanni, G.D., L.S. Watrud, R.J. Seidler, and F. Widmer. 1999b. Fingerprinting of mixed
bacterial strains and BIOLOG Gram Negative (GN) substrate communities by enterobacterial
repetitive intergeneric consensus sequence-PCR (ERIC-PCR). Curr. Microbiol. 38:217-223.
Dionisi, H.M., A.C. Layton, G. Harms, I.R. Gregory, K.G. Robinson, and G. S. Sayler.
Quantification of Nitrosomonas oligotropha-like ammonia oxidizing bacteria and Nitrospira
spp. from full-scale wastewater treatment plants by competitive PCR. 2002. Appl. Environ.
Microbiol. 68: 245-253.
Donegan, K.K., R.J. Seidler, J.A. Doyle, L.A. Porteous, G. Di Giovanni, F. Widmer, and L. S.
Watrud. 1999. A field study with genetically engineered alfalfa inoculated with recombinant
Rhizobium melilotv. effects on the soil ecosystem. J. Appl. Ecol. 36:920-936.
Duggan, P.S., P.A. Chambers, J. Heritage, and J.M. Forbes. 2000. Survival of free DNA encoding
antibiotic resistance from transgenic maize and the transformation activity of DNA in ovine
saliva, ovine rumen fluid and silage effluent. FEMS Microbiol. Lett. 191:71-77.
Fairbrother, A., and L. Kapustka. 2001. In Ferenc, S. Impacts of Low-Dose, High-Potency
Herbicides on Non-target and Unintended Plant Species, ed. S. Ferenc, pp. 1-17. SET AC,
Pensacola, Florida.
Ferenc, S.A. 2001. Impacts of Low-Dose, High-Potency Herbicides on Non-target and Unintended
Plant Species, SETAC, Pensacola, Florida.
Fischhoff, D.A., and L.S. Watrud. 1988. Microbes as a source of genes for agricultural
biotechnology. In Biotechnology: Professional Issues and Social Concerns, eds. P. DeForest,
8-3
-------
M. S. Frankel, J. S. Poindexter, and V. Weil, pp. 65-71. Am. Assn. for Adv. of Sci., Wash.,
DC.
Fletcher, J.S. 1997. Testing and knowledge of organic phytotoxins. In Soil Ecotoxicology, eds. J.
Tarradellas, G. Bitton, and D. Rossel, pp. 271-289. CRC Press, Inc., Lewis Publishers. Boca
Raton, FL.
Fletcher, J.S., F.L. Johnson, and J.C. McFarlane. 1990. Influence of greenhouse versus field testing
and taxonomic differences on plant sensitivity to chemical treatment. Environ. Tox. Chem.
9:769-776.
Fletcher, J.S., M.J. Muhitch, D.R. Vann, J.C. McFarlane, and F.E. Benenati. 1985. Phytotox
database evaluation of surrogate plant species recommended by the U.S. Environmental
Protection Agency and the Organization for Economic Cooperation and Development.
Environ. Tox. Chem. 4:523-5532.
Fletcher, J.,and H. Ratsch. 1991. Plant Tier Testing: A Workshop to Evaluate Non-target Plant
Testing in Subdivision J Pesticide Guidelines. EPA/600/9-91/041. Corvallis, OR: US EPA,
Western Ecology Division.
Fletcher, J., T.G. Pfleeger, and H.C. Ratsch. 1993. Potential environmental risks associated with the
new sulfonylurea herbicides. Environ. Sci. Technol. 27:2250-2252.
Fletcher, J. T.G. Pfleeger, and H.C. Ratsch. 1995. Chlorsulfuron influence on garden pea
reproduction. Physiol. Plant. 94:261-267.
Fletcher, J.S., T.G. Pfleeger, H.C. Ratsch, H.C., and R. Hayes. 1996. Potential impact of low levels
of chlorsulfuron and other herbicides on growth and yield of non-target plants. Environ.
Toxicol. Chem. 15:1189.
Freemark, K,. and C. Boutin. 1995. Impacts of agricultural herbicide use on terrestrial wildlife in
temperate landscapes: A review with special reference to North American. Agric. Ecos.
Environ. 52:67-91
8-4
-------
Freemark, K., and C. Boutin. 1994. Impacts of agricultural herbicide use on terrestrial wildlife: A
review with special reference to Canada. Technical Report No. 196, Canadian Wildlife Service,
Headquarters, Ottawa, Canada.
Greig-Smith, P. W. 1991. The Boxwort experience: effects of pesticides on the fauna and flora of
cereal fields, pp. 333-371. In The Ecology of Temperate Cereal Fields, eds. L.G. Firbank, N.
Carier, J.F. Daiyshire, and G.R. Potts. British Ecological Society Symposium 32. Blackwell
Scientific Publications, Oxford, U.K.
Hardegger, M., P. Brodmann, and A. Hermann. 1999. Quantitative detection of the 35S promoter
and the nos terminator using quantitative competitive PCR. European Food Res. Technol.
209:83-87.
Heid, C. A.. J. Stevens, K. J. Livak. and P. M. Williams. 1996. Real time quantitative PCR. Gen.
Res. 6:986-994.
Heywood, V.H. 1978. Flowering Plants of the World. Mayflower Books, New York.
Helsel, Z.R., E. Ratcliff, and W. Rudolph. 1987. Maleic hydrazide effects on soybean reproductive
development and yield. Agron. J. 79:910-912.
Hill, D.A. 1985. The feeding ecology and survival of pheasant chicks on arable farmland. J. Appl.
Ecol. 22:645-654.
Howe, H.F. 1977. Bird activity and seed dispersal of a tropical wet forest tree. Ecology 58:539-550.
Hume, L. 1987. Long-term effects of 2,4-D applications on plants. I. Effects on the weed
community in a wheat crop. Can. J. Bot. 65:2530-2536.
Jobin, B., C. Boutin, and J.-L. Des Granges. 1997. Effects of agricultural practices on the flora of
hedgerows and woodland edges in southern Quebec. Can. J. Plant Sci. 77:293-299.
Kleijn, D, and G.I. Snoeijing. 1997. Field boundary vegetation and the effects of agrochemical drift:
botanical change caused by low levels of herbicide and fertilizer. J. Appl. Ecol. 34:1413-1425.
8-5
-------
Landis, W.G., L.A. Lenart, and J. A. Spromberg. 2000. Dynamics of horizontal gene transfer and the
ecological risk assessment of genetically engineered organisms. Human Ecol. Risk Assess.
6:875-899.
Lemerle, D., J. A. Fischer, and R.B. Hinckley. 1993. Radiometry accurately measures chlorsulfuron
injury to barley. Aust. J. Agric. Res. 44:13-21.
Lyon, D.J., and R.G. Wilson. 1986. Sensitivity of fieldbeans (Phaseolus vulgaris) to reduced rates
of 2,4-D and dicamba. Weed Sci. 34:953-956.
Maxwell, B., and R. Weed. 2001. Non-target terrestrial plant effects of low-dose, high potency
herbicides. In Impacts of Low-Dose, High-Potency Herbicides on Non-target and Unintended
Plant Species, ed. S. Ferrenc. pp. 85-97. SETAC, Pensacola, Florida.
McCarty, L.S., M. Power, and K.R. Munkittrick. 2002. Bioindicators versus biomarkers in
ecological risk assessment. Human Ecol. Risk. Assess. 8:159-164.
McCarthy, J.F. 1990. Concluding remarks. In Biomarkers of Environmental Contamination eds. J.F.
McCarthy, and L.R. Shugart, pp. 429-430. Lewis Publishers, Chelsea MI.
McKelvey, R.A., J.P. Wright, and J.L. Honegger. 2002. A comparison of crop and non-crop plants
as sensitive indicator species for regulatory testing. Pest Manag. Sci. 58:1161-1174.
Mineau, P., P.J. Sheehan, and A. Baril. 1987. Pesticides and waterfowl on the Canadian prairies: a
pressing need for research and monitoring. Pp. 133-147. ICBP Technical Publication. No. 6.
National Academy of Sciences. 2000a. The Future Role of Pesticides in US Agriculture. National
Academy Press. Washington, DC.
National Academy of Sciences. 2000b. Genetically Modified Pest-Protected Plants: Science and
Regulation. National Academy Press. Washington, DC.
National Academy of Sciences. 2002. Environmental Effects of Transgenic Plants: The Scope and
Adequacy of Regulation. National Academy Press. Washington, DC.
8-6
-------
Neal, R.H . 2002. Identifying and tracking pesticide use in agriculture. ArcUser 5:22-23.
Obery, A.M., and W.G. Landis. 2002. A regional multiple stressor risk assessment of the Codorus
Creek watershed applying the relative risk model. Human Ecol. Risk. Assess. 8:405-428.
Obrigawitch, T.T., G. Cook, and J. Wetherington. 1998. Assessment of effects on non-target plants
from sulfonylurea herbicides using field approaches. Pestic. Sci. 52:199-217.
O'Conner, R.J. 1992. Indirect effects of pesticides on birds. Proc. Brighton crop protection
conference: pests and diseases. 3:1097-1104.
Olszyk, D.M., and L.L. Ransinghe. 1994. Effects of cari>on dioxide on competition between rice
and barnyardgrass. IRRN 19:549-50.
Ostfeld, R.S., C.G. Jones, and J.O. Wolff. 1996. Of mice and mast: ecological connections in
eastern deciduous forests. Bioscience 46:323-330.
Orson, J. 2002. Gene stacking in herbicide tolerant oilseed rape: lessons from the North
American experience. 17 pp. English Nature Research Reports. No. 443
Pan, W. 2002. A comparative review of statistical methods for discovering differentially expressed
genes in replicated microarray experiments. Bioinformatics 18:536-545.
Peterson, G., S. Cunningham, L. Deutsch, J. Erickson, A. Quinlan, E. Raez-Luna, R. Tinch, M.
Troell, P. Woodbury, and S. Zens. 2000. The risks and benefits of genetically modified crops: a
multidisciplinary perspective. Cons. Ecol. 4:13 [online] URL: http://www.consecol.org/
vol4/issl/artl3.
Pfleeger, T., and D. Zobel. 1995. Organic pesticide modification of species interactions in annual
plant communities. Ecotoxicol 4:15-37.
Pfender, W.F., S.P. Maggard, and L.S. Watrud. 1995. Soil microbial activity and plant/microbe
symbioses as indicators for ecological effects of bioremediation technology. Proceedings of
Risk Assessment Methodologies; Conference Held at the University of MD, College Park, MD,
June 22-25,1994. Publ. # 1001 Univ. ofMD Biotechnology Institute # TP 9501, pp. 269-279.
8-7
-------
Porteous, L. A., J.L. Armstrong, R.J. Seidler, and L.S. Watrud. 1994. An effective method to extract
DNAs from environmental samples for polymerase chain reaction amplification and DNA
fingerprint analysis. Curr. Microbiol. 29:301-307.
Porteous, L.A., R.J. Seidler, and L.S. Watrud. 1997. An improved method for purifying DNA from
soil for polymerase chain reaction amplification and molecular ecology applications. Mol. Ecol.
6:787-791.
Purrington, C.B., and J. Bergelson. 1995. Assessing weediness of transgenic crops:industry plays
plant ecologist. Trends Ecol. Evol. 10:340-342.
Quist, D., and I.H. Chapela. 2001. Transgenic DNA introgressed into traditional maize landraces in
Oaxaca, Mexico. Nature 414:541-543.
Ramoni M.F., P. Sebastiani and I.S. Kohane. 2002. Cluster analysis of gene expression dynamics.
Proc. Nat. Acad. Sci. 99:9121-9126.
Ratsch, H., and J. Fletcher. 1991. Plant reproduction and/or life cycle testing. In Plant Tier
Testing: A Workshop to Evaluate Non-target Plant Testing in Subdivision J Pesticide
Guidelines eds. J. Fletcher, and H. Ratsch. pp. 80-89. US EPA, Western Ecology Division,
Corvallis, Oregon EPA/600/9-91/041.
Reganold, J.P., J.D. Glover, P.K. Andrews, and H.R. Hinman. 2001. Sustainability of three apple
production systems. Nature. 410:926-930.
Sambrook, J., and D.W. Russel. 2001. Molecular cloning a laboratory manual. 3rd. Ed. Cold
Spring Laboratory Press, Cold Spring Harbor, New York
Science Advisory Panel. 2001. Final Report of the FTFRA Scientific Advisory Panel (SAP)
Meeting Held on June 27-29,2001: Review of Non-Target Plant Toxicity Tests Under the
North American Free Trade Agreement (NAFTA).
Scientific Methods Workshop: Ecological and Agronomic Consequences of Gene Flow From
Transgenic Crops to Wild Relatives. Meeting Proceedings. 2002. The University Plaza Hotel
and conference Center. Ohio State University. Columbus, OH.
8-8
-------
Sears, M.K., R.L. Hellmich, D.E. Stanley-Horn, K.S. Oberhauser, J.M.. Pleasants, H.R. Mattila,
B.D. Siegfried, and G.P. Dively. 2001. Impact of Bt corn pollen on monarch butterfly
populations: A risk assessment. Proc. Natl. Acad. Sci. USA, Vol. 98, Issue 21,11937-
11942. Proc. Natl. Acad. Sci. USA, 98:11937-11942.
Segarra, A.E.. and J.M. Rawson . 2001. Starlink™ Corn Controversy: Background. CRS report
for Congress. 6pp.
Seidler, R.J., L.S. Watrud, and S.E. George. 1998. Assessing risks to ecosystems and human health
from genetically modified organisms. In Handbook of Environmental Risk Assessment and
Management. Ed. P. Calow. pp. 110-146. Blackwell Science, Oxford, UK.
Sheehan, P.J., A.Baril, P. Mineau, D.K. Smith, A. Harfenist, and W.K. Marshall. 1987. The impact
of pesticides on the ecology of prairie nesting ducks. Technical Report Series No. 19. Canadian
Wildlife Service.
Siciliano, S.D., C.M. Theoret, J.R. de Freitas, P.J. Hucl, and J.J. Germida. 1998. Differences in the
microbial communities associated with the roots of different cultivars of canola and wheat.
Can. J. Microbiol. 44: 844-851.
Simon, R., M.D. Radmacher, and K. Dobbin. 2002. Design of studies using DNA microarrays.
Genetic Epidemiol. 23:21-36.
Snow, A. 1997. Commercialization of transgenic plants: ecological risks. Bioscience 47:86-96
Spray Drift Task Force (1997a) A summary of ground application studies. Report available at
http://www.agdn fl.com/AgDRIFT /PDF_FILES/ground.pdf
Spray Drift Task Force (1997b) A summary of aerial application studies. Report available at
http://www.aedrift.com/AgDRIFT/PDF FILES/aerial.pdf
Stanley, P.I., and A.R. Hardy. 1984. The environmental implications of current pesticide usage on
cereals. In Agriculture and the Environment, ed. D.. Jenkins, pp. 66-72. Institute of Terrestrial
Ecology, Cambridge, U.K
8-9
-------
Stavely, J. 2002a. Comparison of standard Methods for Conducting Aquatic and Terrestrial Plant
Toxicity Tests. Poster at Non-target Plant Risk Assessment Workshop, January 15-17,2002,
Alexandria, VA.
Stavely, J. 2002b. Findings: Non-Target Plant Risk Assessment Workshop for Regulators,
Alexandria, Virginia, January, 2002. ARCADIS G&M, Inc.
Stephenson, G.L., N. Koper, G.F. Atkinson, K.R. Solomon, and R.P. Scroggins. 2000. Use of non
linear regression techniques for describing concentration response relationships of plant species
exposed to contaminated site soil. Envion. Tox. Chem. 19:2968-2981.
Taylor, G. Jr. 2001. Non-target terrestrial plant effects of low-dose, high potency herbicides, pp.
129-158 . In. Ferenc, S. Impacts of Low-Dose, High-Potency Herbicides on Non-target and
Unintended Plant Species, SETAC, Pensacola, Florida.
Teske, M.E., S.L. Bird, D.M. Esterly, T.B. Curbishley, S.L. Ray, and S.G. Perry. 2002. AgDRIFT:
A model for estimating near-field spray drift from aerial applications. Environ. Toxic. Chem.
21:659-671.
Templin, M.F,D. Stoll, M. Schrenk, P.C. Traub, C.F .Vohringer, and T. O . Joos. 2002. Protein
microarrays. Trends Biotechnol., 20:160-166.
Tilman, D. 1988. Plant strategies and the dynamics and structure of plant communities. In
Monograp/25 in Population Biology, eds. S. Levin, and H. Horn, Princeton University Press,
Princeton, NJ, Nr 26. 360 p.
USDA-NASS. National Agricultural Statistics Service (NASS). 1999.1997 Census of Agriculture,
Geographic Area Series, 3 CD-ROM set. USDA-NASS, Washington, DC.
U.S. Environmental Protection Agency. 1992. Framework for Ecological Risk Assessment PB93-
102192 EPA/630/R-92/001.
U.S. Environmental Protection Agency. 1998. Guidelines for Ecological Risk Assessment.
EPA/630/R-95/002F.
8-10
-------
Vierhelig, H, M. Alt, J. Lange, M. Gut-rella, A. Wiemken, and T. Bollen. 1995. Colonization of
transgenic tobacco constitutiviely expressing pathogenesis related proteins by the vesicular
arbuscular mycorrhizal fungus Glomus mosseae. Appl. Environ. Microbiol. 8:3031-3034.
Watrud, L.S., F.J. Perlak, M.T. Tran, K. Kusano, M.A. Miller-Wideman, M.G. Obukowicz, D.R.
Nelson, J.P. Kreitinger, and R. J. Kaufman. 1985. Cloning of the Bacillus thuringiensis subsp.
kurstaki delta-endotoxin gene into Pseudomonas fluorescens: molecular biology and ecology of
an engineered microbial pesticide. In Engineered Organisms in the Environment: Scientific
Issues, eds. H.O. Halvorson, D.Pramer, and M. Rogul. American Society
Watrud, L.S. 1987a. The biological revolution: tools and products of biotechnology. In
Application of Biotechnology: Environmental and Policy Issue, ed. J.R. Fowle, HI, pp. 11-26.
Westview Press, Boulder, Colorado.
Watrud, L.S. 1987b. Developing biotechnology products for agricultural pest control. In Trace
Substances in Environmental Health, ed. D.D. Hemphill, pp. 323-328. University ofMissouri,
Columbia, Missouri.
Watrud, L. S. 2000. Genetically Engineered Plants in the Environment: Applications and Issues. In
Microbial Interactions in Agriculture and Forestry. Eds. N.S. Subba Rao, and Y.R.
Dommergues, pp. 61 -81. Science Publishers, Inc. Plymouth, UK.
Webster, G., T.M. Embley, and J.I. Prosser. 2002. Grassland management regimes reduce small-
scale heterogeneity and species diversity ofbeta-proteobacterial ammonia oxidizer populations.
Appl. Environ. Microbiol. 68:20-30.
Westra, P., G. Franc, B. Cranmer, and T.D'Amato. 1991. Research report on 1988 potato-herbicide
injury research. In Plant Tier Testing: A Workshop to Evalua.e Non-target Plant Testing in
Subdivision J Pesticide Guidelines, eds. J. Fletcher, and H. Ratsch, pp. 98-104. US EPA,
Western Ecology Division, Corvallis, Oregon EPA/600/9-91/041.
Widmer, F., R.J. Seidler, P.M. Gillevet, L.S. Watrud, and G.D. Di Giovanni. 1998. A highly
selective PCR protocol for detecting 16S rRNA genes of the Genus Pseudomonas (Sensu
Stricto) in environmental samples. Appl. Environ. Microbiol. 64:2545-2553.
8-11
-------
Widmer, F., R.J. Seidler, and L.S. Watrud. 1996a. Persistence of NPT-II from transgenic plants in
soil:laboratory and field observations. In Biotechnology Risk Assessment, pp. 189-191.
USEP A/USD A/Environment Canada/Agriculture and Agri-Food Canada. Proceedings of the
Biotechnology Risk Assessment Symposium, June 6-8,1995, Pensacola, FL. Publ. #1002 Univ.
of MD Biotechnology Institute, College Park.
Widmer, F., R.J. Seidler, and L.S. Watrud. 1996b. Sensitive detection of transgenic plant marker
gene persistence in soil microcosms. Mol. Ecol. 5:603-613.
Widmer, F., B.T. Shaffer, L.A. Porteous, and R.J. Seidler. 1999. Analysis of nifH gene pool
compexity in soil and litter at a Douglas fir forest site in the Oregon Cascade mountain range.
Appl. Environ. Microbiol. 65:374-380.
Wiegers, J.K., H.M. Feder, L.S. Mortensen, D.G. Shaw, V.J. Wilson, and W.G. Landis. 1998. A
regional multiple-stressor rank-based ecological risk assessment for the Fjord of Port Valdez,
Alaska. Human Ecol. Risk Assess. 4:1125-73.
Winton, L.M., J.K. Stone, L.S. Watrud, and E.M. Hansen. 2001. Simultaneous one-tube
quantification of host and pathogen DNA using Taq-man real-time PCR. Phytopathology
92:112-116.
Wipff, J. 2002. Gene flow in turf and forage grasses (Poaceae). In: Scientific Methods
Workshop: Ecological and Agronomic Consequences of Gene Flow. pp.143-161
Wozniak, C. 2002. Gene flow assessment for plant-incorporated protectants by the Biopesticide
and Pollution Prevention Division, U. E. P. A. In: Scientific Methods Workshop: Ecological
and Agronomic Consequences of Gene Flow, pp. 162-177.
Yang, Y. H., and T Speed. 2002. Design issues for cDNA microarray experiments. Nature
Reviews Genetics 3:579-588.
8-12
-------
Appendix A Potential research on effects of chemical herbicides on aquatic plants.
-------
Potential research on effects of chemical herbicides on aquatic plants.
Rationale
The aquatic research needs of OPPTS in support of FEFRA and TSCA are quite broad. These
include expanding the types of aquatic tests required to include representative species from a variety of
aquatic plant communities ranging from freshwater wetlands to marine alga systems (see attached
table). The environmental conditions of the test systems needs to be characterized and standardized.
At least two reproductive tests are needed and will have to be developed. The higher tier or level tests
needs to be developed including multi-species microcosm, mesocosm and field tests. Research is also
needed to determine the linkage between laboratory tests and field results. The research needed to
improve the aquatic plant risk assessments will require a wider level of expertise than the terrestrial
component simply because the needs span from fresh water systems to marine systems.
There are currently two aquatic tests required for pesticide registration (listed below). One for
algae and the other for the vascular plant Lemna. This data is extrapolated to protect all freshwater
and marine algae, plants and communities in the United States and therefor probably the world.
a) Algal toxicity tests, Tiers I and II - The intended use is for developing data on the acute
toxicity of chemical substances and mixtures in aquatic environments subject to environmental effects
test regulations and was written specifically for Selenastrum capricornutum (fresh water green alga)
and Skeletonema costatum (marine diatom). It is also used for Anaabaena flos-aquae and Navicula
pilliculosa. The test occurs in flasks and lasts up to 96 hours. The end point is number of cells per unit
volume which is turn is used to determine EC50's
b) Aquatic Plant Toxicity Test Using Lemna spp., Tiers I and II -The intended use is for
developing data on the phytotoxicity of chemicals to the aquatic environment using the freshwater
aquatic plants Lemna gibba and L. minor. The test takes place in a vessel and lasts up to 14 days.
The end point is number of fronds which are used to calculate EC5s,50s, and 90s along with LOECs
(Lowest Effect Concentration) and NOECs (No Effect Concentration).
Examples of Needed Research
The algal test requires experimental evaluation of toxicity endpoints, test organisms and test
conditions. These issues have not been resolved in the scientific literature and a program to resolve
-------
Table 1. Additional species suggested for aquatic testing for the protection of nontarget plants from
phytotoxic chemicals grouped by growth form / environment and type of exposure.
Group
Additional Species
1
Freshwater algae
Green algae: Scenedesmus subspicatus; Chorella
vulgaris; Chlamydomonas reinhardi; Chlamydomonas
eugametos.
Blue-green algae: Anabaena cylindrica; Microcystis
aeruginosa
Diatom: Nitzchia sp.; Craticula cuspidata
2
Marine algae
Diatom: Thalassiosira pseudonana
Dinoflagellate: ND
Red algae: ND
Golden-brown algae: Macrocystis pyrifera
3,5
Floating vascular
Nuphar sp; Nymphaea sp.; Spirodela sp.
4
Submersed vascular
Ceratophyllum sp.; Vallisneria americana', Elodea
canadensis, Egeria densa, Potamogeton perfoliatus;
Najas sp.
6,7
Emergent vascular
Monocot: Spartina pectinata; Scirpus acutus; Phalaris
arundinacea
Dicot: Nelumbo lutea; Rorippa nasturium-aquaticum
ND - not determined
1-4 aquatic exposure
5-7 foliar exposure
-------
Appendix B Example experimental protocol for a study on vegetative and reproductive
responses of a major crop.
-------
Appendix B. Example experimental protocol for a study on vegetative and reproductive
responses of a major crop.
This experiment will address the Pesticide Project Goal of developing
information for improved plant testing guidelines to be used in
evaluating the potential effects of pesticides on terrestrial plants.
The first proposed experiment is a comparison of herbicides effects
on plants growing in different environments, treated with a herbicide
at different growth stages, and representing two crops with different
economic endpoints.
Objectives:
1. To compare the relative response of crop plants to a herbicide
when grown in greenhouse vs. field.
2. To compare the relative response of crop plants to herbicide
exposure at an early vegetative growth stage vs. a mature
reproductive/full developed stage of growth.
3. To compare the relative response to a herbicide of a crop with a
seed production economic endpoint vs. a crop with a storage root
production economic endpoint.
4. To evaluate possible physiological indicators of noninjurious
effects of herbicides on plants.
5. To evaluate the performance of the track sprayer under
experimental conditions and to evaluate herbicide application QA
procedures.
Plant Species:
Soybean, seed, 4 days from planting to germination, 65 days to
flowering (different variety), harvest date will be based on seed
development
Potato, tubers, 10 days from planting to germination, flowering (and
presumably initiation of tuber development) 28 days from germination,
harvested approximately 5 0 days after germination or when tubers are
fully developed
Pots/Soil/Seeding:
Pot Size: Soybean- 6" diameter x 5 3/4" deep green plastic pot,
Potato- 10" diameter x 12" high black plastic pots; saucers will be
placed under each pot
Soil: Sandy loam soil, sterilized by OSU Horticulture Dept. Samples
of the soil used for soybean and potatoes (different batches) will be
send to the OSU soil analysis lab. to determine fertility (N, P and K
concentrations), texture, pH and carbon concentration.
Fertilizer: Osmocote incorporated in soil at time of potting, 10
-------
Appendix C Federal regulation Tor Plant-Incorporated Protectants; Final Rules and
Proposed Rule (40 CFR 152 and 174, Part IV),
-------
i
0
F=3
y
Thursday,
July 19, 2001
Part IV
Environmental
Protection Agency
40 CFR Parts 152 and 174
Plant-Incorporated Protectants; Final
Rules and Proposed Rule
-------
37772 Federal Register/Vol. 66, No. 139/Thursday, July 19, 2001/Rules and Regulations
ENVIRONMENTAL PROTECTION
AGENCY
40CFR Parts 152 and 174
[OPP-300369B; FRL-6057-7]
RIN 2070-AC02
Regulations Under the Federal
Insecticide, Fungicide, and
Rodenticide Act for Plant-Incorporated
Protectants (Formerly Plant-
Pesticides)
AGENCY: Environmental Protection
Agency (EPA).
ACTION: Final rule.
SUMMARY: The substances plants
produce for protection against pests,
and the genetic material necessary to
produce these substances, are pesticides
under the Federal Insecticide, Fungicide
and Rodenticide Act (FIFRA), if humans
intend to use these substances for
"preventing, repelling or mitigating any
pest." In this rule, EPA finalizes certain
of the proposed rules published in 1994,
1996, and 1997. Specifically, EPA
changes the name of this type of
pesticide from "plant-pesticide" to
"plant-incorporated protectant";
clarifies the relationship between plants
and plant-incorporated protectants
under FIFRA; exempts from FIFRA
requirements plant-incorporated
protectants derived through
conventional breeding from sexually
compatible plants; and establishes a
new part in the Code of Federal
Regulations (CFR) specifically for plant-
incorporated protectants. Procedures are
also set forth for Confidential Business
Information (CBI); any claim of
confidentiality must be substantiated
when the claim is made. This rule will
benefit the public by ensuring that
public health and the environment are
adequately protected while reducing
burden on the regulated community,
thereby potentially reducing costs for
consumers.
DATES: This rule is effective September
17, 2001.
FOR FURTHER INFORMATION CONTACT:
Philip Hutton, Biopesticides and
Pollution Prevention Division, Office of
Pesticide Programs (751 lC),
Environmental Protection Agency, 1200
Pennsylvania Ave., NW., Washington,
DC 20460; telephone number; (703)
308-8260; e-mail address;
hutton.phil@epa.gov.
SUPPLEMENTARY INFORMATION:
I. General Information
A. Does this Action Apply to Me?
You may be potentially affected by
this action if you are a person or
company involved with agricultural
biotechnology that may develop and
market plant-incorporated protectants.
Potentially affected categories and
entities mav include, but are not limited
to:
Categories
NAICS codes
Examples of potentially affected entities
Pesticide manufacturers
32532
Establishments primarily engaged in the formulation and preparation of agricultural and
household pest control chemicals
Seed companies
111
Establishments primarily engaged in growing crops, plants, vines, or trees and their seeds
Colleges, universities, and pro-
611310
Establishments of higher learning which are engaged in development and marketing of
fessional schools
plant-incorporated protectants
Establishments involved in re-
54171
Establishments primarily engaged in conducting research in the physical, engineering, or
search and development in the
life sciences, such as agriculture and biotechnology
life sciences
This table is not intended to be
exhaustive, but rather provides a guide
for readers regarding the types of
entities potentially affected by this
action. Other types of entities not listed
in the table could also be affected. The
North American Industrial
Classification System (NAICS) codes
have been provided to assist you and
others in determining whether or not
this action might apply to certain
entities. To determine whether you or
your business may be affected by this
action, you should carefully examine
the provisions in 40 CFR part 174. If you
have any questions regarding
applicability of this action to a
particular entity, consult the person
listed under FOR FURTHER INFORMATION
CONTACT.
B. How Can I Get Additional
Information, Including Copies of this
Document and Other Related
Documents?
1. Electronically.You may obtain
electronic copies of this document, and
certain other related documents that
might be available electronically, from
the EPA Internet Home Page at http;//
www.epa.gov/. To access this
document, on the Home Page select
""Laws and Regulations", "Regulations
and Proposed Rules," and then look up
the entry for this document under the
"Federal Register—Environmental
Documents." You can also go directly to
theFederal Register listings at http;//
www.epa.gov/fedrgstr/. To access
information about EPA's program for
biopesticides go directly to the Home
Page for the Office of Pesticide Programs
at http://www.epa.gov/pesticides/
biopesticides.
2. In person. The Agency has
established an official record for this
action under the docket control number
OPP-300369B. The official record
consists of the documents specifically
referenced in this action, any public
comments received during an applicable
comment period, and other information
related to this action, including any
information claimed as Confidential
Business Information (CBI). This official
record includes the documents that are
physically located in the docket, as well
as the documents that are referenced in
those documents. The public version of
the official record, which includes
printed, paper versions of any electronic
comments submitted during an
applicable comment period, is available
for inspection in the Public Information
and Record Integrity Branch (PIRIB),
Rm. 119, Crystal Mall #2, 1921 Jefferson
Davis Highway, Arlington, VA, from
8:30 a.m. to 4 p.m., Monday through
Friday, excluding legal holidays. The
PIRIB telephone number is (703) 305-
5805.
II. Under What Authority Is EPA
Issuing The Rule?
A. FIFRA
This rule is promulgated under the
authority of FIFRA section 3 and se'ction
25(a) and (b) (7 U.S.C. 136a and 136w(a)
and (b)) and FFDCA section 346a and
371.
FIFRA section 3(a) provides, with
some exceptions, that no person may
distribute or sell in the United States
any pesticide that is not registered
-------
37774 Federal Register/Vol. 66, No. 139/Thursday, July 19, 2001/Rules and Regulations
determining whether a pesticide
chemical residue is safe. EPA must
consider "available information
regarding the aggregate exposure levels
of consumers ... to the pesticide
chemical residue and to other related
substances, including dietary exposure
under the tolerance and all other
tolerances in effect for the pesticide
chemical residue, and exposures from
other non-occupational sources." (21
U.S.C. section 346a(b)(2)(D)(vi)).
Consequently, a finding that a pesticide
qualifies for a tolerance exemption
could also demonstrate that the
pesticide chemical meets the first
exemption criterion of low probability
of risk with respect to human health
risks arising from other non-
occupational routes of exposure. Such a
pesticide also meets the second FIFRA
exemption criterion of no likely
unreasonable adverse effects, with
respect to human health risks arising
from all non-occupational exposures, if
the risks resulting from use of that
pesticide are consistent with the FFDCA
section 408 exemption standard, and the
potential benefits of use outweigh any
human health risk even in the absence
of regulatory oversight.
However, FIFRA does not provide for
exemption of a pesticide in food based
solely upon consistency with the
FFDCA section 408 exemption standard.
At a minimum, EPA also must evaluate
risks arising from occupational exposure
to humans and determine that such
risks meet both exemption criteria. In
addition, EPA must evaluate the risks to
the environment from the pesticide and
determine both that the pesticide poses
only a low probability of environmental
risks, and that use of the pesticide is not
likely to cause any unreasonable
adverse effects on the remainder of the
environment in the absence of
regulation under FIFRA.
IH. What is the Background for this
Rule?
This final rule establishes certain
basic parameters of EPA's regulatory
program under FIFRA for plant-
incorporated protectants. In this rule,
EPA defines the scope of products
subject to FIFRA jurisdiction, and
identifies the category of products over
which it will exert regulatory oversight.
EPA also establishes certain
fundamental definitions to clarify what
will be subject to regulation as a plant-
incorporated protectant. The rule also
finalizes certain regulatory procedures
specific to plant-incorporated
protectants. This document also
provides some guidance on the way in
which the Agency intends to interpret
the existing regulations for these
products until it is able to establish
additional regulations specific to plant-
incorporated protectants.
Specifically, the rule clarifies that
plants used as biological control agents
remain exempt from FIFRA
requirements, but that plant-
incorporated protectants are not.
Second, the rule exempts plant-
incorporated protectants derived
through conventional breeding from
sexually compatible plants. Third, this
final rule establishes a new 40 CFR part
174, specifically for plant-incorporated
protectants; any additional regulations
specific to plant-incorporated
protectants will be codified in 40 CFR
part 174. The final rule also imposes a
• requirement at § 174.71, that any person
producing an otherwise exempt plant-
incorporated protectant for sale and
distribution, who obtains any
information regarding adverse effects of
this otherwise exempt plant-
incorporated protectant on human
health or the environment report that
information to EPA. Finally, the rule
includes a provision that any claim of
confidentiality must be made at the time
of submission and substantiated at the
time the claim is made.
A. What Is a Plant-Incorporated
Protectant?
Plants have evolved, and thus
naturally possess, various mechanisms
to resist pests. The mechanisms of
resistance can be varied, including, for
example, structural characteristics of the
plant, the production of metabolites that
have toxic properties, biochemical
cascades resulting in localized necrosis
of plant tissue, or the production of
specific toxic substances in response to
pest attack. Humans have for
approximately 10,000 years selected and
bred certain plants as sources of, for
example, food, feed, and fiber, and a
frequently selected characteristic was
the ability to resist pests. More recently,
humans have developed scientific
techniques by which traits from any
living organism, including an ability to
resist pests, can be introduced into a
plant. When humans intend to use
substances involved in these
mechanisms in plants for "preventing,
destroying, repelling or mitigating any
pest," the substances are pesticides
under the FIFRA definition of pesticide,
regardless of whether the pesticidal
capability evolved in the plants or was
introduced by breeding or through the
techniques of modern biotechnology.
The genetic material necessary for the
production of such a pesticidal
substance also meets the FIFRA
statutory definition of a pesticide. Such
genetic material is introduced into a
plant with the intent of ultimately
producing a pesticidal effect even
though the genetic material may not,
itself, directly affect pests. The
pesticidal substance, along with the
genetic material necessary to produce it,
produced and used in living plants, is
designated a "plant-incorporated
protectant" by EPA.
Plant-incorporated protectants are
primarily distinguished from other
types of pesticides because they are
intended to be produced and used in the
living plant. This difference in use
pattern dictates in some instances
differences in approach. For example,
because the plant-incorporated
protectant is produced by the plant
itself and used in the living plant,
exposure considerations in risk
assessments may be different, although
as noted in Unit VII.D.2., the risk
assessment framework used for other
types of pesticides can be used for
plant-incorporated protectants.
B. Does the Rule Have Any Relevance to
Other Types of Pesticides?
Nonviable plant tissues, organs, or
parts that are used as pesticides, will
not be subject to the provisions of this
rule, which will be codified in
regulations at 40 CFR part 174. Rather,
such pesticides are subject to the
regulations found in 40 CFR parts 150
through 173 and 40 CFR parts 177
through 180. An example of this type of
pesticide would be the powder,
produced by drying and grinding
cayenne peppers, dusted on plants to
protect them from pests.
Substances that are isolated from a
plant's tissues and then applied to
plants for pest control will not be
subject to the regulations in 40 CFR part
174. Rather these types of pesticides in
formulations such as those for foliar
application are subject to regulations
found in 40 CFR parts 150 through 173
and 40 CFR parts 177 through 180. An
example of this type of pesticide would
be pyrethrum isolated from
chrysanthemum plants, formulated with
other ingredients for foliar application,
and sprayed on other plants for pest
control.
Substances that are synthesized will
not be subject to the regulations in 40
CFR part 174. Such pesticides are
subject to regulations found in 40 CFR
parts 150 through 173 and 40 CFR parts
177 through 180. An example of this
type of pesticide is the herbicide,
atrazine.
C. What is the History of this Rule?
This rule is an additional step in fully
implementing the "Coordinated
Framework for Regulation of
-------
37776 Federal Register/Vol. 66, No. 139/Thursday, July 19, 2001 /Rules and Regulations
A. What Are the Key Features of the
November 23, 1994, Federal Register?
In the November 23, 1994, Federal
Register document (59 FR 60519), EPA
proposed to: first, clarify- how the
exemption at 40 CFR 152.20 relates to
plants used as biological control agents
and to plant-incorporated protectants;
second, exempt under FIFRA section
25(b)(2), plant-incorporated protectants
that are derived from plants closely
related to the recipient plant, except for
a requirement that sellers or distributors
of an otherwise exempt plant-
incorporated protectant submit to EPA
any information they may obtain
regarding potential unreasonable
adverse effects caused by an exempt
plant-incorporated protectant; and third,
establish new part 40 CFR part 174
specifically for plant-incorporated
protectants. This document also
contained a proposed rule on
substantiation of any claim of
confidentiality at the time the claim was
made.
1. Clarification of exemption at 40
CFR 152.20; status of plants used as
biological control agents with regard to
FIFRA requirements. In the November
23, 1994, Federal Register document,
EPA proposed to amend 40 CFR 152.20
to clarify that plants used as biological
control agents are exempt from FIFRA
requirements under section 25(b)(1).
The proposed amendment at 40 CFR
152.20 would also indicate that this
exemption does not apply to plant-
incorporated protectants and would
refer the reader to 40 CFR part 174 for
regulations, including a listing of
exemptions, on plant-incorporated
protectants.
2. Proposed exemption of plant-
incorporated protectants derived from
plants closely related to the recipient
plant. In 1994, EPA described three
options for defining when a plant-
incorporated protectant would be
exempt because it is derived from plants
closely related to the recipient plant.
EPA proposed to exempt plant-
incorporated protectants derived from
plants closely related to the recipient
plant based on the rationale that the
probability of new exposures from this
group of plant-incorporated protectants
is very low. Option l, the Agency's
preferred option, used sexual
compatibility, including hybridization
achieved by wide and bridging crosses,
as a measure of relatedness between
plants. Under this option, plant-
incorporated protectants would be
exempt from all FIFRA requirements,
except for the adverse effects reporting
requirement, if the genetic material that
leads to the production of the pesticidal
substance is derived from plants that are
sexually compatible with the recipient
plant and has never been derived from
a source that is not sexually compatible
with the recipient plant. Recipient plant
was described as the plant into which
the plant-incorporated protectant is
introduced and in which the plant-
incorporated protectant is produced.
Sexually compatible, when referring to
plants, was described as capable of
forming a viable zygote through the
fusion of two gametes including the use
of bridging or wide crosses between
plants.
Option 2 would utilize the rank of
genus as the taxonomic standard for
describing closely related plants such
that plant-incorporated protectants
derived from plants classified in the
same genus as the recipient plant would
be exempt from all FEFRA requirements,
except for the adverse effects reporting
requirement. Taxonomy is a system of
orderly classification of organisms
according to their presumed natural
relationships. Taxonomy reflects current
scientific observations about
phenotypic, and to a certain extent,
genotypic, similarities between
organisms.
Option 3, also an alternative option,
would utilize both the taxonomic rank
of genus and sexual compatibility to
describe closely related plants. This
option would exempt from all FIFRA
requirements, except for the adverse
effects reporting requirement, plant-
incorporated protectants derived from
plants classified in the same genus as
the recipient plant, as well as plant-
incorfjorated protectants derived from
plants sexually compatible with the
recipient plant. Under Options 1 and 3,
plant-incorporated protectants derived
from plants sexually compatible with
the recipient plant would be exempt
even if the source and recipient plants
are classified in different genera.
None of the options offered by the
EPA were intended to exempt a plant-
incorporated protectant that has been
modified so that it is significantly
different functionally from the plant-
incorporated protectant as it occurs in
the source organism (59 FR 60524).
i. Associated definitions. In 1994,
pertinent definitions associated with the
proposed exemptions included:
"Bridging crosses between plants"
would be the utilization of an
intermediate plant in a cross to produce
a viable zygote between the
intermediate plant and a first plant, in
order to cross the plant resulting from
that zygote with a third plant that would
not otherwise be able to produce viable
zygotes from the fusion of its gametes
with those of the first plant. The result
of the bridging cross is the mixing of
genetic material of the first and third
plant through the formation of an
inteimediate zygote.
"Wide crosses between plants" would
be to facilitate the formation of viable
zygotes through the use of surgical
alteration of the plant pistil, bud
pollination, mentor pollen,
immunosuppressants, in vitro
fertilization, pre-pollination and post-
pollination hormone treatments,
manipulation of chromosome numbers,
embryo culture, or ovary and ovule
cultures, or any other technique that the
Administrator determines meets this
definition.
In 1994, EPA also presented a
definition for plant-pesticide, now
termed plant-incorporated protectant,
and definitions of active and inert
ingredient for plant-pesticides.
"Plant-pesticide" was defined as a
pesticidal substance that is produced in
a living plant and the genetic material
necessary for the production of the
substance, where the substance is
intended for use in the living plant.
"Active ingredient," when referring to
plant-incorporated protectants only, was
defined as a pesticidal substance that is
produced in a living plant and the
genetic material necessary for the
production of the substance, where the
substance is intended for use in the
living plant.
"Genetic material necessary for the
production" was defined as: Genetic
material that encodes for a pesticidal
substance or leads to the production of
a pesticidal substance and regulatory
regions. It does not include noncoding,
nonexpressed nucleotide sequences.
"Inert ingredient," when referring to
plant-incorporated protectants only, was
defined as any substance, such as a
selectable marker, other than the active
ingredient, and the genetic material
necessary for the production of the
substance, that is intentionally
introduced into a living plant along
with the active ingredient, where the
substance is used to confirm or ensure
the presence of the active ingredient.
"Living plant" was defined as a plant
that is alive, including periods of
dormancy, and all viable plant parts/
organs involved in the plant's life cycle.
"Noncoding, nonexpressed nucleotide
sequences" were defined as the
nucleotide sequences that are not
transcribed and are not involved in gene
expression. Examples of noncoding,
nonexpressed nucleotide sequences
include linkers, adapters,
homopolymers, and sequences of
restriction enzyme recognition sites.
ii. Potential exemption criterion based
on process. The Agency also requested
-------
37778 Federal Register/Vol. 66, No. 139/Thursday, July 19, 2001/Rules and Regulations
174.71, plant-incorporated protectants
that are derived through conventional
breeding from sexually compatible
plants. The exempt plant-incorporated
protectants represent a subcategory of
the plant-incorporated protectants
described in Option 1 in the November
23, 1994. Federal Register document (59
FR 60522). (EPA is seeking additional
comment in a supplemental document
published elsewhere in this issue of the
Federal Register on whether all plant-
incorporated protectants derived from
plants sexually compatible with the
recipient plant should be exempt from
FIFRA requirements, regardless of how
they are introduced into the recipient
plant.)
The following language appears in 40
CFR 174.25 to describe this subcategory:
A plant-incorporated protectant is exempt
if all of the following conditions are met:
(a) The genetic material that encodes the
pesticidal substance or leads to the
production of the pesticidal substance is
from a plan', that is sexually compatibie with
the recipient plant.
(bj The genetic materia] has never been
derived from a source that is not sexually
compatible with the recipient plant.
The following language addressing
inert ingredients in plants derived
through conventional breeding from
sexually compatible plants is added to
40 CFR 174.485, subpart X:
An inert ingredient, and residues of the
inert ingredient, are exempt if all of the
following conditions are met:
(a) The genetic material that encodes the
inert ingTedient or leads to the production of
the inert ingTedient is derived from a plant
sexually compatible with the recipient food
plant.
Cb) The genetic material has never been
derived from a source that is not sexually
compatible with the recipient food plant.
(c) The resides of the inert ingredient are
not present in food from the plant at levels
that are injurious or deleterious to human
health
1. Associated definitions. Pertinent
definitions associated with the
exemption include:
"Bridging crosses between plants"
means the utilization of an intermediate
plant in a cross to produce a viable
zygote between the intermediate plant
and a first plant, in order to cross the
plant resulting from that zygote with a
third plant that would not otherwise be
able to produce viable zygotes from the
fusion of its gametes with those of the
first plant. The result of the bridging
cross is the mixing of genetic material
of the first and third plant through the
formation of an intermediate zygote.
"Cell fusion" means the fusion in
vitro of two or more cells or protoplasts.
"Conventional breeding of plants"
means the creation of progeny through
either: The union of gametes, i.e.,
syngamy, brought together through
processes such as pollination, including
bridging crosses between plants and
wide crosses: or vegetative
reproduction. It does not include use of
any one of the following technologies:
Recombinant DNA; other techniques
wherein the genetic material is extracted
from an organism and introduced into
the genome of the recipient plant
through, for example, micro-injection,
macro-injection, micro-encapsulation;
or cell fusion.
"Genome" means the sum of the
heritable genetic material in the plant,
including genetic material in the
nucleus and organelles.
"Recombinant DNA" means the
genetic material has been manipulated
in vitro through the use of restriction
endonucleases and/or other enzymes
that aid in modifying genetic material,
and subsequently introduced into the
genome of the plant.
"Sexually compatible," when
referring to plants, means a viable
zygote is formed only through the union
of two gametes through conventional
breeding.
"Source" means the donor of the
genetic material that encodes a
pesticidal substance or leads to the
production of a pesticidal substance,
"Vegetative reproduction" means: In
seed plants, reproduction by apomixis; '
and in other plants, reproduction by
vegetative spores, fragmentation, or
division of the somatic body.
"Wide crosses" means to facilitate the
formation of viable zygotes through the
use of.surgical alteration of the plant
pistil, bud pollination, mentor polien,
immunosuppressants, in vitro
fertilization, pre-pollination and post-
pollination hormone treatments,
manipulation of chromosome numbers,
embryo culture, or ovary and ovule
cultures.
Pertinent associated definitions in 40
CFR 174.3, several of which are
discussed in Unit VTI.B.8., include:
"Active ingredient" means a
pesticidal substance that is intended to
be produced and used in a living plant,
or in the produce thereof, and the
genetic material necessary for the
production of such a pesticidal
substance.
"Genetic material necessary for the
production" means both: Genetic
material that encodes a substance or
leads to the production of a substance,
and regulatory regions. It does not
include noncoding, nonexpressed
nucleotide sequences.
"Inert ingredient" means any
substance, such as a selectable marker,
other than the active ingredient, where
the substance is used to confirm or
ensure the presence of the active
ingredient, and includes the genetic
material necessary for the production of
the substance, provided the genetic
material is intentionally introduced into
a living plant in addition to the active
ingredient.
'Living plant" means a plant, plant
organ, or plant part that is alive, viable
or dormant. Examples of plant parts
include, but are not limited to, seeds,
fruits, leaves, roots, stems, flowers and
pollen.
"Noncoding, nonexpressed nucleotide
sequences" means the sequences are not
transcribed and are not involved in gene
expression. Examples of noncoding,
nonexpressed nucleotide sequences
include, but are not limited to, linkers,
adaptors, homopolvmers, and sequences
of restriction recognition sites.
"Pesticidal substance" means a
substance that is intended to be
produced and used in a living plant, or
in the produce thereof, for a pesticidal
purpose during any part of a plant's life
cycle (e.g., in the embryo, seed,
seedling, mature plant).
"Plant-incorporated protectant"
means a pesticidal substance that is
intended to be produced and used in a
living plant, or in the produce thereof,
and the genetic material necessary for
the production of such a pesticidal
substance. It also contains any inert
ingredient contained in the plant, or
produce thereof.
"Produce thereof," when used with
respect to plants containing plant-
incorporated protectants only, means a
product of a living plant containing a
plant-incorporated protectant, where the
pesticidal substance is intended to serve
a pesticidal purpose after the product
has been separated from the living
plant. Examples of such products
include, but are not limited to,
agricultural produce, grains and lumber.
Products such as raw agricultural
commodities bearing pesticide chemical
residues are not "produce thereof
when the residues are not intended to
serve a pesticidal purpose in the
produce.
"Recipient plant" means the living
plant in which the plant-incorporated
protectant is intended to be produced
and used.
Other definitions, relevant for plant-
incorporated protectants only, can be
found at 40 CFR 174.3. In this final rule,
"plant" means an organism classified
using the 5-kingdoro classification
system of Whittaker (Ref. 1) in the
kingdom, Plantae. Therefore, the term
"plant" includes, but is not limited to,
bryophytes such as mosses,
pteridophytes such as ferns,
-------
68 Microbial Interactions in Agriculture and Forestry (Volume II)
Tabic X Examples of non-taiget and unintended effects of engineered plants
Trail
Plant
Effects
References
Insect resistance
cotton
and
Changes in size and diversity
of soil microbial, nematode
Donegan et al., 1995
potato
and trucToarthopod
populations; changes in soil
Hizyme activity
Donegan et al., 1996
tobacco
Changes in soil respiration;
changes in size and diversity of
protozoa, nematode and
mictoarthopod populations
Donegan et al, 199?
Disease resistance
tobacco
Decrease and delay in
aibuscular mycorrhizal
infection
Vierhelig et al., 1995
Herbicide resistance
Arabidopsis
Gene outcrossing
Bergelson et al., 1998
beets
Gene outcrossing
Dieti-Pfeilstetter and
Kirchner, 1998
canola
Gene outcrossing
Chevre et al., 1997;
Lefol et al , 1991;
Pumngton ic
Bergelson, 1995
canola
Change in endophytic and
rhizosphere microbial
populations
Siciliano et al., 1998
Specialty uses:
Di Giovanni et al.,
1999;
Donegan et al., 1999
Ligrun-Peroxid ase
alfalfa
Changes in rhizosphere and
soil microbial populations
alfalfa
Reduced shoot biotnass and
changes in shoot
macronutrient content
Donegan et a!., 1999
alfalfa
Reduced shoot biomass;
changes in macronutrient and
Biiomutrienl content; decreased
mytorrtuxal infection
Watrud et al, 1998
Auxin, Enzymes
aspen
Altered wood anatomy and
shoot growth; change in lignin
structure
Tuominen et al., 1995;
Lapierre et al., 1999
Pigments
pet -i lia
Loss of color
MacKenzie, 1990
-------
Appendix E Agenda for a "Scientific Methods Workshop: Ecological and Agronomic
Consequences of Gene Flow from Transgenic Crops to Wild Relatives,"
Columbus OH, 5-6 March, 2002.
-------
Scientific Methods Workshop:
Ecological and Agronomic Consequences of Gene Flow from
Transgenic Crops to Wild Relatives
Meeting Proceedings
The University Plaza Hotel and Conference Center
The Ohio State University
Columbus, OH
March 5th and 6th, 2002
Steering Committee:
Dr. Allison Snow (Chair and Co-PI), Ohio State University
Dr. Carol Mallory-Smith (Co-PI), Oregon State University
Dr. Norman Ellstrand, University of California at Riverside
Dr. Jodie Holt, University of California at Riverside
Dr. Hector Quemada, Crop Technology Consulting, Inc., Kalamazoo, MI
Logistical Coordinator: Dr. Lawrence Spencer, Ohio State University
-------
Table of Contents
Speakers/Titles
Allison Snow, Ohio State University
Introduction
Bert Abbott, Clemson University
"Molecular Genetic Assessment of the Risk of Gene Escape
in Strawberry, a Model Perennial Study Crop."
Paul Arriola, Elmhurst College
"Gene flow and hybrid fitness in the Sorghum bico/or-
Sorghum hatepense complex"
Lesley Blancas, University of California, Riverside
"Patterns of genetic diversity in sympatric and allopatric
Populations of maize and its wild relative teosinte in
Mexico: evidence for hybridization"
Norm Ellstrand, University of California, Riverside
"Gene Flow from Transgenic Crops to Wild Relatives:
What Have We Learned, What Do We Know, What
Do We Need to Know?"
Jodie Holt, University of California, Riverside (Plenary Speaker)
"Prevalence and Management of Herbicide-Resistant Weeds"
Diana Pilson, University of Nebraska
"Fitness and population effects of gene flow from transgenic
sunflower to wild Helianthus annuus"
Hector Quemada, University of Western Michigan
"Case Study: Gene flow from commercial transgenic
Cucurbits pepo to 'free-living' C. pepo populations"
Christiane Saeglitz, Aachen University of Technology
"Monitoring the Environmental Consequences of
Gene Flow from Transgenic Sugar Beet"
-------
Neal Stewart, University of North Carolina, Greensboro
"Gene flow and its consequences: Brassica napus
(canoia, oilseed rape) to wild relatives"
Steve Strauss, Oregon State University
"Gene flow in forest trees: From empirical estimates to
transgenic risk assessment"
Joseph Wipff, Pure Seed Testing, Inc.
"Gene flow in turf and forage grasses (Poaceae)"
Chris Wozniak, U.S. Environmental Protection Agency
"Gene Flow Assessment for Plant-Incorporated Protectants
by the Biopesticide and Pollution Prevention Division, U.S. EPA"
Robert Zemetra, University of Idaho
'The Evolution of a Biological Risk Program: Gene flow between
Wheat (Triticum aestivum L.) and Jointed Goatgrass
{Aegilops cylindrica Host)"
-------
Proceedings
Page 3
SUMMARY
Gene flow from transgenic plants to wild relatives is one of the major research areas targeted
by USDA's Biotechnology Risk Assessment Research Grants Program (BRARGP). We received
funds for a two-day workshop that will bring together researchers who study the prevalence
and consequences of gene flow from transgenic crops to weeds and other wild relatives. On
the first day, speakers will discuss the general context for gene flow research, the information
needs of USDA-APHIS, EPA, and the biotechnology industry, and case studies of specific crop-
wild complexes, including cucurbits, brassicas, sunflower, sorghum, rice, wheat, maize,
strawberry, poplar, and turfgrasses. Written summaries of these talks are included below. On
the second day, break-out groups will discuss the advantages and disadvantages of various
approaches for studying the occurrence of gene flow and various effects of gene flow (fitness
effects of transgenes in wild relatives, effects on population dynamics, indirect community
effects, and effects on the genetic diversity of wild relatives). The crops, wild relatives, and
regulatory issues we discuss will focus on the USA, but much of the workshop will be relevant
to similar situations in other countries. Proceedings from the workshop will be posted on an
internet website that will be publicized in professional journals and newsletters. Bridging the
fields weed science and plant ecology, this workshop will help define the most appropriate and
rigorous empirical methods available for studying questions related to gene flow from
transgenic crops to weedy and wild relatives.
BACKGROUND AND GOALS
Gene flow between crops and free-living, noncultivated plants is often considered to be an
undesirable consequence of adopting transgenic crops (e.g., NRC 1989, NRC 2000). This
process occurs when pollen moves from a crop to its wild or feral relative - or vice versa- and
genes from their offspring spread further via the dispersal of pollen and seeds. In addition,
some crops, such as oats, radish, and oilseed rape, can proliferate as feral weeds. Although
crops and weeds have exchanged genes for centuries, transgenes can confer novel, fitness-
related traits that were not available previously, and the same transgenes can be introduced
into many different crops, increasing the potential for their escape (e.g., resistance to the
herbicide glyphosate). A fundamental question, then, is what impacts could single or multiple
transgenes have on the abundance and distribution of wild relatives? From a regulatory
perspective, it is useful to compare the effects of transgenes to effects of nontransgenic crop
genes that spread to wild and/or weedy populations, keeping in mind that certain traits
developed through the introduction of transgenes (e.g. herbicide tolerance, herbivore and
pathogen resistance, and resistance to harsh environmental conditions) have been produced
through traditional breeding as well.
As a starting point, we need to determine which crops hybridize spontaneously with wild and/or
weedy relatives in a given country or region. In cases such as sunflower, squash, and radish,
the crop and the weed represent different forms of the same species, and crop-to-wild plant
gene flow occurs whenever these forms grow near each other. In sunflower and radish, crop
genes are known to persist for many generations in wild populations, even when first-
generation wild-crop hybrids produce fewer seeds per plant than wild plants (e.g., Whitton et
al. 1997, Snow et al. 2001). Gene flow can also occur when crops and weeds are more
Gene Row Workshop, The Ohio State University, March 5 and 6, 2002
t
-------
Appendix F Examples of meetings on biosafety and risk assessment of engineered plants
(Watrud, 2000).
-------
66
Table 1.
Microbial Interactions in Agriculture and Forestry (Volume II)
Examples of meetings on biosafety and risk assessment of engineered plants
Meeting/Title
Location/Year
Held
Editor/Publisher/Date/Organizers
1st International Symposium on the Kiawah Island,
Biosafety Results of Field Tests of SC. 1990
Genetically Modified Plants and
Microorganisms
Pestiadal Transgenic Plants: Product Annapolis,
Development, Risk Assessment and MD, 1990
Needs
Workshop on Safeguards for Planned Ithaca, NY,
Introduction of Transgenic Oilseed 1990
Symposium on Ecological Implications College Park,
of Transgenic Plant Release MD, 1992
2nd International Symposium on the Goslar,
Biosafety Results of Beid Tests of Germany,
Genetically Modified Plants and 1992 .
Microorganisms
Toward Enhanced and Sustainable Taipei,
Agricultural Productivity in the Taiwan, 1993
2000's. Breeding Research and
Biotechnology
3rd International Symposium on the Monterey,
Biosafety Results of Field Tests of CA, 1994
Genetically Modified Plants and
Microorganisms
OECD Workshop on Ecological Queenstown,
Implications of Transgenic Crop Plants New Zealand,
Containing Bacillus thuringirnsis 1994
Toxin Genes
Herbicide-resistant Crops: a Bitter or Memphis, TN,
Better Harvest? 1995
Dialogue on Risk Assessment of Domach,
Transgenic Plants: Scientific, Switzerland,
Technological and Societal 1997
Perspectives
4th International Symposium on the Tsukuba-
Biosafety Results of Field Tests of inachi. Japan,
Genetically Modified Plants and 1997
Microorganisms
Virus-resistant Transgenic Plants: Godollo,
Potential Ecological Impact Hungary, 1997
5th International Symposium on Braunschweig,
The Biosafety Results of Beid Tests of Germany,
Genetically Modified Plants and 1998
Microorganisms
MacKenzie, D.R., Henry, S.C. (eds.).
Agricultural Research Institute,
1991.
US EPA, Office of Pesticide
Programs, 1991
USD A, Animal and Plant Health
Inspection Service, 1990
Levin, M and R.J. Seidler (eds.),
Blackwell Scientific Publ., Oxford,
UK. Mol. EcoL 3:1-90,1994
Casper, R., Landsmann, J. (eds ).
Biologische Bundesanstalt fur
Land- and Forstwirtschaft, 1992
Academia Sinica, Nankang,
Taichung District Agricultural
Improvement Station, 1994
Jones, D.D. (ed.). University of
California, 1994
Hokkanen, H.M.T. (ed.). University
of Helsinki, Finland, 1994
Southern Weed Science Society,
Champaign, IL, 1995
Heaf, D. (coordinator), Ifene, UK,
1997
Matsui. S., Miyasaki. S., Kasamo, K.
(eds.). Japan International Research
Center for Agricultural Sciences,
1997
Tepfer, M (ed.). Springer-Verlag,
Berlin, Germany, 1997
Sduemaiui, J. and R. Casper
(Organizers)
Biologische Bundesanstalt fur
Land- und Forstwirtschaft
Braunschweig, Germany
-------
Appendix G Biotechnology references recommended by the USEPA Biotechnology Steering
Committee.
-------
April 11.2002
C:\WlNDOWS\TEMP\c.lotus.notes.data\references and other resources, wpd
Biotechnology References
Recommended By the Biotechnology Steering Group
1. EPA's Bt crops reassessment document. All together it exceeds 400 pages so you might want
to start with the Overview. The Science Assessment Chapters are a thorough review of
what we had for the initial registration and what we have learned since. See
http://www.epa.gov/pesticides/biopesticides/reds/brad_bt_pip2.htm. Also, Janet
Andersen can provide a wordperfect file of the document
2. USDA Biotechnology Risk Assessment Research Grants Program -
http://www.reeusda.gov/1700/funding/brargp.htm
The workshop proceedings and contributed papers from Ecological and Agronomic
Consequences of Gene Flow from Transgenic Crops to Wild Relatives can be accessed at
http://www.biosci.ohio-state.edu/~lspencer/gene flow.htm. This workshop was funded
by USDA biotech risk assessment research grants program and Chris Wozniak from
BPPD attended. Chris has a great deal of expertise in this area.
3. EC-supported research into the safety of Genetically Modified Organisms -
http://europa.eu.int/comm/research/qualitv-of-life/gmo/
4. USDA's Biotechnology Risk Assessment Research Grants Program Home Page (look for the
summaries of sponsored research) -
http://www.reeusda.gov/crgam/biotechrisk/biotech.htm
(Also found on this site are three summaries of research in 1994,1995 and 1996
sponsored by USDA, EPA and Environment Canada.)
5. Genetically Modified Pest-Protected Plants: Science and Regulation (NAS) —
http://www.nap. edu/books/0309069300/html/
6. Environmental Effects of Transgenic Plants: The Scope and Adequacy of Regulation -
http://www.nap.edu/catalog/10258.html
7.The Plant Journal - http://www.blackwell-science.com/tDi/qm
(NB: The Kuiper et al. paper on biotech foods).
8. Information Systems for Biotechnology
http://www.isb.vt.edu/news/2002/news02.Apr.html
(NB: This latest issue (April 2002) of the ISB News contains a summary of the gene flow
conference held in Columbus, Ohio recently and a summary of the recent series of reports
in PNAS on non-target effects from Bt corn. Some might be interested in scanning
earlier issues in the ISB archives.)
-------
e. Nelson, GC (2001) Genetically Modified Organisms in Agriculture: Economics and
Politics, Academic Press, New York.
f. NRC (National Research Council) (2002) Environmental Effects of Transgenic Plants:
The Scope and Adequacy of Regulation, National Academy Press, Washington
DC
g. NRC (National Research Council) (2000). Transgenic Plants and World Agriculture,
National Academy Press, Washington DC
h. Shelton AM, Zhao J-Z, Roush RT (2002) Economic, Ecological, Food Safety, and
Social Consequences of the Development of Bt Transgenic Plants, Ann. Rev.
Entomol. 47: 845-881.
i. Wolt, JD, R.K.D. Peterson (2000) Agricultural Biotechnology and Societal Decision-
Making: The Role of Risk Analysis, AgBioForum 3(l):291-298.
12. Ecological Effects
a. NRC (National Research Council) (1996) Ecologically Based Pest Management,
National Academy Press, Washington DC
b. NRC (National Research Council) (2001) Ecological Monitoring of Genetically
Modified Crops: A Workshop Summary, National Academy Press, Washington,
DC.
c. Nielsen KM, Bones AM, Smalla K, van Elsas JD (1998) Horizontal Gene Transfer
from Transgenic plants to Terrestrial Bacteria - A rare Event? FEMS Microbiol.
Revs. 22: 79-103.
d. Rissler, J, Mellon M (1996) The Ecological Risks of Engineered Crops, MTT Press,
Cambridge, MA
13. Resistance Mangement
a. Binns MR, Nyrop JP, van der Werf W (2000) Sampling and Monitoring in Crop
Protection, CABI Publishing, New York.
b. Bourguet D, Genissel A, Raymond M (2000) Insecticide Resistance and Dominance
Levels, J. Econ. Entom. 93:1588-1595.
c. Caprio MA (2001) Source-Sink Dynamics Between Transgenic and Non-Transgenic
Habitats and Their Role in the Evolution of Resistance, J. Econ. Entom. 94: 698-
705.
3
-------
q. McKensie JA (1996) Ecological and Evolutionary Aspects of Insecticide Resistance,
RG Landes Co., Austin TX.
r. Mitchell PD, Riedell WE (2001) Stochastic Dynamic Population Model for Northern
Corn Rootworm (Coleoptera: Chrysomelidae), J. Econ. Entom. 94:599-608.
s. NRC (National Research Council) (2000) Genetically Modified Pest-Protected Plants:
Science and Regulation, National Academy Press, Washington, DC.
t. Pedigo LP, Buntin GD (1994) Handbook of Sampling Methods for Arthropods in
Agriculture, CRC Press, Boca Raton FL.
u. Siegfried BD, Spencer T, Nearman J (2000) Baseline Susceptibility of the Com
Earworm (Lepidoptera: Noctuidae) to the CrylAb Toxin from Bacillus
thuringiensis, J. Econ. Entom. 93:1265-1269.
v. Sivakumar MVK, Gommes R, Baier W (2000) Agrometeorology and sustainable
agriculture, Agric. Forest Meteorol. 103: 11-26.
w. Thomas, M. B. (1999) Ecological approaches and the development of "truly
integrated" pest management, Proc. Natl. Acad. Sci. USA 96:5944-5951.
x. Venette RC, Moon RD, Hutchison WD (2002) Strategies and Statistics of Sampling
for Rare Individuals, Ann. Revs Entomol. 47: 143-174.
y. Williamson, M. 1996. Can the Risks from Transgenic Crop Plants be Estimated?
Trends Biotechnol. 14:449-450.
5
-------
TECHNICAL REPORT DATA
(Pfease read Instructions on the reverse before completing)
1. REPORT NO.. 2.
£PAHoo]£'03!oz4
3. RECIPIENTS ACCESSION NO.
4. TITLE ANCf SUBTITLE ^ .
fcesettrch plan .* Effects of Cnem/daf rrzyp/ctdes
d>ndL Alon-tdrged
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7 AUTHORS . , . ,->/-*
-D&vid. M. Ofstyk., Ph.D.
8. PERFORMING ORGANIZATION REPORT NO.
IV2P-03-/I3
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Western jzostoqy Divfsfo^
fl&b tonal Healm * Environ. Effects im-
M Stf. 3S*- .Sb.\ CorMttis.GH 47333
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME A^D ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
@j£S£Arch Plan
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This project supports EPA's mission to protect human health and to safeguard the natural environment — air, water, and
land — upon which life depends. Specifically, we address EPA's responsibility to prevent pollution and reduce the impacts
from pollution to communities and ecosystems (Government Performance and Results Act (GPRA) Goal 4, "Safe
Communities")- To achieve this goal, EPA's Office of Prevention, Pesticides, and Toxic Substances (OPPTS) requires
scientifically credible information and methods for use in assessing health and ecological risks from products used in
commerce, including chemical pesticides and genetically engineered plants. OPPT regulates chemical and biological
pesticides primarily under the Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) administered through the Office
of Pesticide Programs (OPP). Other acts and programs, especially the Toxic Substances Control Act (TSCA), and the
Federal Food, Drug and Cosmetic Act (FFDCA) are administered by OPPTS's Office of Pollution Prevention and Toxics
(OPPT) to provide for protection of the environment from chemicals and biological pesticides. In the past, protection of
ecological resources has received minimal attention under these regulations compared to concerns regarding impacts on
human health. Recently, however, awareness of adverse effects from drift of new low-dose high-toxicity herbicides to non-
target crops and native vegetation has heightened awareness of the need to improve tests for effects of chemical herbicides
to plants. Similarly, public concern regarding the release of genetically engineered plants and the adoption of the "Final
Rules and Proposed Rules for Plant-Incorporated Protectants" (40CFR Parts 152 and 174) have increased the need for tools
to evaluate the risks from engineered plants and gene flow from engineered crops to other plant species. Thus, OPP and
OPPT need tools to assess ecological risks from transgenic crops, improved methods for spatially explicit ecological risk
assessments, new methods to provide for efficient and effective gathering and interpretation of herbicide hazard
identification and dose-response data, and investigations of the potential effects of high priority hazards.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Grouo
hltih plants jrisk assessment/
qetdfc- ejagmeen«3/
farrestna ( plants
18 DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report)
21. NO. OF PAGES
20. SECURITY CLASS (This Page)
22. PRICE
EPA Form 2220-1 (Rev. 4-77 ) PREVIOUS EDITION IS OBSOLETE forms/adminrteairpt.frm 7/8/99
------- |