Findings: Non-Target Plant
Risk Assessment Workshop
for Regulators
Prepared for
Environmental Fate and Effects Division
Office of Pesticide Programs
U.S. Environmental Protection Agency
Washington D.C.
Prepared by
ARCADIS
May 6, 2002
£? ARCADIS
Infrastructure, buildings, environment, communications
-------
-------
Appendices
Contents
A. List of Acronyms and Abbreviations
B. Non-target Plant Risk Assessment Workshop Participant List
C. Issues Paper: Non-target Plant Risk Assessment Workshop for Regulators
D. Presentations Given at Non-Target Plant Risk Assessment Workshop
E. Non-target Plant Risk Assessment Workshop Agenda
-------
Appendix A
List of Acronyms and Abbreviations
A.I.
Active ingredient
APHIS
Animal and Plant Health Inspection Service
BBA
Federal Biological Research Centre for Agriculture and Forestry
EEC
Expected Environmental Concentration
EC25
The concentration that results in a 25% reduction in the test endpoint being measured
relative to the control
EC50
The concentration that results in a 50% reduction in the test endpoint being measured
relative to the control
Ecx
The concentration that results in a reduction of "x" percent in the test endpoint being
measured relative to the control.
EPA
Environmental Protection Agency
EPPO
European and Mediterranean Plant Protection Organization
EU
European Union
EXAMS
Exposure Analysis Modeling System (a model)
FIFRA
Federal Insecticide, Fungicide and Rodenticide Act
GENEEC
Generic Estimated Environmental Concentration (a model)
GIS
Geographical Information Systems
GLP
Good Laboratory Practices
Kd
Soil-water partition coefficient
Koc
Octanol-water partition coefficient
LOC
Level of Concern
NOEC
No observed effect concentration
OPPTS
Office of Prevention, Pesticides and Toxic Substances
OPP
Office of Pesticide Programs
OPPT
Office of Pollution Prevention and Toxics
PEC
Predicted Environmental Concentration
PMRA
Pest Management Regulatory Agency
PRZM
Pesticide Root Zone Model (a model)
RQ
Risk Quotient
T&E
Threatened & Endangered
TER
Toxicity Exposure Ratio
TAI
Technical active ingredient
TGAI
Technical grade active ingredient
TSCA
Toxic Substances Control Act
USD A
U.S. Department of Agriculture
USGS
U.S. Geological Survey
-------
Appendix B
Non-target Plant Risk Assessment Workshop Participant List
-------
NON-TARGET PLANT RISK ASSESSMENT WORKSHOP PARTICIPANTS LIST
NAME
Phone
Number
Fax Number
Business Address
E-mail Address
Jose Luis Alonso-Prados
34-91-347-14-79
same
INIA-Pcsticidc Group
Department Proteccion Vegetal Ctra.
Corufia km 7.2 2800
Madrid, Spain
prados@inia.es.
Karl Arne
206-553-2576
206-553-1775
US EPA
Pesticide Specialist
Office of Ecosystems and Communities
EPA Region 10 (ECO-084)
1200 6,h Ave
Seattle, WA 98101
Arnc.karl@cpa gov
Paul Ashby
0190 55794
Pesticides Safety Directorate (UK)
Pesticides Regulatory Specialist
(Ecotoxicology)
Mallard House
Kings Pool 3
Peasholme Green York, UK Y0I 7PX
Paul.Ashby@psd.defra.gsi.gov.uk
Terrell (Tern) Barry
916-324-4140
916-324-4088
California Department of Pesticide Regulations
Senior Environmental Research Scientist
Environmental Monitoring Branch
P.O. Box 4015
Sacramento, CA 95812
tbarry@cdpr.ca.gov
Web site: www.cdpr.ca.gov
David Bays
703-605-0216
703-308-6466
U.S. Environmental Protection Agency
Ariel Rios Building (7503C Rm 911AA)
1200 Pennsylvania Avenue
Washington, DC 20460
Bays.David@cpa.gov
Norman Birchficld
703-605-0582
703-305-6309
U.S. Environmental Protection Agency
Ariel Rios Building (7507C Rm 102IE)
1200 Pennsylvania Avenue
Washington, DC 20460
Birchficld.Norman@epa gov
-------
NAME
Phone
Number
Fax Number
Business Address
E-mail Address
James Carlcton
703-305-5736
703-305-6309
U.S. Environmental Protection Agency
Ariel Rios Building (7507C Rm I008Q)
1200 Pennsylvania Avenue
Washington, DC 20460
Carlcton Jamcs@cpa.gov
Michael (Mike) Davy
703-305-7081
703-305-6309
U.S. Environmental Protection Agency
Ariel Rios Building (7507C Rm I012F)
1200 Pennsylvania Avenue
Washington, DC 20460
Davy Michael@epa.gov
Damn Dantin
850-934-9383
850-934-2406
U S Environmental Protection Agency
Gulf Ecology Division
1 Sabine Island Drive
Gulf Breeze, FL 32561
Dantin.Darrin@epa gov
Neils Elmcgaard
45 89201585
45 89201414
Senior Biologist
National Environmental Research Institute
Department of Terrestrial Ecology
P. O. Box 314
Vcjlsovcj 25
DK-8600 Silkcborg
Denmark
ne@dnui.dk.
Philip Errico
703-305-6663
703-308-1825
U.S. Environmental Protection Agency
Ariel Rios Building (7505C Rm 241)
1200 Pennsylvania Ave
Washington, DC 20460
Errico.philp@cpa gov
James Fairchild
573-876-1871
573-876-1896
U.S. Geological Survey
4200 New Haven Rd
Columbia, MO 65201
Jamcs_Fairchild@usgs.gov
John Fletcher
541-754-4374
541-754-4799
US EPA WED
200 SW 35th St
Corvallis, OR 9733
Fletcher John@cpa gov
jflctchcr@ou.edu
-------
NAME
Phone
Number
Fax Number
Business Address
E-mail Address
Derek Francois
613-736-3726
613-736-3710
Pest Management Regulatory Agency
(PMRA)
Sir Charles Tuppcr Building
2720 Riverside Drive
Postal Locator 6604E
Ottawa, ON, Canada
K1AOK9 Canada
Dcrek_Francois@hc-sc gc.ca
Patrick (Jerry) Hannun
703-305-5190
703-305-6409
U.S. Environmental Protection Agency
Ariel Rios Building (7507C-Rm 1008M)
1200 Pennsylvania Avenue
Washington, DC 20460
Hannan. Patrick@epa.gov
Russell Jones
703-308-5071
703-308-7026
U S. Environmental Protection Agency
Ariel Rios Building (7511C-9I0W25)
1200 Pennsylvania Avenue
Washington, DC 20460
Jones.Russell@epa.gov
Thomas Kopp
703-305-5405
703-305-6309
U.S. Environmental Protection Agency
Ariel Rios Building (7507C-Rm 1008B)
1200 Pennsylvania Avenue
Washington, DC 20460
Kopp.Thomas@epa.gov
Ted Kuchnicki
613-736-3733
613-736-3710
Pest Management Regulatory Agency
(PMRA)
2720 Riverside Drive
Postal Locator 6604E
Ottawa, ON, Canada
KIAOK9 Canada
Ted_Kuchnicki@hc-sc gc.ca
Christine Kula
49 531 299 3611
49 531 299 3605
Federal Biological Research Centre for
Agriculture and Forestry
Department of Plant Protection Products and
Application Techniques
Biology Division
Mcsscweg 11/12
D 38104 Braunschweig
Germany
C.K.ula@BBA.DE
-------
NAME
Phone
Number
Fax Number
Business Address
E-mail Address
Leo Lasota
703-308-8126
703-308-8091
U.S. Environmental Protection Agency
Herbicide and Insecticide Branch
Biological and Economic Analysis Division
Ariel Rios Building (7503C Rin 911AA)
1200 Pennsylvania Avenue
Washington, DC 20460
Lasota. leo@cpa gov
Elizabeth Leovcy
703-305-7328
703-305-6309
U S. Environmental Protection Agency
Ariel Rios Building (7507C)
1200 Pennsylvania Avenue
Washington, DC 20460
Leovcy. El izabcth@cpa gov
Michael Lewis
850-934-9382
850-934-2406
U.S. Environmental Protection Agency
USEPA Environmental Effects Research Lab
Gulf Ecology Division/ORD
Sabine Island Drive
Gulf Breeze, FL 32561-5299
Lewis.Michael@epa.gov
Robert Luttik
31 30 2742795
31 30 2744401
National Institure for Public Health and the
Environment (RIVM)
Centre for Subsiances and Risk Assessment
P O Box 1
3720 BA Bilthoven
The Netherlands
Robert Luttik@rivm ill
Nicholas Mastrota
703-305-5249
703-305-6309
U.S. Environmental Protection Agency
Ariel Rios Building (7507C Rm I034P)
1200 Pennsylvania Avenue
Washington, DC 20460
Mastrota Nicholas@cpa gov
Brian Montague
703-305-638
703-305-6309
U.S. Environmental Protection Agency
Ariel Rios Building (7507C Rm 1034V)
1200 Pennsylvania Avenue
Washington, DC 20460
Montague Brian@cpa gov
Edward (Ed)
Odcnkirchen
703-305-6449
703-305-6309
U S. Environmental Protection Agency
Ariel Rios Building (7507C Rm 1034T)
1200 Pennsylvania Avenue
Washington, DC 20460
Od c n k i rch e n. Ed wa rd@e pa. go v
-------
NAME
Phone
Number
Fax Number
Business Address
E-mail Address
David Olyzk
541-754-4397
541-754-4799
U.S. Environmental Protection Agency
WED US EPA
200 SW 35th St.
Corvalhs, OR 97333
olszyk.david@mercury.cor.epa.
gov
Rick Pctrie
703-305-7358
703-305-6309
U.S. Environmental Protection Agency
Ariel Rios Building (7507C Rm 1012L)
1200 Pennsylvania Avenue
Washington, DC 20460
Pctric.Rick@epa.gov
Thomas Pflecgcr
541-754-4374
541-754-4799
US EPA WED
200 SW 35th St
Corvalhs, OR 9733
Pflccger.thomas@cor.uscpa.gov
Daniel Rieder
703-305-5314
703-305-6309
U.S. Environmental Protection Agency
Ariel Rios Building (7507C Rm 102 IF)
1200 Pennsylvania Avenue
Washington, DC 20460
Rieder.Daniel@cpa.gov
Kent Smith
202-720-3186
202-720-3191
United States Department of
Agriculture
Office of Pest Management Policy (OPMP)
Agricultural Research Service (ARS)
1400 Independence Ave, SW
Room 3859 , South Ag Building
Wahington, DC 202560-0315
KSmith@ars.usda.gov
Web site: www.ars.usda.gov/opmp
Jerry Smrchek
202-564-7628
202-564-7430
U.S. Environmental Protection Agency
Ariel Rios Building (7403 M)
1200 Pennsylvania Avenue
Washington, DC 20460
Smrchek.Jcrry@epa.gov
Jane Stavcley
919-544-4535 Ext.
238
919-544-5690
ARCADIS
4915 Prospectus Drive, Suite F
Durham, NC 27713
jstaveley@arcadis-us.com
lngrid Sunzenaucr
703-305-5196
703-305-6309
U.S. Environmental Protection Agency
Ariel Rios Building
1200 Pennsylvania Avenue
Washington, DC 20460
Sunzenaucr.lngrid@cpa.gov
-------
NAME
Phone
Number
Fax Number
Business Address
E-mail Address
Chris Lcc-Stccrc
61 2 6250 0256
62 2 6250 7554
Risk Assessment and Policy Section
Environment Australia CATEB
GPO Box 787
Canberra, ACT
Australia 260-1
Chris.Lcc-Stccre@ea.gov.au
Stephanie Syslo
703-305-6355
703-305-6309
U.S. Environmental Protection Agency
Ariel Rios Building (7507C Rm 1021 F)
1200 Pennsylvania Avenue
Washington, DC 20460
Syslo.Stephanic@cpa gov
Dick Watkins
214-665-6591
214-665-7263
Louisiana State Project Officer
Pesticide Applicator and Certification and
Training Program Coordinator
United States Environmental Protection
Agency
Region 6 (6PD-P)
1 1445 Ross Ave., Suite 1200
Dallas, TX 75202
Watkins.dick@cpa.gov
-------
NON-PARTICIPANT SUPPORTERS OF THE NON-TARGET PLANT RISK ASSESSMENT WORKSHOP
NAME
Phone
Number
Fax Number
Business Address
E-mail Address
Harold Coble
919-513-2124
919-513-1995
Agronomist
U.S. Food and Drug Administration
1017 Main Campus Drive
Suite 2500
Raleigh, NC 27606
harold_coblc@ncsu.cdu
Andy Hart
44 (0)1904 462053
44 (0) 1904 462111
Wildlife Ecotoxicology
Unit 5129 2053
Central Science Laboratory
Sand Hutton
York Y04I 1 LZ
UK
a.hart@csl gov.uk
Website: www.csl.gov.uk
Christian Kjaer
45 89 20 14 00
45 89 20 14 14
Department of Terrestrial Ecology
P O. Box 314
Vcjlsovej 25
DK-8600 Silkeborg
Denmark
ckj@dmu dk
Frank dc Jong
31 30 2742795
31 30 2744401
National Institute for Public Health and the
Environment (RIVM)
Centre for Substances and Risk assessment
P O.Box 1
3720 BA Bilthovcn
The Netherlands
Frank. dejong@RIVM.NL
-------
Appendix C
Issues Paper: Non-target Plant Risk Assessment Workshop for Regulators
-------
Issues Paper:
Non-Target Plant Risk Assessment Workshop for Regulators
01/09/02
Prepared for
and
Modified by:
Environmental Fate and Effects Division
Office of Pesticide Programs
U.S. Environmental Protection Agency
Washington D.C.
Prepared By:
Jane Staveley
ARCADIS G&M, Inc.
-------
1.0 Introduction, Purpose and Goals of Workshop 1
1.1 International harmonization 2
1.2 Trends in ecological risk assessment for pesticides 2
2.0 Issues for Discussion 4
2.1 What is the best way to conduct a non-target plant risk assessment? 4
2.2 Which pesticides/uses require assessment? 4
2.3 Applicability of the ecological risk assessment framework to non-target plants . . 4
2.4 Problem formulation 5
2.5 The use of a "tiered" approach 5
2.6 Chemical properties 6
2.7 Exposure assessment 6
2.8 Effects assessment 6
2.9 Risk characterization and regulatory applications 9
2.10 On-going research 10
2.11 Research needs 10
2.12 Action items 10
3.0 Non-target Plant Risk Assessment: Background and Summary of Recent Events 11
3.1 History of non-target plant risk assessment in the United States (U.S. EPA) ... 11
3.1.1 Pesticides 11
3.1.2 Industrial chemicals 13
3.2 History of non-target plant risk assessment in Canada 14
3.3 Recent efforts in North America 15
3.3.1 Ecological Incident Information System (EIIS) 15
3.3.2 1999 ILSI Workshop 15
3.3.3 OPP Efficacy data 16
3.3.4 Identified OPP Research Needs 16
3.4 Approaches from other countries 17
3.4.1 European Union 17
3.4.2 European and Mediterranean Plant Protection Organization (EPPO) ... 18
3.4.3 Ministry of Agriculture, Fisheries and Food (MAFF) 20
4.0 Appendix 23
4.1 List of Acronyms 23
4.2 References 25
4.3 ILSI Workshop Summary and Recommendations 28
i
-------
List of Figures
Figure 1. The framework for ecological risk assessment 35
Figure 2. Conceptual Model for Terrestrial Non-Target Plants 36
Figure 3. Simplified diagram of the sub-scheme for evaluation
of the risk of a plant protection product to higher plants 37
Figure 4. Conceptual Model of the Source, Transport, Fate and Effects of
Low Dose, High Toxicity Herbicides on Non-Target Species and Ecosystems 38
List of Tables
Table 1. Comparison of approaches for aquatic non-target plants 39
Table 2. Comparison of approaches for terrestrial non-target plants 40
Table 3. Comparison of assumptions and models used for exposure assessment 41
ii
-------
1.0
IiNTRODUCTlOiN, PURPOSE \ND GOALS OK WORKSHOP
The Office ofPesticide Programs (OPP) and the Office of Pollution Prevention and Toxics
(OPPT) in the U.S. Environmental Protection Agency (EPA) are convening an international
workshop to explore various regulatory approaches for assessing the risk of chemicals to non-
target plants. In North American countries, risk to non-target plants has historically received
less attention than risk to aquatic and terrestrial animals. As a result of this imbalance,
methodologies and approaches for assessing risk to non-target plants are less advanced
(Benenati, 1990; Boutin et al., 1995: Fletcher, 1991; Kapustka, 1997; Lewis. 1990; Wang and
Freemark, 1995). A similar situation exists in most European countries where data on the
toxicity of chemicals to plants has not been routinely required (OECD, 1994).
During the last decade, many regulatory agencies in North America and Europe have received an
increased number of incident reports concerning adverse effects to non-target plants from the use
of pesticides. At the same time, scientific advisory groups, stakeholders, and researchers have
urged regulators to produce more realistic and scientifically sound risk assessments and to reduce
the uncertainty in their risk assessments. This increased interest in risks to non-target plants
coupled with an emphasis on producing more refined risk assessments, such as probabilistic risk
assessments (ECOFRAM, 1999a; 1999b). presents regulators with an exciting opportunity to
advance the risk assessment approach for non-target plants.
This international workshop is being convened to gather relevant information from regulatory
bodies and scientists in various countries throughout the world and to promote sharing of
information and risk assessment methods. The purpose and goals of the workshop are as
follows:
improve communication and collaboration among the international community regarding
regulatory approaches for conducting risk assessments for non-target plants;
discuss the elements of a risk assessment framework and the issues associated with
exposure and ecological assessments for non-target plant risk assessments;
identify and prioritize research topics which will address the uncertainties in non-target
plant risk assessments;
develop a set of recommendations for advancing non-target plant risk assessments.
Although the workshop is focused on non-target plant risk assessment for pesticides, the
discussion of approaches and the recommendations resulting from the workshop are expected to
be applicable to non-pesticidal chemicals.
This document is divided into four chapters. Chapter 1 describes the reasons for convening the
workshop, while Chapter 2 frames the issues that will be discussed in the workshop. Chapter 3
summarizes existing and developing approaches, including activities and issues that have been
raised in the U.S., Canada, and Europe. Background material along with references and a list of
acronyms are provided in Chapter 4.
1
-------
1.1
International harmonization
Evaluating and managing the risks of pesticides to the environment is the responsibility of
regulatory authorities in a number of countries. While it may be difficult to develop one
globally harmonized approach for evaluating and regulating pesticide risk, it is useful to
determine what commonalities exist in the various approaches and work towards developing
more compatible approaches The benefits of producing harmonized methods and approaches
are many, including a saving of resources, improved communication and scientific assessments,
and consistent and improved regulatory decisions. Under the North American Free Trade
Agreement (NAFTA), the United States and Canada have agreed to harmonize their testing
requirements and to share resources m reviewing data. The U.S. EPA and the Canadian Pest
Management Regulatory Agency (PMRA) have made significant progress towards harmonization
in a number of areas, and both agencies are also participating in international harmonization
efforts through the Organization for Economic Co-operation and Development (OECD). With
the increased globalization of environmental and economic issues, international collaboration in
this area will benefit both the regulators and the regulated community.
1.2 Trends in ecological risk assessment for pesticides
The term "ecological risk assessment" may be defined differently in different countries, and
sometimes within the same country. It has been defined as "the process that evaluates the
likelihood that adverse ecological effects may occur or are occurring as a result of exposure to
one or more stressors3' (U.S. EPA, 1998) or ::the practice of determining the nature and
likelihood of effects of [human] actions on animals, plants and the environment" (SETAC,
1997). In Europe, the term "environmental risk assessment" is sometimes used.
In the United States, the "EPA Framework for Ecological Risk Assessment" (U.S. EPA, 1992)
has served as a model for the development of a more quantitative assessment process, followed
by the development of ecological risk assessment guidance ("Guidelines for Ecological Risk
Assessment," U.S. EPA, 1998). As described in the Framework document, the overall risk
assessment (ERA) process has three phases: Problem Formulation, Exposure and Effects
Analyses, &nd Risk Char&cteriza:ion (see Figure I). Related risk assessment and risk
management frameworks have also been developed in other countries as well (Power and
McCarty, 1998).
Historically, the U.S. pesticides program has focused on exposure and effects analysis, with less
attention given to the problem formulation and risk characterization phases. (The risk
characterization phase describes the likelihood and magnitude of adverse effects on multiple
species as a result of exposure to a stressor or multiple stressors). With the publication of the
EPA Framework, the Agency was directed to incorporate problem formulation and risk
characterization into the risk assessment process. In the past, the problem formulation phase has
2
-------
not been historically stated explicitly, but recent comments from the Scientific Advisory Panel
(SAP) suggest that it should be an integral part of the process as well.
In the United States, significant progress has been made in advancing ecological risk assessment
methods relative to pesticide regulation In response to a 1996 FIFRA Scientific Advisory Panel.
EPA has been working towards developing tools and methodologies to move beyond a
deterministic approach and towards a probabilistic assessment of risk (U.S. EPA. 2000a; U.S.
EPA, 2000b; U.S. EPA, 2000c). The Ecological Committee on FIFRA Risk Assessment
Methods (ECOFRAM). a group comprised of experts from government agencies, academia,
contract laboratories, environmental advocacy groups, and industry, worked on identifying and
developing probabilistic tools and methods for terrestrial and aquatic assessments under the
FIFRA regulatory framework. The effort was divided into two work groups, aquatic and
terrestrial, each of which released their draft findings in 1999. With the exception of algae,
though, consideration of non-target plants was intentionally omitted from the ECOFRAM
process. Although the U.S. EPA is developing an implementation plan that incorporates
probabilistic tools and methods for conducting ecological risk assessments for pesticides, it is not
focusing on plants at this time.
-i
J
-------
2.0
Issues for Discussion
2.1 What is the best way to conduct a non-target plant risk assessment?
This may require iwo separate answers: ideally, how would you approach plant risk assessment
and realistically (e.g.. wuh current knowledge and data limitations), what is the best way to
approach it?
2.2 Which pcsticidcs/uses require assessment?
If there is no potential for environmental exposure to pesticides, an assessment of ecological
effects i$. not needed. Restricted uses, such as in closed-system greenhouses, indoors, and
swimming pools do not trigger plant testing in the U.S or in Canada. In Canada, testing is
required for uses in open-system greenhouses and forestry, for outdoor domestic and industrial
uses, and for nondomestic uses on food and nonfood crops in terrestrial and aquatic
environments. Data requirements in the U.S. are similar, except that currently phytotoxicity data
are not required if the pesticide is a non-herbicide or a water-insoluble fungicide applied to
food/feed crops, and has low volatility and low water solubility. What criteria are used in other
countries to eliminate phytotoxicity as a concern due to a lack of exposure potential?
Is it sufficient to eliminate pesticides from further consideration of risk to non-target plants based
upon their uses9 Would other decision criteria be useful? These might include physical-
chemical properties. QSAR evaluation, mode of action, mode of uptake, application rates and
methods, number and timing of applications, and known information about fate and transport in
aquatic and terrestrial environments.
In 1994, the SAP endorsed phytotoxicity testing for all pesticides having outdoor use sites. This
means screening of insecticides and fungicides as well as herbicides for phytotoxicity. Is this
screening approach supportable?
2.3 Applicability of the ecological risk assessment framework to non-target plants
A number of stakeholders have recommended that pesticide risk assessment for non-target plants
be conducted within an ecological risk assessment context, such as the EPA Framework (U.S.
EPA, 1992; U.S. EPA, 1998), and OPP is mandated to follow this guidance. What is your
opinion of this Framework? Are there other ecological risk assessment schemes that should be
considered, and how do they compare or contrast with the EPA Framework?. What sort of
modifications to the EPA Framework would be appropriate?
What is the role of stakeholder input to non-target plant risk assessments? How can flexibility be
incorporated into the risk assessment process to address risk management concerns, e.g..
economic plants, endangered species?
4
-------
2.4
Problem formulation
According to the EPA Framework, the first phase of the ecological risk assessment process is the
problem formulation phase This phase is a planning and scoping process that establishes the
goals, breadth, and focus of the risk assessment. The end product is a conceptual model that
identifies the environmental values to be protect (the assessment endpoints), the data needed, and
the analyses to be used. In this first stage, there will probably be different data requirements for
different pesticides. The logic underpinning the problem formulation step will need to be clearly
articulated to provide regulators with a useful tool and also allow the regulated community to
understand what information needs to be submitted for a particular product. Furthermore, for
comparative risk assessments, it may be necessary to specify a consistent base set of information
for all pesticides. Is this base set necessary and what would it consist of?
In problem formulation, the goals are developed, the assessment endpoints are selected, the
conceptual model is developed, and an analysis plan is prepared. What would a generic problem
formulation look like for non-target plants? What level of protection would be considered
adequate for non-target plants?
What are the goals of a problem formulation, i.e., what are we trying to protect? How do we
define non-target plants? Why are non-target plants important?
What are the appropriate assessment or analysis endpoints? (Note: assessment endpoints have
been defined as "explicit expressions of the actual environmental value that is to be protected,
operationally defined by an ecological entity and its attributes," U.S. EPA, 1998). To answer this
question, we may need to determine what level of organization is important. For example,
individuals may be important for threatened and endangered species, while populations and/or
communities are important for other species. What are the spatial and temporal scales we should
be concerned about? What is the potential for recovery and what level of protection is desired
over scale of organization, space and time?
A conceptual model is a written description and visual representation of predicted relationships
between ecological entities and the stressors to which they may be exposed. Several examples of
conceptual models for non-target plant risk assessment are attached (Figures 2, 3 and 4). Does
the development of a conceptual model have utility? What should a generic model include?
2.5 The use of a "tiered" approach
There is widespread agreement that an iterative or "tiered" approach is useful for assessing risk
to non-target organisms. In a tiered approach, generic and screening-type data can be used at
lower tiers, more specific and definitive data at intermediate tiers, and highly refined information
at the higher tiers. A tiered approach can also be used within a deterministic risk assessment. At
what point in the risk assessment, should additional data be required? What are the triggers for
5
-------
advancing from a lower tier to a higher tier0 How can flexibility and how much of it should be
built into the tiers?
2.6 Chemical properties
In general, the approaches used currently, e U.S.EPA and EPPO. state that the technical grade
of the active ingredient is tested at lower tiers, while the formulated product is tested at higher
tiers. What are the advantages and disadvantages of each, and what criteria are important in
making the decision as one moves through the tiers'7
What sort of information on the properties of the pesticide (mode of action, mode of uptake, etc )
can be used to better refine the direction of the risk assessment, or be used as decision points to
step out of the process?
2.7 Exposure assessment
Should the exposure assessment be performed entirely separate from the effects assessment, or
should both be performed together?
How is the exposure assessment conducted? Is the assumption of 100% of the maximum
application rate appropriate, or some lower percentage? What generic assumptions are used?
What improvements can be suggested to provide alternatives to the maximum exposure estimates
in a deterministic assessment, e.g., consideration of long-range transport?
Should an in-crop scenario and an out:of-crop scenario be considered separately? (See
discussion in Hart, 1999) Or should the exposure values be aggregated before comparing them to
toxicity values, and if so, how should this be done?
For deterministic risk assessment, what types of models are currently available for creating
distributions of exposure concentrations? (Some of the matrices to be considered include air,
soil, and water). What are the input data needs, and how is the output presented? How do these
models/approaches account for spatial and temporal variability? What sort of scale (local,
regional, landscape) is addressed? Can we compare and contrast these approaches, e.g.,
AgDRIFT vs. Ganzelmeier?
2.8 Effects assessment
A number of protocols are available for conducting toxicity tests with aquatic and terrestrial non-
target plants. Some of these protocols have been used routinely for many years, e.g., population
growth tests for microalgae and duckweed, seedling emergence and vegetative vigor tests for
terrestrial crop species, while other protocols are in early stages of development and validation,
e.g., tests with submersed and emergent aquatic plants, reproduction or life-cycle tests for
6
-------
terrestrial plants. In addition, there is little information available on the use of certain plant
groups such as mosses, ferns, liverworts or conifers in toxicity testing.
What characteristics are important in selecting a toxicity test species? For example:
geographic occurrence is widespread or occurs in the area of pesticide use;
representative of species likely to be exposed (occupies similar niche and has similar
means of exposure);
data exists on its use as a test organism;
conditions for acceptable growth in laboratory or greenhouse are known and can be
simulated;
unusual equipment, procedures or conditions are not required;
its response is easy to observe using readily available techniques;
more than one response parameter can be observed;
known to be sensitive to at least some chemicals;
the coefficient of variation for response parameters is low in the controls (so that effects
can be distinguished)
Please comment on these and list other factors as appropriate.
What types of tests are most useful for non-target plant risk assessment? What new protocols
should be considered? What types of effects should be measured (e.g., shoot height, shoot
weight, visual effects, functional measures, etc.)? Should laboratory growth conditions be
optimal or realistic? Do the measurement endpoints have objective indices that can be reliably
and reproducibly quantified?
Are appropriate exposure routes, e.g., aqueous, foliar, soil, used in existing test protocols? Are
any important routes of exposure currently being missed for a particular group of plants? For
example, is it necessary to include foliar exposure to floating aquatic plants or soil pore-water
exposure for terrestrial plants? How important is timing of application, e.g., flowering period?
Should efficacy screening data generated by registrants be used? Why or why not? Can we
extrapolate data to other species of plants? How do other countries use efficacy data? What is
needed in terms of standardization of screening tests to make this information more usable in a
risk assessment?
How many and which species should be tested? The existing EPA/PMRA approach uses 5
aquatic plants and 10 terrestrial crop plants which serve as surrogates for all plants. The new
testing proposal (Davy et al., 2001) recommends expanding the number of aquatic species to 11
and the number of terrestrial species to 26. What information is needed to assess the required
number of species, and can we determine this information from existing data? What other effect
assessment approach can be used in determining species to be tested? How do we identify key
species without the use of mapping? How do we identify key species more qualitatively?
7
-------
Boutin and Rogers (2000) have reported that more than 10 species should be tested, while
industry data (Mctvelvey et al.. 2001) indicates that 10 is sufficient. The number of species
needed to establish the sensitivity distribution appears to be dependent upon the pesticide. If the
mode of action is common to all plants and the mode of uptake is also non-specific, the response
should fall within a fairly narrow range, e g . industr> data on atrazine. However, a broader
sensitivity distribution would be expected with more specific mode of action/mode of uptake
(MO.A/MOU). Can a basic approach be identified to develop answers to this question?
In selecting a range of species to test, a taxononnc approach has been used most commonly. This
approach is based on the assumption that testing a variety of plant families will provide a better
estimate of the range of effects that can be expected among all plants. Are there approaches
other than the taxonomic one that should be considered? Should a consistent set of species be
tested at the lower levels to allow for comparisons? It has been suggested that flexibility in
selecting the test species should be allowed, and that the selection should be dependent upon
chemical properties and use patterns.
Are there groups of plants that are currently omitted from testing that are ecologically important
or more sensitive to some pesticides than currently tested species, e.g., mosses, ferns, liverworts,
and conifers?
What endpoints should be generated from the toxicity tests, and do they vary for species (algae,
aquatic macrophytes, terrestrial plants). The EC50 has been used for algae and duckweed
because their generation times are relatively short. The more conservative EC25 has been used
for other vascular plants because they have longer generation times and testing is typically
performed on only a portion of their life cycle. What models can extrapolate from individuals to
population level? Should test and assessment endpoints be assessed separately?
Do we need to consider reproductive endpoints? Is the difference between a growth endpoint
versus a reproductive endpoint for a particular species greater than the differences in growth
endpoints between species? (If not, resources should most likely be focused on obtaining growth
endpoint data for more species). What reproductive testing endpoints, e.g., seed formation, pollen
viability, etc., would be important?
Are there other types of effects that need to be considered (physiological and biochemical effects,
such as oxygen evolution, carbon fixation, etc.) that are not being adequately assessed in current
tests that measure gross somatic effects?
Are multi-species tests useful, e.g., microcosm, mesocosm? If so, how can they be used?
What are suitable approaches for extrapolating toxicity between species? Should uncertainty
factors be used in the denominator of a risk quotient in a deterministic assessment, and if so,
should these factors be fixed (as is currently done in the U.S. for risk assessment of industrial
chemicals), or should they vary depending upon the number of species tested (and tests
8
-------
completed)? What basis should be used to determine an uncertainty factor if it is allowed to be
variable? How many species are needed to drop the uncertainty factor? Can the HD5 (Hazardous
Dose for 5% of the total species) approach be used, as suggested by Hart (1999)?
2.9 Risk characterization and regulatory applications
Environmental risk characterization is the summarizing step of risk assessment. It is the
integration of the exposure assessment with the ecological effects assessment into a
comprehensive, scientifically defensible description of the likelihood and extent of adverse
effects to the environment from the use of a pesticide. Key components of the risk
characterization discuss the ecological significance of the adverse effects, including
consideration of the types and magnitudes of the effects, their spatial and temporal patterns, and
the likelihood of recovery; strengths, limitations, and uncertainties of the risk assessment. The
risk characterization should provide risk managers with a clear, concise, and informative
discussion of the potential risk from a pesticide.
Besides the risk quotient, what other information should be used in making risk characterization
decisions, e.g., consideration of the pattern of toxicity?
What are appropriate ways to express uncertainty? Can uncertainty be qualified in a simplistic
manner even at the lower screening levels of assessment? For example, if an ECx is generated
with a 95% confidence interval, can this information be incorporated into the risk quotient in a
deterministic assessment?
How can uncertainty for an endpoint not studied be expressed? How can uncertainty for an
expected effect in a plant not studied be expressed?
What is the best way to present the results of the risk characterization to facilitate risk
management decisions? What is the nature of the risk characterization used by risk managers?
What are some of the limitations of risk assessment in helping us understand potential for
adverse effects to non-target plants.
How can the concept of recovery be incorporated into the risk characterization, or should it be?
What is the role of mitigation and risk reduction measures? How can the risk characterization be
used to provide input to the development of mitigation options?
What is the role of post-registration monitoring? What are the advantages and disadvantages of
obtaining monitoring data?
9
-------
2.10 On-going research
Are you involved in or aware of any research currently in progress that would help advance the
tools and methods for non-target plant risk assessment?
2.11 Research needs
What are the important research initiatives that would advance the state of non-target plant risk
assessment, both in the short term (within the next five years) and the long term (five years or
longer)? Please prioritize these needs in each category'
2.12 Action items
What do you see as the next steps for advancing non-target plant risk assessment? For example:
identify an acceptable screening level assessment that could be considered a "first cut"
obtain agreement on two or three high priority research initiatives for each stage of the
risk assessment and develop scopes of work for these projects
determine model input parameters for each level of refinement in the risk assessment
coordinate input to and involvement with OECD
10
-------
3.0 Non-target Plant Risk Assessment: Background and Summary of Recent
Events
3.1 History of non-target plant risk assessment in the United States (U.S. EPA)
In the U.S. EPA, two of the most important programs for evaluating the hazards posed to aquatic
and terrestrial plants by the use and release of chemical substances in the environment are the
pesticides program and the toxic substances program. In general, pesticides are regulated under
FIFRA (the Federal Insecticide. Fungicide and Rodenticide Act), while toxic substances are
regulated under a separate law. the Toxic Substances Control Act (TSCA). A more detailed
description of these programs is presented below.
3.1.1 Pesticides
The Office of Pesticide Programs (OPP) in EPA is responsible for evaluating the effects of
pesticides on human health and the environment. In their risk assessment process. OPP considers
the risks of pesticides to non-target plants. Requirements for phytotoxicity testing of pesticides
are found in the Code of Federal Regulations (40 CFR Part 158) and are supplemented by the
non-target plant test guidelines, "Pesticide Assessment Guidelines Subdivision J, Hazard
Evaluation: Non-target Plants" (Hoist and Ellwanger, 1982). The Subdivision J guidelines
describe procedures for effects assessment (three tiers of testing), but do not address the exposure
assessment process. The assumptions used in the exposure assessment process are described in
"Hazard Evaluation Division Standard Evaluation Procedure: Ecological Risk Assessment
(Urban and Cook, 1986). As currently required, Tier I testing (for both aquatic and terrestrial
plants) is conducted at a concentration representative of "maximum exposure" conditions. This
concentration is derived from the maximum label application rate or three times the expected
environmental concentration.
Assumptions used to generate the maximum exposure concentration include direct application to
a one acre pond to a depth of 15 cm (for aquatic exposures) and direct application to a 3-cm deep
column of soil with a bulk density of 1.5 g/cmJ (for terrestrial exposures). Overspray exposure is
estimated at 60% of the maximum application rate. Exposure via spray drift, runoff and washoff
is estimated at 10% of the maximum label application rate for exposure to aquatic plants (Urban
and Cook, 1986).
For the aquatic plant exposure, OPP uses the models GENEEC and PRZM/EXAMS. (A
description of these models can be found on the following Web site:
http://www.epa.gov/oppefedl/models/water/index.htm. The screening model GENEEC2 uses a
chemical's label application information, its soil/water partition data and its degradation kinetics
to estimate maximum exposure values in the same "standard" agricultural field/farm pond
scenario as used in the more refined PRZM/EXAMS simulations. The program is generic in that
it does not consider differences in climate, soils, topography or crop in estimating potential
pesticide exposure. OPP currently uses linked PRZM and EXAMS models for refined estimates
-------
of pesticide concentrations in surface waters for drinking water and aquatic exposure
assessments. Each PR2M modeling scenario represents a unique combination of climatic
conditions, crop specific management practices, soil specific properties, site-specific hydrology,
and pesticide-specific application and dissipation processes. Each simulation is conducted using
36 years of rainfall data to cover year-to-year variability in runoff. Daily edge-of-field loadings
of pesticides dissolved in runoff waters and sorbed to entrained sediment, as predicted by PRZM,
are discharged into a standard water body (for ecological assessments, the "farm pond")
simulated by the EXAMS model. The farm pond scenario is based on a 10-ha field draining into
a 1 - ha by 2- meter deep water body.
For terrestrial plants, OPP assumes runoff and drift are from one acre to an adjacent acre. For
terrestrial .plant exposure from spray drift, OPP uses 5% of the maximum label application rate
(this is reduced to 1% when ground equipment is used for application). For terrestrial plant
exposure from runoff, values of 1, 2 or 5% of the maximum label application rate are selected,
depending upon water solubility. When considering semi-aquatic plants, such as those growing
in wetlands, the same calculations for spray drift and runoff as described for terrestrial plants are
used; however, a 10-acre to 1-acre scenario is used.
Currently OPP uses the TerrPlant model, which is an Excel-based program for determining the
EEC's (estimated environmental concentrations) and RQ's (risk quotients) for terrestrial and
semi-aquatic plants. The exposure concentrations from the runoff and/or drift are divided by
plant toxicity endpoints (EC25, EC05 or NOAEC) to produce RQs for terrestrial plants.
Terrestrial acute exposure concentrations from runoff and drift are estimated for plants existing
in dry as well as in semi-aquatic (wetland) areas. This model does not utilize the fate
characteristics of the pesticide in deriving the concentrations. In general, plant EECs estimations
depend on the application rate, application method (ground or aerial), solubility of the active
ingredient and drift conditions. In addition to determining the EECs, the TerrPlant Model can
also provide estimations of acute plant RQs.
Effects upon aquatic and terrestrial plants are estimated based upon the results of the toxicity
tests described in the Subdivision J guidelines. Tier I tests are conducted using a single
"maximum exposure" concentration. If effects at the Tier I concentration are > 25% for
terrestrial plants or >50% for aquatic plants, Tier II testing is required. Tier II testing is dose-
response testing that provides EC25, EC50 and NOEC estimates. Tier III testing is field testing,
to be triggered on a case-by-case basis. To date, only one Tier III aquatic plant test and one Tier
III terrestrial plant test have been required by EPA. The Subdivision J guidelines do not mention
microcosm or mesocosm testing, although such tests are addressed (for aquatic organisms) in
Subdivision E (Hazard Evaluation: Wildlife and Aquatic Organisms).
The Subdivision J guidelines describe testing procedures for five species of aquatic plants: a
green alga, a blue-green cyanobacterium, a freshwater diatom, a marine diatom, and a floating
vascular plant. Three types of tests for terrestrial plants are covered in Subdivision J: seed
germination, seedling emergence, and vegetative vigor. The seed germination test is no longer
12
-------
required as a separate test since this process is covered in the seedling emergence test. Test
species include six species of dicots from at least four families, including soybean and a root
crop, and four species of monocots from at least two families, including corn. Typically, these
requirements are filled by testing exclusively with crop species. Efficacy screening data are not
used because of the following problems' lack of standardized test methods, species,
measurement endpoints. data analysis and reporting; variability and subjectivity in visual
assessments; and lack of availability of the final formulation for use in early efficacy tests.
In 1994, discussions between industry and EPA led to the publication of the Pesticide
Reregistration Rejection Rate Analysis (U.S. EPA, 1994) which analyzed reasons for rejection of
submitted studies. A number of issues related to non-target plant testing were covered in this
document. In 1996, EPA/OPPTS published a set of ecological effects test guidelines (designated
Series 850) that were designed to harmonize test procedures between the pesticides program and
the toxic substances program within the Agency. The Series 850 Guidelines only address
procedures for performing effects assessments, though, and do not discuss how these data are
used in the risk assessment process.
The Series 850 guidelines are currently being updated and further harmonized in response to
comments from a Scientific Advisory Panel, the public, the Reregistration Rejection Rate
Analysis, and recent developments in non-target plant testing methods. Areas of modification
include reduction in the length of the tests for aquatic plant species, i.e., from 5 to 4 days for
algae, from 14 to 7 days for duckweed; addition of specific minimum germination percentages
for terrestrial species; specification of the use of the technical grade active ingredient (TGAI) for
aquatic plant tests and the use of the typical end use product (TEP) for terrestrial plant tests; and
flexibility regarding the substitution of weeds or native plants and/or adding species to the test
battery of crop plants.
3.1.2 Industrial chemicals
Under TSCA, the Office of Pollution Prevention and Toxics (OPPT) evaluates the hazard posed
by the production/manufacture, use and disposal of industrial chemicals to aquatic and terrestrial
organisms found in the environment. OPPT uses a four-tiered testing scheme which begins with
a deterministic screening level aquatic or terrestrial risk assessment. Depending upon the quality
and quantity of the data, uncertainty factors ranging from 1 to 1,000 are used. The lowest ECx is
divided by the uncertainty factor to obtain the level of concern (LOC), and the risk quotient is
then determined by the ratio of the EEC to the LOC. Since the majority of exposure scenarios
considered by OPPT are aquatic, the typical plant toxicity data available are limited to the results
of an algal toxicity study. Terrestrial plant testing is part of the scheme, however, and the
authority exists to request additional studies on both aquatic and terrestrial plants if necessary
(Smrchek et al., 1993). Other recent activities in which EPA has been involved are discussed in
Section 3.3.
13
-------
3.2 History of non-target plant risk assessment in Canada
The Canadian government initiated the development of plant toxicity testing guidelines in 1990.
and Environment Canada published "Proposed Guidelines for Registration of Chemical
Pesticides: Non-target Plant Testing and Evaluation'" in 1993 (Boutin et al., 1993). This proposal
described a four-tiered approach in which tests increase in complexity with each succeeding tier.
Tier I involves screening at a single dose, and includes algal growth and vegetative vigor for
rooted aquatic and terrestrial vascular plants. The single dose or '"Maximum Challenge
Concentration" is based upon the same assumptions described earlier for risk assessments in the
U.S.. except that Canada calculates the EEC from oversprav as 100% of the maximum
application rate rather than 60% (Boutin et al . 1995). Efficacy data developed by the registrant
may be considered in Tier I.
Tier II consists of definitive (multi-concentration) tests, including algal growth, duckweed
growth using a foliar oversprav exposure, seed germination/root elongation for terrestrial
vascular plants, and vegetative vigor for rooted aquatic and terrestrial plants. Tier III includes
additional testing with aquatic plants (submerged and emergent species) and special testing,
while Tier IV includes microcosm/mesocosm and field testing. The algal species include 3
freshwater and 3 marine species; these are the same as the U.S. EPA Subdivision J species with
the addition of a representative marine dinoflageliate and a representative marine green or
golden-brown alga. (Marine species testing may be waived if the product is unlikely to occur in
the marine environment). Testing with vascular plants includes 30 species from 10 families (for
herbicides) and 10 species from 6 families (for non-herbicides).
Testing is conducted with the TAI or the formulated product, depending upon the particular test.
In general, earlier tiers of testing are conducted with the TAI and later tiers with the formulation.
Any statistically significant phytotoxicity at Tier I triggers Tier II testing for that species.
Progression from Tier II to Tier III is triggered when the EEC is greater than the EC50 (for algae
or duckweed) or the EC25 (for terrestrial plant seed germination/root elongation) divided by a
safety factor of 10. Progression to Tier III based upon vegetative vigor occurs when the EEC is
greater than the EC25 for 25% of the plant species or 50% of the plant families.
Although the Canadian proposal has been published, it was not fully adopted by the PMRA
during 1993-1998. As of 1998, the PMRA harmonized with EPA's guideline for non-target plant
testing until such time that a revised guideline is developed between the U.S. and Canada.
14
-------
3.3 Recent efforts in North America
3.3.1 Ecological Incident Information System (EI1S)
In the United States, ecological incident reports play an essential role in the risk assessment and
decision-making process. Widespread ecological incidents for a pesticide may confirm a risk
that was predicted by risk assessment models, or it may indicate that the actual risk is greater
than that predicted by the model. Recognizing the importance of this data, EPA created a
relational database to efficiently store and manage information contained in ecological incident
reports. This database is called the Ecological Incident Information System, or EIIS.
Information obtained from EIIS must be interpreted with great care because of limitations and
potential bias of the data. Data in EIIS are extracted from written reports of ecological incidents.
The two primary sources of incident reports are pesticide registrants and voluntary submissions
by government agencies. Currently, few standards exist for investigation, diagnosis and
reporting of incidents, and only a small portion of all actual incidents is reported to authorities.
Out of those, only a fraction is adequately investigated to identify the cause of the incident, and
only a fraction of those are submitted to the Agency. Despite these limitations, the EIIS database
can be a useful tool for identifying patterns and trends in risks of pesticides to non-target plants.
During the last decade, the Agency has received numerous incident reports concerning adverse
effects to non-target plants from the use of pesticides. This increase is due, in large part, to
improvements in analytical methodology. Scientific literature has also shown that widespread
use of pesticides are associated with damage to non-target terrestrial plants (Pimentel and Levitan
1986). Adverse effects resulting from exposure to pesticides include plant death, reduction in
biomass without recovery, loss of biomass with recovery, and stimulation or excessive plant
growth.
3.3.2 1999 ILSI Workshop
In 1999, the U.S. EPA sponsored an international workshop, organized by the International Life
Sciences Institute (ILSI), entitled "Impacts of Low-dose, High Toxicity Herbicides on
Unintended Plant Species." The report of this workshop will be published by SETAC (Ferenc, in
press). During this workshop, issues were raised about the adequacy of currently used test
species to represent non-tested plant families and species, the predictive value of adverse effects
on early plant growth to adverse effects on reproduction and yield, extrapolation from the
greenhouse to the field, and the relationship of single exposures to multiple exposures. (See
Appendix 4.3 for more details)
15
-------
3.3.3 OPP Efficacy Data
Although the U.S. EPA waived the submission of efficacy data on crops, this data can be
required when needed, e.g.. Public Interest Findings. Special Reviews or in some cases
Emergency Exemptions. The Agency is currently developing criteria for submission of efficacy
data to aid in the assessment of risk to non-target plants from the use of pesticides.
3.3.4 Identified OPP Research Needs
Below is a list of OPP priority research needs for non-target plant risk assessments. This list
may change depending upon recommendations of the Scientific Advisory Panel (SAP. 2001
meeting £>n ''Review of Non-target Plant Toxicity Testing Under NAFTA"). The SAP report to
the Agency is expected to be received in the first part of November.
© Develop test battery sensitivity ratios for existing test species vs proposed species.
• Conduct research to determine:
- If annual crop plants serve as surrogates for native plants, and/or for
perennial/woody plants.
- If terrestrial vascular plants can serve as surrogates for emerged vascular rooted
macrophytes.
- If one taxonomic group can serve as surrogate to another taxonomic group of
plants.
- The most sensitive growth stages/endpoints - do we need reproductive
endpoints?
- If C02, 02, other gaseous endpoints are more sensitive or more predictive than
current biomass and plant height measurements of effects.
- Laboratory to field ratio.
- The most appropriate species for field monitoring purposes
© Develop test methods for woody plants tests, life cycle and partial life cycle tests,
soil toxicity test, and field tests.
• Determine species of high importance/dominance in each eco-region that could
serve as test species representative of the eco-region.
® Determine linkages between plant loss and effects on wildlife.
• Assess causes of long-range pesticide transport, and the impacts to plants from
exposure.
• Assess feasibility of using GIS mapping.
® Assess the feasibility/applicability of new technologies: satellite infrared imaging,
mRNA.
• Collection efficiency of plants: How does a plants geometry and surface texture
affect it's ability to capture spray drift?
® Phytotoxic effects from concentrated and dilute sprays: Are standard phytotoxicity
tests using dilute sprays covering large areas analogous to concentrated droplets
impacting discrete spots on plant surfaces?
16
-------
® Exposure to runoff of phvtotoxic chemicals: Are current phytotoxicity tests
adequate to measure the effects from pesticide runoff?
3.4 Approaches from other countries
3.4.1 European Union
For the European Union, requirements for risk assessment of pesticides are described in EC
Directive 91/414/EEC. with Annexes II and III setting out the data requirements and Annex VI
outlining the principles of risk assessment. The approaches used to characterize risks from
pesticides are comparable to those used in the U.S. (Campbell et al., 1999). A tiered system is
used in \yhich the first tier provides a conservative evaluation. Procedures for "higher tier"
evaluation are not well developed, which prompted a group of experts from 14 EU member
states, the U.S., and Canada to hold a workshop under the auspices of the EC, SETAC-Europe
and OECD (Campbell et al., 1999). This workshop, termed "HARAP" (Higher-Tier Aquatic
Risk Assessment of Pesticides), focused on aquatic risk assessment only.
Campbell et al. (1999) summarized the procedures for aquatic risk assessment for pesticides in
the EU. Initial peak concentrations are calculated from look-up tables of spray drift
(Ganzelmeier et al., 1993), assuming that drift enters a 30 cm deep, static, water body and
assuming instantaneous mixing in the water column. Long-term exposure concentrations are
calculated with appropriate fate models. A working group known as FOCUS has been
developing a number of standard European scenarios to estimate predicted environmental
concentrations (PECs) in surface water. Toxicity values are then divided by the resulting
exposure concentration to determine a'toxicity: exposure ratio (TER). If the TER for acute
toxicity and exposure is <100 or for chronic toxicity and exposure is <10, then further evaluation
is required.
Recommendations for improving higher-tier aquatic risk assessments were developed during the
HARAP workshop. One of the suggestions of this workshop was to make better use of
information on the fate and effects properties of the chemical (termed "interrogating the core
data"). For example, comparing the initial (instantaneous) PEC for a compound that dissipates
rapidly to toxicity data generated under constant exposure conditions for a number of days,
weeks or months, could overestimate the potential for effects (especially for chronic studies).
Using time to effects and PECs based upon time-weighted average concentrations could be more
appropriate in these scenarios. Laboratory studies conducted by using modified exposure
regimes (e.g., incorporating other fate processes, using variable dosing, etc.) could also be used.
To reduce uncertainty, additional organisms could be used in single-species tests. The workshop
participants recommended that "in general, for compounds which do not appear to be selective to
aquatic organisms (i.e., all standard test organisms respond at a similar - within an order of
magnitude - concentrations), it is suggested that eight species be used as a minimum to describe
the distribution of sensitivities of aquatic organisms." (Campbell et al., 1999). If a particular
17
-------
group of organisms is identified as sensitive, then it would be appropriate to test eight species
from that group (except in the case of fish, where it is recognized that sensitivity distributions are
relatively narrow). The additional species data could be used to reduce the uncertainty factor, or
in probabilistic effects assessment. Population-level studies are another approach; these can be
experimental or modeling studies. Indoor multi-species tests and field studies were also
discussed for use at higher tiers.
There is little guidance in EC Directive 91/414/EEC specifically regarding non-target plants,
other than evaluation of the effects upon algae. According to a survey of OECD member nations
conducted m 1994 (OECD, 1994), few countries routinely required data on the ecotoxicology of
pesticides to non-target plants at that time. The most common requirement was for a growth
inhibition.test on algae when the pesticide was used outdoors. This is listed as always required in
Denmark, Germany, Italy. Austria, Finland. Norway, and Sweden, and sometimes required in the
Netherlands. Portugal, United Kingdom, Australia, Canada, Switzerland and the U.S. The only
country surveyed that did not have any requirements for an algal toxicity test was Japan. Most
countries require the algal test on the active ingredient, although sometimes testing on the
formulation is requested. Countries that also ask for algal testing for pesticides with indoor uses
include Germany, Italy, the Netherlands, Australia, Austria, Sweden, and Switzerland.
Terrestrial plant tests (seed germination and vegetative vigor) were required less frequently.
Only Switzerland always requires both these tests, for both indoor and outdoor uses, and they are
conducted with the formulated product. The member countries of the European Community did
not report requiring terrestrial plant tests at the time of the survey. Tests on aquatic plant growth
were reported as being not required or required only infrequently, except in the U.K. This
survey report did not contain information on how the data were used in risk assessment.
3.4.2 European and Mediterranean Plant Protection Organization (EPPO)
The EPPO approach ("Decision-Making Scheme for the Environmental Risk Assessment of
Plant Protection Products'") is a tiered scheme that incorporates the elements of ecological risk
assessment to a large degree. Chapter 12 of this scheme discusses risk assessment for "non-target
higher plants." The first step is not called "problem formulation" but involves consideration of
the properties of the pesticide and its use pattern. If there is no potential for exposure, the
conclusion of "negligible risk" is reached. If exposure is possible, pathways of drift and gaseous
transport are considered. Predicted environmental concentrations (PEC) are compared to toxicity
endpoints to develop Exposure-to-toxicity Ratios (ETRs). The PEC values are determined by
considering the fraction of the applied pesticide that could be expected at a certain distance from
the treated area. The denominator of the equation is the EC50. As non lethal endpoints (e.g.
biomass, plant height, number of leafs, percent chlorosis, percent ground cover) are determined
for the calculation of the EC50 value, this value is proposed as the basis for further assessment
steps and not a NOEC value. If the ETR due to drift events at 1 meter distance from the treated
area is less than 1. the application is classified as low risk. If the ETR due to drift events at 5
18
-------
meter distance from the treated area is greater than 1. the application is classified as high risk. In
the case of orchards, the distances from the treated area are 3 and 7. respectively.
For gaseous transport and run-off events a 'safety factor' of 10 is used. No scientific reason is
available for explaining this safety factor of 10. but it is chosen because it offers roughly the
same protection level as provided in the drift scenario (differences in percentage of drift at 1 and
5 meter( or 3 and 7m)).
In the EPPO scheme, all plant protection chemicals (not just herbicides) are tested.
A differentiation is made between herbicides and growth regulators and other plant protection
products. For this latter group, screening data (including efficacy data) can be used to show
whether the compound has phvtotoxic properties or not. Data for at least 6 species from 6
different^families (including monocots and dicots) should be made available. The plants should
be tested at the maximum recommended application rate or higher. If phytotoxicity is observed
(50% effect in one or more plants). Tier 2 is triggered and thereafter the same route as herbicides
and growth regulators is followed.
A minimum number of 6 species, representing families for which significant herbicidal activity
has been claimed/found should be tested (with a range of concentrations, preferably in a
geometric progression, but covering the EC50 value, e.g., OECD 208). Statistical methods will be
applied to calculate the 5th percentile of the available toxicity data. These methods apply
uncertainty factors that are dependent on the sample size (the safety factors decrease with
increasing sample size).
For post emergence herbicides and other pesticides, the vegetative vigour test should be used
unless the mode of action indicates a specific test otherwise, e.g. growth regulators influencing
flowering, or some inhibitors of cell division causing flower sterility. For pre-emergence
herbicides, the seedling emergence test is more applicable. The definitive tests should be
conducted using an appropriate lead formulation.
The EPPO scheme also incorporates uncertainty analysis and risk management components. The
uncertainty analysis involves re-examining the assumptions used in the original risk
characterization and refining them to reflect potential variation. The risk management phase
discusses various mitigation alternatives. A simplified diagram of the EPPO scheme for the
evaluation of the risk of plant protection products to terrestrial higher plants is presented in
Figure 1.
The EPPO approach is compared to the existing approach used in the U.S. and the proposed
approach developed in Canada in 1993 in Tables 2 and 3.
Please note that the scheme described in this chapter is still a draft and probably will become
operative in March 2002. But it is still possible that minor adjustments will be made. The
predicted environmental concentrations that are used in the non-target terrestrial higher plant
19
-------
scheme will be provided by two other EPPO schemes, e.g. air and soil. These schemes are not
described in this document.
3.4.3 Ministry of Agriculture, Fisheries and Food (MAFF)
In the U.K . the Pesticides Safety Directorate of the Ministry of Agriculture, Fisheries and Food
(now called the Department for Environment, Food and Rural Affairs) commissioned a study
entitled ''Assessing Pesticide Risks to Non-target Terrestrial Plants" (Breeze et al., 1999).
Section Six of this study ("Options for testing and risk assessment," by A.D.M. Hart), provides a
summary of different approaches to non-target plant risk assessments. These approaches (by the
U.S., Canada, EPPO, OECD and the Global Crop Protection Federation, an industry association)
were used as the basis for a round-table discussion. The concepts discussed and conclusions
reached are summarized below
In answering the question "what are we trying to protect?" the group recognized that it would be
impractical to test all species requiring protection, but that the most important common species
should be included. Consideration should be given to including groups, such as ferns, mosses,
liverworts, and conifers, that have typically not been included in the past. Risk assessment should
consider desirable species inside the target area as well as outside; however, the level of risk that
is acceptable will generally be higher in the cropped area. Although the goal is to protect plant
populations or communities, it is more practical at present to focus on the individual. Research is
needed to determine if effects on mature plants are important and if so, whether they can be
predicted from tests of germination and early growth.
It may be appropriate to separate "in crop" versus "out of crop" exposure scenarios. For the "in
crop" scenario, the exposure can be simulated by a test in which the pesticide is applied as
recommended. For the "out of crop" scenario, spray drift accounts for the majority of exposure
and can be estimated; the Ganzelmeier estimate was recommended. Although vapor drift may
contribute, it is too complex to include in routine assessment and would only be included based
upon the properties of the pesticide. Runoff exposure may be estimated as well. These routes
must be aggregated before comparing the total exposure with toxicity; however, this is not a
straightforward process. It seems sensible to consider the exposure to two life stages (seeds and
young plants) separately. Consideration should also be given to mature plants at flowering and
fruit formation. Combining the factors of location (in/out of crop) and life stage (seed/young
plant) leads to four separate exposure scenarios in each risk assessment.
In the toxicity assessment, it seems efficient to make use of efficacy screening data to the extent
possible. There is a need to define the minimum type and quality of data required at each stage of
the risk assessment. It also seems efficient to use maximum challenge concentration tests at
lower tiers of assessment and dose-response tests at higher tiers. In higher tier assessments, the
formulated product should be used. At lower tiers it may be efficient to use the active ingredient
but it may be necessary to incorporate an uncertainty factor to account for additional toxicity of
20
-------
the formulated product. Should tests with mature plants be considered important, protocols must
be developed that are repeatable. relevant to field use conditions and exposure patterns, relevant
to the modes of action of the active ingredient, and realistic or at least interpretable from an
ecological viewpoint. Effects to consider are growth of the adult plant, seed production,
flowering, fruit production, and vegetative propagation.
Because it is impossible to determine toxicity to all plant species, it is necessary to select a
reasonable number and extrapolate to other species. To date, the approach has been to pick a
number of species and in some schemes (e g., Canada) use a safety factor for extrapolation.
However, alternative statistical approaches are more objective, although as yet unproven for
plants. It may be possible to use the HD5 approach that has been developed for birds and
mammals. This approach characterizes the shape and breadth of the sensitivity distribution and
uses it to predict the dose that would be hazardous to the most sensitive 5% of species. If the
distribution-based approach is successful, it will provide a clear basis for deciding which and how
many species should be tested. In more refined assessments, tests might be conducted with
particular species of ecological importance. It was also noted that available data indicate that
grasses show a more consistent response than broadleaved plants, suggesting that more of the
latter should be tested to characterize the range of variation for a particular chemical.
The conclusion about the format of the risk assessment was that the structure should be flexible,
beginning with existing screening data, then conducting additional dose-response tests as needed,
then special studies if required.
Consideration should be given to using probabilistic methods, and case studies could be
developed by using existing information to compare these approaches with more conventional,
deterministic ones. In a probabilistic assessment, at least one of the input variables (toxicity or
exposure) is a distribution rather than a fixed point estimate, and the output is generally also
expressed as a distribution. Some ways to express the resulting risk include graphical overlays of
the cumulative distributions of toxicity and exposure; calculating a distribution of toxicity-
exposure ratios; or a repetitive simulation of exposure of a large number of individuals, taking
values at random from the input distributions and assigning a response based upon the extent to
which the exposure exceeds the sensitivity of the individual. There are two major advantages of
probabilistic assessments: they enable uncertainties to be dealt with explicitly, objectively and
quantitatively; and they can be used to produce results that describe the frequency and magnitude
of effects.
The degree of refinement required in the risk assessment depends on how close the actual risk is
to the threshold of acceptability. One can refine the exposure estimates, the toxicity estimates, or
both, depending upon how much each option will contribute towards reducing uncertainty and
how much each option will cost. A useful scheme is one in which the pesticide use is categorized
as low, medium or high risk, such that no further refinement is needed for either the low or high
risk categories to allow regulatory decisions to be made. The medium risk category typically
benefits from additional refinement to provide sufficient information for decision-making.
21
-------
A five-tiered process was suggested, as follows:
0 - potential for exposure is evaluated
1 - preliminary screening using efficacy studies
II - initial assessment using worst-case TER
III - initial assessment using most-likely TER.
IV - refined assessment where required, including special studies and attention to key
species
Tiers 0 - If serve to identify low-risk pesticides which require no further assessment, Tier Iff
serves to identify high-risk pesticides which are likely to require risk mitigation, and Tier IV
refines the assessment for pesticides of intermediate risk potential. The toxicity test to be
conducted assumes germination/emergence tests with young plants. Other types of effects
(including those on mature plants) could be addressed through appropriate tests or the use of
uncertainty factors. The exposure is equal to the maximum application rate or a fraction thereof
(for out-of-crop scenarios) and does not address routes such as vapor drift and run-off.
Uncertainty factors could be developed here as well. This is especially important for the worst-
case TER (Level II), since pesticides which pass the trigger at this level would be presumed safe.
The approach presented would be used in a flexible way to meet the needs of each assessment.
For example, the screening data in Tier I together with other information (e.g., mode of action)
could be used to focus testing at higher tiers on particular-species and growth stages.
-------
4.0 Appendix
4.1 List of Acronyms
ACPA - American Crop Protection Association
ASTM - American Society for Testing and Materials
CFR - Code of Federal Regulations
EEC - Expected Environmental Concentration
EC - European Communities
EC25 - the concentration that results in a 25% reduction in the test endpoint being measured
relative to the control
EC50 - the concentration that results in a 50% reduction in the test endpoint being measured
relative to the control
ECx - the concentration that results in a reduction of "x': percent in the test endpoint being
measured relative to the control.
ECOFRAM - Ecological Committee on FIFRA Risk Assessment Methods
EPA - Environmental Protection Agency
EPPO - European and Mediterranean Plant Protection Organization
ERA - ecological risk assessment
EU - European Union
EXAMS - Exposure Analysis Modeling System (a model)
FIFRA - Federal Insecticide, Fungicide and Rodenticide Act
GENEEC - Generic Estimated Environmental Concentration (a model)
GIS - Geographical Information Systems
HARAP - Higher-tier Aquatic Risk Assessment for Pesticides
HD5 - Hazardous dose for 5% of total species
ILSI - International Life Sciences Institute
LOC - Level of Concern
MAFF - Ministry of Agriculture, Fisheries and Food
MOA - mode of action
MOU - mode of uptake
NAFTA - North American Free Trade Agreement
NOEC - no observed effect concentration
NTP - non-target plants
OECD - Organization for Economic Co-operation and Development
OPPTS - Office of Prevention, Pesticides and Toxic Substances
OPP - Office of Pesticide Programs
OPPT - Office of Pollution Prevention and Toxics
PEC - Predicted Environmental Concentration
PMRA - Pest Management Regulatory Agency
PRZM - Pesticide Root Zone Model (a model)
QSARs - Quantitative Structure Activity Relationships
RQ - Risk Quotient
23
-------
SAP - Scientific Advisory Panel
SET AC - Society of Environmental Toxicology and Chemistry
T&E - Threatened & Endangered
TEP - typical end use product
TER - Toxicity Exposure Ratio
TAI - technical active ingredient
TGAl - technical grade active ingredient
TSCA - Toxic Substances Control Act
24
-------
4.2 References
Benenati. F. 1990. Keynote address: Plants - keynotes to risk assessment. In W. Wang, J.W.
Gorsuch and W.R. Lower, eds.. Plants for Toxicity Assessment. STP 1091. American Society
for Testing and Materials. Philadelphia. PA, pp. 5-13.
Boutin, C., K.E. Freemark, and C.J. Keddv. 1993. Proposed Guidelines for Registration of
Chemical Pesticides: Non-target Plant Testing and Evaluation. Technical report series no.
145. Canadian Wildlife Service (Headquarters), Environment Canada, Ottawa.
Boutin. C , K E. Freemark, and C.J. Keddy, 1995. Overview and rationale for developing
regulatory guidelines for nontarget plant testing with chemical pesticides, Environmental
Toxicology and Chemistry, Vol. 14, No. 9, pp. 1465-1475.
Boutin, C. and C. A. Rogers, 2000. Pattern of sensitivity of plant species to various herbicides -
an analysis with two databases. Ecotoxicology 9: 225-271.
Breeze, V.G., E.J.P. Marshall, A. Hart, J.A. Vickery, J. Crocker, K. Walters, J. Packer, D.
Kendall, J. Fowbert and D. Hodkinson, 1999. Assessing pesticide risks to non-target
terrestrial plants: A desk study. Commissioned by MAFF Pesticides Safety Directorate.
Campbell, P.J., D.J.S. Arnold, T.C.M. Brock, N.J. Grandy, W. Heger, F. Heimback, S.J. Maund,
and M. Streloke, eds., 1999. Guidance Document on Higher-Tier Aquatic Risk Assessment
for Pesticides (HARAP), a special publication of SETAC.
j
Davy, M., R. Petrie, J. Smrchek, T. Kuchnicki and D. Francois. 2001. Proposal to update non-
target plant toxicity testing under NAFTA.
ECOFRAM (Ecological Committee on FIFRA Risk Assessment Methods), 1999a. ECOFRAM
Aquatic Draft Report.
ECOFRAM (Ecological Committee on FIFRA Risk Assessment Methods), 1999b. ECOFRAM
Terrestrial Draft Report.
European and Mediterranean Plant Protection Organization (EPPO), updated. Decision-Making
Scheme for the Environmental Risk Assessment of Plant Protection Products.
Ferenc, S., ed. (In press). Impacts of Low-Concentration, High-Potency Herbicides on Nontarget
Crops and Unintended Plants
Fletcher, J.S., 1991. A brief overview of plant toxicity testing. In W. Wang, J.W. Gorsuch and
W.R. Lower, eds., Plants for Toxicity Assessment. Vol. 2. STP 1115. American Society for
Testing and Materials, Philadelphia, PA, pp. 5-11.
25
-------
Ganzelmeier, H., D. Rautmann, R. Spangenberg, M. Streloke, M. Herrman, H.J Wenzelburger
and H.F. Walter. 1995. Studies on the spray drift of plant protection products. Report.
Mitteilungen aus der Biologischen Bundesanstalt fur Land-und Fortwirtschaft, Berlin-
Dahlem. Germany.
Hart, A.D.M., 1999. Section Six Options for testing and risk assessment, in Breeze et al..
Assessing pesticide risks to non-target terrestrial plants: A desk study. Commissioned by
MAFF Pesticides Safety Directorate.
Hoist. R.W. and T.C. Ellwanger, 1982. Pesticide Assessment Guidelines: Subdivision J Hazard
Evaluation: Nontarget Plants, EPA 540/9-82-020.
Kapustka. L.A., 1997. Selection of phytotoxicity tests for use in ecological risk assessment. In
W Wang. J.W. Gorsuch and J.S. Hughes, eds., Plants for Environmental Studies. CRC
Press/Lewis Publishers. Boca Raton, FL, pp. 515 - 548.
Lewis, M.A., 1990. Are laboratory-derived toxicity data for freshwater algae worth the effort?
Environ. Toxicol. Chem. Vol 9, pp. 1279-1284.
McKelvey, R., Honegger, J., and Wright, J. A comparison of crop and non-crop plant sensitivity
to eleven herbicides in laboratory testing (in review, 2001).
Organization for Economic Co-operation and Development (OECD), 1994. Data requirements
for pesticide registration in OECD member countries: survey results. Environment
Monograph No. 77, Environment Directorate, OECD, Paris, France.
Pimentel, D. and L. Levitan. 1986. Pesticides. Amounts applied and amounts reaching pests.
Bioscience 36(2): 86-91.
Power, M. and L.S. McCarty, 1998. A comparative analysis of environmental risk assessment/risk
management frameworks. Environmental Science & Technology Vol. 32, No.9, pp. 224A -
231 A.
Science Advisory Panel, 2001. Proposal to update non-target plant toxicity testing under NAFTA
Shaw, J.L. and J. A. Gagne, 2000. Terrestrial non-target plant risk assessment for pesticides: A
generic problem formulation and proposed protection goals. Prepared for the American Crop
Protection Association Environmental Risk Analysis Oversight Work Group.
Smrchek, J., R. Clements, R. Morcock, and W. Rabert, 1993. Assessing ecological hazard under
TSCA: methods and evaluation of data. In: W.G. Landis, J.S. Hughes, and M.A. Lewis, eds.,
Environmental Toxicology and Risk Assessment. ASTM STP 1179, American Society for
Testing and Materials, Philadelphia, PA, pp. 22-39.
26
-------
Society of Environmental Toxicology and Chemistry (SETAC), 1997. Ecological Risk
Assessment Technical Issue Paper. Pensacola. FL.
Taylor, G.E . 1999. Ecological risk characterization of low dose, high toxicity herbicides.
Unpublished SETAC manuscript from 1LS1 Workshop.
Urban. D.J. and N. J. Cook, 1986. Hazard Evaluation Division Standard Evaluation Procedure:
Ecological Risk Assessment. EPA 540/9-86-167
U S. Environmental Protection Agency, 1992. A framework for ecological risk assessment.
Office of Research and Development. Washington, DC. EPA/630/R-92-001.
U S. Environmental Protection Agency, 1994. Pesticide Reregistration Rejection Rate Analysis:
Ecological Effects. Office of Prevention, Pesticides and Toxic Substances. EPA 738-R-94-
035.
U.S. Environmental Protection Agency, 1998. Guidelines for ecological risk assessment. Office
of Research and Development, Washington, DC. EPA/630/R-95/002F.
U.S. Environmental Protection Agency, 2000a. A progress report for advancing ecological risk
assessment methods in OPP: A consultation with the FIFRA Scientific Advisory Panel,
Overview Document. http://www.epa.gov/scipolv/sap/2000/April/probover.pdf
U.S. Environmental Protection Agency, 2000b. Technical progress report of the implementation
plan for probabilistic ecological assessments: Aquatic systems.
http://www.epa.gov/scipolv/sap/2000/April/probaq.pdf
U.S. Environmental Protection Agency, 2000c. Technical progress report: Implementation plan
for probabilistic ecological assessments: Terrestrial systems.
http://www.epa.gov/scipolv/sap/2000/April/probter.pdf
Wang, W. and K. Freemark, 1995. The use of plants for environmental monitoring and
assessment. Ecotoxicol. Environ. Saf. 30:289-301.
27
-------
4.3 ILSI Workshop Summary and Recommendations
In 1999. the OPPTS sponsored an international workshop titled "Impacts Of Low-dose. High
Toxicity Herbicides on Unintended Plant Species" This workshop was attended by 40 scientists
from acadeima. government, industry,, and ecological groups. Five research papers were
presented lor discussion at the workshop This workshop provided a forum for discussion and
analysis of public concerns regarding low dosage, high toxicity herbicides. Both the manuscripts
and workshop were guided by the following questions'
What are the current methods used to quantify the exposure to unintended/nontarget plants
from long-range transport of low-dose/high potency herbicides and how can they be
improved?
What needs to be done to evaluate the biological effects of chronic low-dose exposures and
acute low-dose exposures?
Are the current species and testing endpoints in laboratory and greenhouse studies sufficient
for evaluating the effects of low-dose, high potency herbicides on unintended or nontarget
plants?
Is the current technology sufficient to adequately evaluate unintended or nontarget plant
impacts at the individual, community and ecosystem levels?
What data and methods are needed to better characterize the likelihood, magnitude, and
severity of adverse ecological effects of these herbicides on unintended or nontarget plants
How can we better identify and characterize risks to native plant population structure at the
ecosystem level?
The workshop papers are to be published later in 2001 by The Society for Environmental Toxicity
and Chemistry (SETAC).
The five papers presented at the ILSI workshop were:
1. "Low Dose, High Toxicity Herbicides: An Historical Perspective of Environmental Concerns
to Frame the Issues", Anne Fairbrother and Lawrence A. Kapustka.
2. "Nontarget Aquatic Plant Effects Of Acetolactate Synthase Inhibiting Herbicides", Hans G.
Peterson.
3. "Ecological Risk Characterization of Low Dosage high Toxicity herbicides", George E.
Taylor, Jr.
4. "Exposure To Low-Dose, High Toxicity Herbicides - Literature Review", Donald Waite.
5. "Non-target Terrestrial Plant Effects of Low Dose, High Toxicity Herbicides", Bruce
Maxwell and Rebecca Weed.
In the workshop a focus was placed on the ALS inhibiting herbicides, including imidaolines,
sulfonylureas, triazolopyrimidine sulfonamides and pyrimidyl thio-benzates, as much less data
exist for other low-dose, high potency herbicides. However, the principles discussed might apply
28
-------
equally to non-ALSase inhibiting compounds. The questions posed to both the authors and the
workshop participants were intended to frame the discussions on the current state of scientific
knowledge and provide insight and direction to future approaches to risk assessment of these
compounds.
Alleged incidents of low-dose, high potency herbicide drift from application sites to distant
nontarget sites are of concern. The veracity of these reports has been challenged; however few
data are available to either support or refute such reports. In particular, a series of alleged
incidents were reported in the state of Washington. Workshop participants were in agreement
that there is no current or soon to be expected testing methodology that can prevent unforeseen
incidents, such as have been reported on cherry trees in the Northwest. An uncommon
combination of natural and anthropogenic events may have occurred to result in the reported
impacts. Modelling and testing methodologies do not currently exist that will allow us to predict
when such a confluence of factors might be expected to occur. Indeed, test methods currently
used for the detection of compounds in plant tissues, soil and sediment cannot measure residues
at the very low levels encountered with these herbicides.
Several new methods for detection of exposure were discussed but inherent problems with each
hinder the practical application of these methods for routine testing at this time. High
performance liquid chromatography with detection by mass spectrometry (HPLC/MS) is
considered the most sensitive and specific methodology for detection of the low-dose, high
potency herbicides. Under certain conditions, this methodology also may allow for quantification
of exposure. However, HPLC/MS is expensive and issues of the tests' sensitivity remain.
Although HPLC mass spectrometry costs have begun to decline, maintenance and technical skill
requirement factors are high.
Immunoassay methods exist, but currently are not very specific and cross-reactivity is a
confounding factor. The costs associated with the two test systems, immunoassay and
HPLC/MS, are both high although the cost streams are significantly different. Immunoassay kits
are relatively inexpensive on an individual basis, however very few field samples can be tested
with a kit due to the necessary number of replicates and control dilutions. The high cost of
reagents and controls keeps the test cost high at an application level; running a very large number
of samples is expensive.
Reasonably, analytical methods should be capable of detecting the compounds at concentrations
that are predicted by bioassay results to cause effects. However, ideal analytical methods also
should be able to detect levels that cause persistent or significant effects under natural conditions.
Ultimately, analytical methods that are able to detect very low levels will be needed to consider
the function of distance from the site of application, and degradation processes and rates in plants.
The general understanding from the workshop participants is that the analytical technology is not
yet available for practical widespread application, but these newer methods are being refined and
may provide the necessary sensitivity and specificity required in the foreseeable future, at
reasonable expense.
29
-------
A confounding problem with detection of exposure relates to the quick degradation by plants of
the parent compound. Within hours after exposure, the plant already has significantly reduced the
amount of parent compound so collecting plant tissues for analysis may not be the most
efficacious approach to determining exposure to non-target and unintended plants. Although soil
microbes degrade the compounds as well, concentrations more representative of exposure might
be found by testing soil rather than by testing plant tissues Metabolic by-products of these
herbicides might serve as biomarkers of exposure, but some metabolites are not specific to the
low-dose, high potency herbicides. Specific amino acid chains resulting from exposure to
ALSase inhibitors might for example provide "signature fingerprints" of exposure to these
compounds Analytically, if an amino acid is found to be a signature metabolite, this certainly
provides a more cost-effective diagnostic. Unfortunately, while signature metabolites may
correlate, well with exposure, they most likely will not correlate well with damage. In addition,
metabolite levels in plants (e.g , alpha keta butyric acid) may increase with the environmental
stresses associated with exposure under natural conditions, confounding the uncertainty in
extrapolating from the laboratory to the field. Unless the rate of metabolite formation can be well
characterized, it will not be possible to determine the timing of exposure. Even if signature
amino acid metabolites for the ALSase inhibiting herbicides are found, these will only indicate
that a plant was exposed to an ALS inhibitor. It would not indicate when, how much, or which
one. It is then necessary to correlate the presence of these biomarkers with some measure,
perhaps in soil or other environmental media that will correctly identify the chemical and allow
the final link to causality. In tenns of a forensic diagnostic, such methods may have applicability
but it still is necessary to be able to actually measure the compound to confirm correlation of
effects to exposure. However if signature amino acid chains are tested for and not found to be
present, this is indicative of non-exposure. Chemical or biomarker tagging of the compound
might provide a feasible mechanism for detection of exposure; timing of exposure and extent of
adverse effects would likely not be obtainable through this approach.
A suite of plant species currently is used for laboratory, and in some case field, testing of low-
dose, high potency herbicides in pre-registration risk assessment. However, this suite of species
was not developed specifically for these compounds, and generally is not representative of all
aquatic and terrestrial species potentially exposed under conditions of use. The current screening
level testing is a simplistic approach; running through a battery of tests under conditions that
maximize activity, then finding the lowest concentration for effect on the most sensitive species.
The intent has been to find a point below which there is confidence that no detrimental effects
will occur.
There was lengthy discussion of the representativeness, or lack thereof, of the suite of species
used in the test batteries. For example, although upwards of 50 species from a variety of plant
families might be used in pre-registration efficacy studies, the species used are almost exclusively
crop and weed species. The endpoint of these efficacy tests typically is plant death, and is used to
reflect a predicted level of control (e.g., 80%). Dose-response assessment cannot be conducted on
this type of data. In addition, given that the tests are conducted for efficacy determination, they
may not adhere to standardized GLP protocols. Review of efficacy data however, might aid in
30
-------
identifying plant families likely to be more sensitive and therefore serve as a basis for establishing
appropriate test species
Flisk assessment of ALSase inhibitor compounds has not expanded much beyond a tier I or
screening level and there is a lack of methods to extrapolate beyond these data to other life stages,
species or to the ecosystem level. Development of approaches to higher level testing and
extrapolation for plant species are well behind that of avian and mammalian risk assessment.
Given this, extrapolation methods for plants represent a significant void for risk assessment on
plant species.
It is unclear whether more test species are necessary or warranted to evaluate adverse effects of
the low-dose, high potency herbicides in laboratory or field studies, however, new and innovative
methods for determining the appropriate species to evaluate were discussed. These tests are
protective, but not predictive. Additional information is needed to allow for prediction of effects
under natural conditions and hence protect nontarget plants and ecosystems. There may be only a
few species in a community responsible for the maintenance of the structure and ostensibly the
functioning of the community. If the goal is to conserve the ecosystem within which these
chemicals will be used, the aim should be to protect the ecologically important species, including
the primary drivers of the ecosystem. This information can be used to narrow the focus of
potential species for evaluation of impacts. A broader approach is to consider potential
environmental impacts on a regional basis and take advantage of new methodologies, such as GIS
and drift models, to determine a priori which natural or agricultural ecosystems might be "at risk"
of exposure to a pesticide based on the expected or predicted post-registration use.
For example, it is possible to examine'the spatial distribution of the particular target crop on
electronic maps, overlay those maps with wind speed and weather data, and again, overlay with
the spatial distribution of rare endangered species. Through all these layers can be seen the
area(s) potentially at risk of high exposure and the species of interest in these areas. Incident
reports may be used to identify or pre-locate areas for GIS application when considering a new
compound from a class similar to the incident compound. Data layering exercises like this do not
reduce the uncertainty associated with the individual data used in the study, but this approach is
amenable to use in probabilistic analysis. Probabilistic analysis may allow for movement away
from the necessary identification of the most sensitive species. Species may instead be selected
for functional and/or physiological characteristics allowing for better prediction of environmental
effects.
Systems identified as "at risk" might then be grouped based on similarities of structure and
function. Within groups, plant species might then be identified, such as key species, potentially
more sensitive species, representatives from various trophic levels, etc., for consideration in a risk
assessment. Under current approaches, these species would likely be represented in laboratory
and/or field studies by the small suite of available surrogate species. Initially, this approach
would generate different lists of test species for different compounds but it would take an equally
immense effort to identify the 20 species that represent a one-size fits-all test for every
31
-------
compound; and it is highly unlikely that such a definitive list could be derived. Workshop
participants discussed an approach that could be used to integrate the information from GIS site-
based studies with a generic set of criteria to determine the most appropriate suite of test species
for use.
Once a list of criteria is established, the suite of test species developed has to include at least one
species that satisfies each of the criteria. One criterion might be "present in habitat;" at least one
species present in the habitat(s) predicted to be exposed under field application and use should be
included or represented in the test. At least one species that serves as an important component of
the food web in the system should be tested, as well as at least one species that serves another
important process in the ecosystem, such as nitrogen fixation. A species that exhibits a
reproductive response to at least one class of herbicide at concentrations encountered in the field,
using models that predict concentrations if actual data are lacking, should be included. Another
criteria might be a close taxonomic and/or physiologic relationship to the target weed. But one
criterion is very important: it must be possible to grow the plant under practical experimental
conditions. The idea here is not to identify a list of plants that each meets all criteria but rather to
identify a list of plant species that collectively meets all criteria. This criteria-based exercise is
not meant to exclude any of the current standard tests and might be more appropriately applied in
a later tier process using the first criteria coupled with the regional or "at risk" ecosystem-based
approach. For example, once an "at risk'" system is identified, specific species from these habitats
that meet the criteria may be considered.
A straw set of criteria for aquatic plant test species were developed during the workshop
discussion for illustrative purposes:
Present in habitat/potentially affected/proximity to application site
Important in the system's food web
Important in some ecological process (e.g., nitrogen fixation, habitat/soil fixation).
Little intraspecific variation in response.
Sexually reproducing and exhibits sexual reproduction response to some herbicide(s).
Asexually reproducing and exhibits asexual reproduction response to some herbicide(s).
Roots into substrate (for sediment/soil exposure).
Similar physiology as target weed/plant.
Can be worked with under laboratory or field conditions.
A potential problem with this approach is that 10 or 15 species might meet three-quarters of the
criteria and the most sensitive species might still be missed. For instance, the terrestrial
solanaceous plants nightshade and tomato, would likely meet the same criteria but the herbicide
effects might be seen at much lower levels of exposure in the tomato than in the nightshade.
However, this is a problem in the current testing protocols as well.
It may be that one or several of the current test species satisfy all of the above criteria. If not,
additional species might need to be considered to provide the missing information. A potential
32
-------
problem with this approach is that if a species has not been used extensively, there may not be
Good Laboratory Practice (GLP) protocols for its use. Adding species to testing protocols or
developing new testing protocols is a nontrivial exercise. It can take two or three years to develop
the protocol, ring test to get the protocol standardized, and apply GLPs. It also is possible to
conduct nonstandard tests under GLP conditions. The decision must be made whether the added
information collected by testing these additional species exceeds the loss of "standardization" of
protocol. Over time. GLP protocols might be developed if new species become commonly used.
Workshop participants were in agreement that the scientific community should and could
improve upon the plant physiology database in this area through targeted research efforts.
Questions to be addressed include: What do we know about how plants respond to different
classes of.chemicals and, in this case, what do we know about their ability to degrade these kinds
of chemicals? What do we know- about differences across plant families in terms of mechanisms
of chemical uptake? What are the physiological processes in the plants that are the most sensitive
to ALSase inhibitors? Basic research may be needed in order to understand how plants take up
and respond to chemicals. A research program might be designed that could answer these
questions in three to five years.
Workshop participants also agreed that for practical consideration, a comprehensive examination
of extant data should be conducted prior to the generation of new data. Particular attention
should be given to evaluation of the validity of these data. This process can be used to highlight
significant data gaps and identify what studies need to be repeated to validate the existing data.
Data also may be rejected for various reasons in this process, such as non-reproducibility. In
addition, a list might be compiled of the aquatic and terrestrial species known to be sensitive to
these herbicide classes, along with the available information on sensitivities. Is floral tissue most
sensitive, or seed, or reproductive endpoints, or vegetative growth? With this list, it may be
possible to develop pertinent test endpoints, whether for deterministic or probabilistic analysis.
With the development of new or additional endpoints is linked the issue of extrapolation from
laboratory or field test-based endpoints to those of interest in natural or agricultural ecosystems.
There is a significant lack of data and methods for extrapolation of the current testing data for
adequate evaluation of the impacts of the low-dose, high potency herbicides on terrestrial or
aquatic species in these ecosystems. Protocols are needed for reproductive effects studies and
currently, few studies have considered early plant growth effects and effects on reproduction.
Attempts to develop protocols for mature plants such as cherries and pears, as well as protocols
for other vascular plants, including economic and native species should be made. In concert,
methods are needed to extrapolate from: laboratory, micro- and mesocosm tests to the field; crops
to native plants; seedlings to mature plants; species tested to species untested; and serrestrial
plants to aquatic plants. Another important point that was brought up was whether data generated
from testing terrestrial plant species can be applied to rooted aquatic macrophytes.
Finally, there is a significant need for analytical methods capable of detecting the chemical at
concentrations predicted on the basis of the bio-assays to effect sensitive species. Also methods
->
J J
-------
for evaluating deposition mechanisms and distribution on plants or canopies, and for comparing
the amount of drift relative to the amount applied are needed. It mav be possible to incorporate
such methods into existinu drift models. The resulting data must be coupled with information on
exposure as a function of distance fiom the site of application with real time measurement, to
correlate exposure with effects with confidence. There was discussion but no agreement among
the workshop participants that post-registration and/or post-application monitoring might provide
a means by which to theoretically "validate" the safety determination made as a result of testing.
The ILSI workshop participants identified several specific basic research needs:
Useful Residue Detection Methods - Low dose, high potency herbicides have great potential
to move from the treated site to non-target areas during and after application. Current
analytical methods are incapable of detecting low dose herbicides at residue levels that may
injure plants, rendering foliar plant samples useless for assessment purposes. Field residue
detection methods that are sensitive, specific and easily used are needed.
Screening Methods Useful for Protection of Ecosystems - Pesticide screening tests that
provide data necessary for ecological risk assessments are needed. Key plant species that
represent ecoregions within Canada, the U.S., and Mexico should be identified and considered
for testing purposes. GIS vegetation mapping of ecoregions that includes key ecological
species, endangered and threatened plants, pristine areas, and possibly organic farms would be
invaluable in this identification process.
Probabilistic Risk Assessment Data Needs - The move from deterministic risk assessment to
probabilistic risk assessment for pesticides requires an improved understanding of data
uncertainties. Baseline research is needed to establish plant sensitivity ranges and confidence
limits. Some uncertainty issues include: the adequacy of current test species to represent non-
tested plant families and species; the predictive value of adverse effects on early plant growth
(first 14-21 days) for extrapolation to adverse effects on reproduction and yield; the predictive
value of adverse effects on plant growth observed in the greenhouse for extrapolation to
adverse effects in the field; and the relationship of single exposures to multiple exposures or
exposures to multiple toxicants. Also, baseline plant population data are needed to conduct
landscape or population effects risk assessments.
Mechanisms of Long-range Residue Transport and Plant Effects - Factors responsible for
low-dose, high potency herbicide movement great distances from the target site of application
must be identified and evaluated. The frequency and duration of exposures to non-target
plants should be determined, and resultant acute/chronic effects on plants evaluated. Existing
ozone plant toxicity research may have relevance to low-dose herbicide plant toxicity.
34
-------
+ + + + + + + +
Figure 1. The framework for ecological risk assessment
Planning
(Risk Assessor/
Risk Manager/ -
Interested Parties
Dialogue)
Ecological Risk Assessment
PROBLEM FORMULATION
to
CO
>-
<
z
<
u
Characterization
of
Exposure
Characterization
of
Ecological
Effects
RISK CHARACTERIZATION
T
JL
>
w
z
CD
O
CD
(J)
o
ja
rV
0
c
-*
73
o
O
o
(/>
03
sr
0)
o
03
r*
CO
0
o
ra
(n
V)
X
Communicating Results
to the Risk Manager
T
JL
Risk Management and
Communicating Results to
Interested Parties
Source¦ U.S EPA, 1998. Guidelines for Ecological Risk Assessment Office of Research & Development,
Washington DC. EPA/630/R-95/002F
35
-------
Figure 2. Conceptuiil Model for Terrestrial Non-Target Plants
Source Shaw, J.L. and J.A Gagne, 2000 Terrestrial non-targe! plant risk assessment for pesticides A generic problem formulation and p/oposed protection
goals Prepared for the American Crop Protection Association Environmental Risk Analysis Ovet sight Work Group
36
-------
Figure 3. Simplified diagram of the EPPO scheme for evaluation
of the risk of a plant protection product to terrestrial higher plants
EMISSION
j. tWvJ:..! ' Iffe'
Properties of pioduct
and active ingredients
,Vv,f5% V •:vWj
Source: European and Mediterranean Plant Protection Organization (EPPO), updated. Decision-Making Scheme
for the Environmental Risk Assessment of Plant Protection Products.
37
-------
Figure 4. Conceptual Model of the Source, Transport, Fate and Effects of
Low Dose, High Toxicity Herbicides on Non-Target Species and Ecosystems
SOURCE
Agricultural Application
Drift
Contaminated
Irrigation Water
Wndblown Dust
DEPOSITION
FlowetfFruit
ORGAN-LEVEL OR
ORGANISMAL EFFECT
POTENTIAL ECOSYSTEM
OR ECONOMIC CONSEQUENCE
Foliage
Soil Surface
Surface Waters
F1 cv\er Development
Fruit Development
Leaf Physidcgy and
Grovth
Physiology and GraMh of
Fee-Uving Soil Fungi and
Mcrcfces
| Physiology and Growth of
Mychcnizae
Physiology and Growth of
Rcots
Harvest
Seed Recnjtment
Wildlife Forage
Insect Foraging
Han/est/Yield
Caiton Fixation
Bcmass
[Nutrient Cycling
iPhylloplane Mcnxrganisms
Wildlife Habitat
Physiology and Growth of
Aquatic Vascular and Norv
Vascdar Rants
Physidogy and Growth of
Mcrctes and Firgi
Sal Biogeochem'stry
Organic Matter Deccrrposition
Mnearlization
Nutrient Cycling
lmm±ilation of Toxic Elements
Water Use in Rants
Nutrient Uptake and Biogeochemsiry
Species Composition
Herbage Quality
Drought Resistance
j&oamss Accumulation
(Nutrient Cyding
(Speaes Interactions/Comrrunity
Dynamics
/ildlife Habitat
Ftecreation
Detritus
Jsogeochemstry
Source: Taylor, G 1999 Ecological Risk Characterization of Low Dose, High Toxicity Herbicides Unpublished
SET AC manuscript form I LSI workshop
38
-------
Table 1. Comparison of approaches for aquatic non-target plants
U.S. EPA (1982)
Environment Canada (1993 draft)
EPPO
Description of tiers
Tier I: single dose test
Tier II: dose-response
Tier III' field testing
Tier r single dose test for algae and rooted
vascular plants
Tier II: dose-response test for algae,
duckweed, and submersed or emergent
vascular plants
Tier III: submersed and emergent plants
Tier IV: microcosm, mesocosm and field
testing
Tier 1 dose-response test for
green algae and
duckweed
Tier 2' dose-response testing for
rooted macrophyte
Triggers to next tier
Tier I to II. >50% effect
Tier II to III: case-by-case
Tier I to II: any effect
Tier II to III:
EEC> 0 1 EC50 (algae, duckweed)
EEC> EC25 (submersed, emergent)
Tier III to IV- EEC> 0 1 EC25
Tier 1 to Tier 2 PEC/NOEC
between 0 1 and 1
Number of test species
3 freshwater algae, 1 marine alga,
floating vascular plant (duckweed)
3 freshwater algae, 3 marine algae, floating
vascular plant (duckweed), rooted aquatic plants
(submersed and emergent species)
green alga, duckweed, rooted
macrophyte
Exposure regime
aquatic
aquatic except for duckweed test which is by
foliar overspray
aquatic (transport pathways to
water consider drift, gaseous
transport, run-off)
Product form
TGAI
TGAI for algae; TEP for duckweed, submersed
and emergent species
not stated
39
-------
Table 2. Comparison of approaches for terrestrial non-target plants
U.S. EPA (1982)
Environment Canada (1993 draft)
EPPO .
Description of tiers
Tier I. single dose test
Tier II: dose-response
Tier III' field testing
Tier I: single dose vegetative vigor test Plant
screening (efficacy data) considered
Tier II. dose-response test for vegetative
vigor and seed germination/root
elongation
Tier III special testing with single species
Tier IV: microcosm, mesocosm and field
testing
Tier 1 single dose tests, emergence and
vegetative vigor Plant screening (efficacy
data) considered
Tier 2 dose-response tests for seedling
emergence and vegetative vigor
Triggers to next tier
Tier I to II >25% effect for
any species
Tier II to III: case-by-case
Tier I to II any effect
Tier II to III.
EEC> EC25 (vigor)
EEC> 0 1 EC25 (germination)
Tier III to IV EEC> EC25 for 25% of species
or 50% of families (vigor)
Tier 1 to Tier 2 >50% effect for any species
Number of test species
10 crop species' 6 dicots (at
least 4 families, to include
soybeans and a root crop) and 4
dicots (at least 2 families
including com).
For herbicides 30 species (10 families)
For non-herbicides 10 species (6 families)
In Tier 1 at least 6 species (at least 6 families)
In Tier 2 at least 6 species repiesenting
families for which significant herbicidal
activity has been claimed/found
Exposure regime
Soil for seedling emergence test,
foliar for vegetative vigor test
Soil for seed germination/root elongation test,
foliar for vegetative vigor test
Soil and foliar
Product form
TEP
TGAI at lower tiers, TEP at higher tiers
An appropriate lead formulation
40
-------
Table 3. Comparison of assumptions and models used for exposure assessment
Exposure Pathway
United States
Canada
EPPO
Leaching
Not part of terrestrial higher plant
scheme
Run-off
1, 2 or 5% of maximum label rate
depending upon solubility. Use
standard scenario (1 acre pond, 15 err
deep) or models (GENEEC at lower
levels of refinement, PRZM3 at higher
levels of refinement)
10% of maximum label rate
Considered as a minor route of
exposure and only assessed in
case the uptake of the active
substance is mainly via the roots
Overspray
60% of maximum label rate
100% of maximum label rate
Not specified but normally in
Europe 100% of application rate
Spray drift
10% of maximum label rate for aquatic
plants; 5% for terrestrial plants (1% if
ground equipment used) At higher
tiers, use AgDRIFT
10% of maximum label rate
Adjusted based upon distance
from treated area For pre-
emergence applications, also
consider density of vegetation
Gaseous transport
Not considered as route of direct
exposure
Considered as a minor route of
exposure and only assessed in
case gaseous transport can cause
higher concentration as drift
events. For pre-emergence
applications, also consider density
of vegetation
Surface water receiving
models
GENEEC2 (screening),
PRZM/EXAMS (higher levels)
Not part of terrestrial higher plant
scheme
41
-------
Appendix D
Presentations Given at Non-target Plant Risk Assessment Workshop
-------
NONTARGET PLANT RISK ASSESSMENT
WORKSHOP FOR REGULATORS
Elizabeth Leovey, Acting Director
Environmental Fate and Effects Division
EPA/Office of Pesticide Programs
-------
Background
¦ During the last decade, regulators have
received an increased number of incident
reports concerning adverse effects to non-
target plants from the use of pesticides
¦ Researchers, scientific advisory groups, and
stakeholders are urging regulators to produce
more realistic and scientifically sound risk
assessments and to reduce uncertainty in
their risk assessments.
-------
Purpose of This Workshop
Exciting opportunity to explore the
various approaches which regulators
throughout the international community
use in assessing the risk of chemicals to
non-target plants.
-------
Goals of the Workshop
¦ Improve collaboration and
communication among the international
community regarding regulatory
approaches for conducting risk
assessments
-------
Goals of the Workshop
¦ Discuss the elements of a risk
assessment framework and the issues
associated with exposure and ecological
assessments for non-target plant risk
assessments
-------
Goals of the Workshop
¦ Identify and prioritize research topics
which will address the uncertainties in
non-target plant risk assessments.
¦ Develop a set of recommendations for
advancing non-target plant risk
assessments.
-------
Scope of the Workshop
Although this workshop is focused
mostly on pesticides, the discussion of
approaches and recommendations
resulting from this workshop should be
applicable to chemicals other than
pesticides.
-------
NON-TARGET PLANT RISK
ASSESSMENT FOR TOXIC SUBSTANCES
IN THE U.S. EPA, OFFICE OF
POLLUTION PREVENTION AND TOXICS
Presented bj J)iv Jen y Smrchfk, Senior Biologist,
Existing CjieiteiiSa^ Assessment Branch (ECAB)
RiskpA^iit'^sment Division (RAD)
Office of PoIIutiipl'i Prevention and Toxics (OPPT)
Office of PreventioiirPts'ticides and Toxic Substances (OPPTS)
U.S. Environmental Protection Agency, Washington, D.C.
e-mail: smrchek.jerrv'foepa.gov
Tel: 202-564-7628; FAX: 202-564-7450
NON-TARGET PLANT RISK ASSESSMENT WORKSHOP
FOR REGULATORS,
Alexandria, Virginia
January 15, 2002
-------
OUTLINE OF PRESENTATION
Introduction
-OPPTS Organization Chart
-OPPT Organization Chart
-Missions of OPPT and RAD
-Major legislative mandates/authorities
-Two-alternative approach used by OPPT to assess and
control toxic chemical hazards
-International activities
-Why do OPPT and OPP work closely together?
-U.S. EPA Ecological Risk Assessment Framework
Hazard/Risk Assessment Process in OPPT
*
-Brief history and overview
-Some characteristics of the OPPT process (Aquatic
emphasis)
-Testing Scheme I: Environmental Effects of Chemicals
-Testing scheme characteristics, iterations, and expansion
-Testing Scheme II: Environmental Effects of Endocrine
Disrupters
-Testing Scheme III: Sediment Effects of Chemicals
-Criteria for selection, and characteristics of test species
-Testing triggers or decision criteria
-Hazard ranking criteria
-OPPTS 850 Test guidelines
-Other assessment tools
-------
-Uncertainty and uncertainty factors
-How the OPPT hazard process operates
-Diagram of process
-Current OPPT deterministic risk assessment process
compared to probabilistic risk assessment process (OPP)
Hazard/Risk Assessment of Toxic (Industrial) Chemicals On
Aquatic and Terrestrial Plants
-Importance of plants under TSCA and basis of plant
hazard/risk assessment
-Differences in OPPT and OPP processes
-Aquatic plant species (algae, Lemna spp.)
-Terrestrial plant species (10 agronomic [crop] species)
-Description (Testing Scheme IV): Plant Effects of Chemicals
OPPT Problem Areas and Research Needs in Plant
Ecotoxicology
-Test guidelines
-Test species
-Test endpoints and triggers
-Substrate or support medium
-Uncertainty factors and hazard ranking criteria
-Other
Conclusions and Future Directions
-------
INTRODUCTION
-------
Office of Prevention, Pesticides, and Toxic Substances
-------
Office of Pollution Prevention and Toxics
New Organization
Office Director
Deputy Office
Director
Interagency
Testing
Committee
-------
Risk Assessment Division
Associate
Director
-------
MISSIONS OF OPPT AND RAD
9 OPPT is the lead office for implementing the Toxic
Substances Control Act (TSCA), the Pollution Prevention
Act (PPA), several provisions of the Federal Food, Drug
and Cosmetic Act (FFDCA), and the Residential Lead-
Based Paint Hazard Reduction Act,
e Core TSCA activities: reduce risk of existing and new
chemicals in the marketplace,
• Manages the Design for the Environment (DfE) program: a
voluntary, cooperative partnership among EPA, industry,
public-interest groups, and other stakeholders to promote
environmentally beneficial and economically feasible
manufacturing technologies by manufacturers,
e Manages the asbestos, lead and PCB programs to reduce
risk of these chemicals,
e Manages the Green Chemistry program to promote safer,
more environmentally friendlier chemical production
processes,
® Implements pollution prevention: pollution should be
prevented or reduced at the source wherever feasible
(source reduction),
-------
0 Manages the Persistent, Bioaccumulative and Toxic (PBT)
initiative: a strategy to identify, take action, and reduce risk
from these highly toxic, persistent chemicals that can
bioaccumulate in organisms,
e Manages High Production Volume (HPV) Challenge
Program to determine risk of chemicals with high
production quantities, marketed internationally by means
of basic health and environmental hazard testing,
• Implements the Children's Health Rule: health testing
initiative and reduction in exposure to improve
children's health,
• Manages endocrine disrupter program to screen and
identify chemicals that are potential EDs; later to test these
chemicals and reduce their risk,
9 OPPT works with other Agency offices, e.g., Office of
Water, Office of Solid Wastes, ORD, on problems and
activities of mutual interest, such as PBTs, chemical
identification, biological test method development, and
testing initiatives.
-------
9 The Risk Assessment Division (RAD) is responsible for
assessing the health and environmental hazards/risks of
existing and new chemicals, and microorganisms;
integrated hazard/risk assessments are prepared,
® RAD manages (in coordination with the Chemical Control
Division of OPPT) all aspects of U.S. participation in the
OECD Screening Information Data Sets (SIDS) program
on HPVs,
e RAD reviews and evaluates health and environmental
effects test data submitted under TSCA or other authorities
administered by OPPT,
® RAD works on developing and updating health and
environmental effects test methods and guidelines in
support of OPPT programs; coordinating guidelines
development with OPP and other national and international
standards organizations (e.g., ASTM International, OECD),
9 RAD works on developing and updating techniques,
procedures, and criteria for assessing the health and
environmental hazards/risks of chemicals and
microorganisms and ensuring that procedures and criteria
employed by OPPT are consistent with Agency guidelines
and with the current scientific state-of-the-art.
-------
MAJOR LEGISLATIVE MANDATES/AUTHORITIES
The Toxic Substances Control Act (TSCA). Public Law 94-469.
October 11. 1976.
• Congress established a number of new requirements and
authorities for identifying as well as controlling toxic
chemical hazards to human health and the environment.
There are several important sections in TSCA, specific to
this process,
« Section 4: Existing Chemicals and Mixtures. Listed on the
TSCA Chemical Inventory (80,000+). Toxicity data are
usually available and exposures are often both aquatic and
terrestrial. Toxicity testing can be required if OPPT can
demonstrate that the chemical either "...may present an
unreasonable risk of,injury to health or the environment...,"
or "...will be produced in substantial quantities, and ... it
enters or may reasonably be anticipated to enter the
environment in substantial quantities." A test rule or
consent order is written to obtain testing.
• Section 5: New (Premanufacture Notice or PMN)
Chemicals. OPPT is given notice prior to the manufacture
or importation into the US of a new chemical or an existing
chemical with a "significant new use." Prior to this
notification no toxicity information is required and for
most PMNs there is little or no toxicity information (95%).
-------
OPPT has 90 days to conduct a hazard/risk assessment and
require generation of toxicity information. A potential risk
must be demonstrated to require toxicity testing. A 5(e)
order is written to develop information. Exposures
are generally only aquatic (99%); analogue data and
Quantitative Structure Activity Relationships (QSARs) are
used in the review. More than 35,000 new chemicals have
been submitted to OPPT since 1979.
$ Section 6: Regulation of Hazardous Chemical Substances
and Mixtures. Finding is made that there is a reasonable
basis to conclude "...the manufacture, processing,
distribution in commerce, use, or disposal of a chemical"
or mixture (or any combination of these activities),
presents or will present an unreasonable risk of injury to
health or the environment. Toxicity data are present and a
regulatory action (prohibition, limitation) is taken.
9 Section 8: Reporting and Retention of Information. OPPT
can request toxicity information under section 8(d); or the
chemical industry can be required to submit certain toxicity
information under section 8(e) for a chemical, or
immediately if one "...presents a substantial risk of injury
to health or the environment..." Applies to both existing
and new chemicals. Aquatic or terrestrial toxicity data or
both are available and a rule is published by OPPT,
requiring this information.
-------
• Other TSCA sections: These deal with voluntary
submissions of toxicity information from manufacturers,
processors, distributors, trade associations, other
government agencies (domestic or foreign), academia, and
other standards setting organizations. The toxicity and
hazard/risk of a chemical or group of chemicals is
determined. Petitions (under TSCA section 2.1) can be
submitted by any citizen and will cause OPPT to review
them and take action.
TSCA is less site-specific, in comparison to assessing effects
of pesticide applications to crops. Industrial chemicals, in
general, exhibit a less specific mode of action (for example,
narcosis) than pesticides. The "field" is considered more in a
general sense than in OPP. Therefore, the process of hazard/risk
assessment and exposure (fate) assessment in OPPT must be
considered in a more general sense and must be more flexible.
An industrial chemical that is reviewed under TSCA, may or may
not be toxic; pesticides are always toxic to something.
Major Difference Between OPPT and OPP Chemical Review
Process:
OPPT: The Chemical is JUDGED INNOCENT UNTIL
PROVEN GUILTY.
OPP: The Chemical is JUDGED GUILTY UNTIL
PROVEN INNOCENT
-------
The Pollution Prevention Act (PPA). Public Law 101-508.
October 27. 1990.
@ This act does not require any toxicity testing and is not as
detailed as TSCA in terms of regulation of chemicals,
t Intent of the Act: Pollution should first be prevented or
reduced at the source,
• Pollution that cannot be prevented should be recycled,
e Pollution that cannot be prevented or recycled should be
treated,
® Disposal or release into the environment is intended to be
employed only as a last resort.
The best way to clean up the environment: Prevent
environmental deterioration (e.g., RELEASE OF TOXIC
CHEMICALS) in the first place.
-------
TWO-ALTERNATIVE APPROACH USED BY OPPT TO
ASSESS AND CONTROL TOXIC CHEMICAL HAZARDS
1. Pollution Prevention Approach
9 Less reactive,
® Actions usually taken before or early-on to limit toxic
effects to aquatic and terrestrial organisms,
9 Source reduction, alternative production processes,
alternative (less toxic) chemicals, other limitations are
used,
« Deterministic hazard/risk assessment processes used to
evaluate degree and extent of toxicity after testing.
2. TSCA Regulatory Approach
9 Partially reactive (for existing chemicals, not for new
chemicals),
9 Regulatory structure used to delimit toxic effects and is put
in place, for example, controls, limits, mitigation measures,
9 Laboratory toxicity test results (early on) viewed as
indicative of later, more widespread (field) adverse effects,
for example, after the chemical enters commerce,
-------
« An elaborate hazard/risk assessment process has been put
in place over the past 20 years; it is used to effectively
categorize TSCA chemicals and carry out the requirements
of TSCA. This process is discussed in great detail later.
3. Both Approaches
• Goal: Regulatory limits, measures taken, and discussions
lead to eventual process changes, source reduction,
mitigation or prevention of adverse effects on organisms
occurring in the environment.
-------
INTERNATIONAL ACTIVITIES
OPPT/RAD is involved in a number of international activities
with a variety of international agencies and organizations,
including the EU, OECD, BIAC, UNEP, FAO, WHO, AEGL,
NACEC, IPCS, IPPC, WTO, and ISO. OPPT/RAD also deals
with many legislative mandates/authorities, agreements,
conventions, and treaties, for example, NAFTA, EC Directives,
MAI, and mutual acceptance of data (MAD) as an OECD
member country.
Activities include test guideline updating and development,
hazard/risk assessment, outreach, labeling, trade, technical
cooperation, hazard classification, sustainable development,
reviewing specific chemicals of international interest, pesticide
joint reviews, risk reduction/management strategies, reviewing
environmental exposure of chemicals, Green Chemistry aspects,
and participating in various workshops and meetings in
toxicology.
There is an increasing awareness that chemical safety must be
dealt with by implementing a coordinated global strategy. It is
realized that what occurs in one country can affect other
countries and thus there should be a harmonization of programs
for assessing and managing the hazard/risk of chemicals
marketed worldwide. Thus, it is veiy important that OPPT
remain heavily involved in international activities.
-------
WHY DO (SHOULD) OPPT AND OPP WORK
CLOSELY IN HAZARD/RISK ASSESSMENT OF
TOXIC (INDUSTRIAL) CHEMICALS AND
PESTICIDES?
* As part of the same larger office, OPPTS, we are required
to work together to carry out our missions. For example, a
program or action may affect both OPPT and OPP. We
have several activities where we have, or are developing
harmonized OPPTS approaches or products (e.g.,
harmonized ecological effects test guidelines; plant non-
target aquatic testing guidelines between OPPTS and the
Canadian Pest Management Regulatory Agency),
0 A chemical registered as a pesticide (under FIFRA) may
also be marketed as a new chemical (under TSCA). An
inert ingredient of a pesticide formulation registered with
OPP may also be submitted to OPPT for approval to be
marketed as a TSCA Section 5 new chemical. To avoid
unnecessary duplication and to maximize available
resources, as well as to minimize redundant ecotoxicity
testing, harmonization of regulatory activities is encouraged
between the offices
& Commonalities in hazard/risk assessment, and other
regulatory activities exist for both industrial chemicals and
pesticides, and should be emphasized,
-------
9 A "unified" front and agreement between offices must be
established and presented, especially in the international
arena (e.g., OECD, EU, NAFTA) where the U.S. position,
Agency, or OPPTS position on an issue is at variance with
positions of other countries. Such differences must be
supported by the "best" science (e.g., support from valid,
scientifically sound test guidelines and assessment
procedures),
® Decision making for both industrial chemicals and
pesticides must be consistent and uniform; similar
decisions should be able to be reached when the same
chemical is reviewed by each office,
• Resources expended by a submitter, petitioner, or registrant
when dealing with OPPTS, will decrease when the
approaches to hazard/risk assessment used in both OPPT
and OPP are similar,
• Results of ORD and other research in ecotoxicology could
be applicable to both offices.
-------
THE U.S. EPA ECOLOGICAL RISK ASSESSMENT
FRAMEWORK
© Framework for Ecological Risk Assessment first published
in 1992,
® Guidelines for Ecological Risk Assessment published in
May. 1998; these expanded upon and replaced the 1992
Framework,
9 Purpose: to help improve the quality of ecological risk
assessments at EPA while increasing the consistency of
assessments among the Agency's program offices and
regions,
• How: identified structure for the guidelines, issues to
address, developed case studies, and prepared a set of
peer-reviewed issue papers,
«> OPPTS has been actively involved in this assessment work,
9 These guidelines must be followed in OPPT Hazard/Risk
Assessment.
-------
Planning
(Risk Assessor/
Interested Parties
Dialogue)
Ecological Risk Assessment
PROBLEM FORMULATION
~
9
&
v>.
5"
Characterization
Of
Exposure
2a
Characterization
Of
Ecological
Effects
2b
o
>
in
2
<6
ft
rt
M
4©
ft
s
e
6!
«
3
TJ
"t
O
«
x
x
-4
3
¦O
e
M •
*»*¦
O
so
s
c
V7
RISK CHARACTERIZAITON
£
Communication Results to the Risk Manger
I
Risk Management and Communicating
Results to Interested Parties
Figure 1-1. The framework for ecological risk assessment (modified from U.S. EPA, 1992a).
(See: U.S. EPA. 1998. Guidelines for Ecological Risk Assessment)
EPA/630/R-95/OO2F)
-------
Fisure 1-2 The ecological risk assessment framework, with an expanded view of each phase. Within each phase,
rectangles designate inputs, hexagons indicate actions, and circles represent outputs. Problem formulation, analysis,
and risk characterization are discussed in Sections 3,4 and 5, respectively. Section2 and 6 descnbe mteract.ons between
risk assessor and risk manager. (See: U.S. EPA. 1998) —
-------
U.S. EPA ECOLOGICAL RISK ASSESSMENT PROCESS
• Problem Formulation
-Purpose of assessment
-Problem is defined
-Plan for analyzing and characterizing risk is
determined
-Integration of available information on sources,
stressors (industrial chemicals), effects (survival,
growth, reproduction), and ecosystem and receptor
characteristics
-Generate assessment endpoints: an explicit expression
of the environmental value that is to be protected (e.g.,
non-target plant populations/communities),
operationally defined by an ecological entity and its
attributes (e.g., survival, growth, and reproduction of
aquatic and terrestrial plant populations)
-Generate conceptual models: a written description and
visual representation of predicted relationships
between ecological entities and the stressors to which
they may be exposed
-Regulatory context/management goal: protect "the
environment" from an unreasonable risk of injury
(TSCA section 2[b][l] and [2]); protect the
aquatic/terrestrial environment
-------
Analysis
-Characterization of exposure: how exposure to
stressors is likely to occur
-Characterization of effects: given this exposure, the
potential and type of ecological effects that can be
expected. Determine strengths/limitations of effects
data, and ecosystem and receptor characteristics.
Analyze data to characterize the ecological responses
under the circumstances defined in the conceptual
model(s) in a stressor-response profile (summarizes
these data on the effects of a stressor and the
relationship of these data to the assessment endpoint
with point estimates or stressor-response curves)
Risk Characterization
-Exposure and stressor-response profiles are integrated
through the risk estimation process
-Summary of assumptions, uncertainties, and strengths
and limitations of the analyses
Communicating Results to the Risk Manager
-------
HAZARD AND RISK ASSESSMENT PROCESS IN OPPT
-------
A NEW HAZARD ASSESSMENT/IDENTIFICATION
PROCESS WAS DEVELOPED AFTER THE PASSAGE OF
TSCA. A BRIEF HISTORY AND OVERVIEW:
1. Late 1970's. Implementation of TSCA. How?
® Hazard assessment schemes/methods examined from
pesticides (and from, e.g., AIBS, Monsanto,
Conservation Foundation, OECD),
0 Symposia in water pollution, aquatic toxicology
(ASTM),
$ Aquatic toxicology research, effluent testing, U.S. Fish
and Wildlife Service research, universities,
2. 1981. Development of Preliminary Environmental
Effects Testing Scheme.
3. 1982. Surrogate Species Workshop (Topics: which
species to test, how many, representativeness, what
sensitive test species?).
4. 1983. Testing Triggers Workshop (Topics: development
of criteria, etc., within testing scheme to proceed or stop,
refinement of test species).
5. 1983. Development of Document: Testing for
Environmental Effects Under the Toxic Substances
Control Act.
-------
6. 1984. Document on Assessment (Uncertainty) Factors
Developed. Topic discussed in detail at the June 2001
Scientific Advisory Panel (SAP) meeting.
7. Late 1970's to mid-1980's. Development of OPPT
Environmental Effects Test Guidelines: Federal Register,
Vol. 50 (188), under TSCA Section 4(b), September,
1985.
8. 1986-1987. Test Guidelines expanded to include
microcosm tests.
9. 1987. Workshop On the Use of Biological Mechanisms
in Ecotoxicity Testing Decisions. To begin updating of
1983 documents (No. 4).
10. 1990-1996. Harmonization of industrial chemicals
(OPPT) and pesticides (OPP) ecological effect test
guidelines. Public drafts published, April 1996, EPA
Series 850.
11. 1990-present. Continued refinement of hazard
assessment/identification through publication (ASTM
1993), participation in national (ASTM) and international
(OECD) hazard assessment/test guideline activities,
contracted research, and Agency activities (e.g.,
Ecological Risk Assessment Guidelines; Guidance on
Characterization of Risk).
-------
12. Early- 1990's-present. Harmonization activities with
OECD: Development of new test guidelines,
revision/updating of existing guidelines.
13. 1997-present. OPPTS harmonization with Canada Pest
Management Regulatory Agency of aquatic/terrestrial
plant testing under NAFTA. SAP meeting held June 27-
29, 2001.
14. 2001-present. Harmonization of industrial chemicals
(OPPT) and pesticides (OPP) ecological effects test
guidelines. Contract funded to finalize 1996 drafts.
15. 2002. As part of this contract, a plant risk assessment
workshop is held, January 15-17.
-------
SOME CHARACTERISTICS OF THE OPPT HAZARD/RISK
ASSESSMENT PROCESS
® Hazard x Exposure = RISK,
e Hazard Assessment (or Identification) Process is
compatible with the Risk Assessment Process (U.S. EPA
1992, 1998),
e Process is scientifically valid (e.g., in development, use)
and accepted by the U.S. ecotoxicological community,
e It is general, so that all classes of industrial chemicals can
be evaluated in a common process; it is usually not site-
specific, but can be easily adapted to a specific site,
e It is applicable to, and protective of both aquatic and
terrestrial organisms,
• It is able to discriminate between "degrees" of toxicity or
on a scale from low to medium (moderate) to high toxicity
to organisms,
e Process is consistent, logical, available (references), and
transparent,
e It has a sequential or tiered structure,
-------
It currently operates mainly at the organism and less so at
the population levels. Eventually, must also operate at the
cellular and community/ecosystem levels,
If there is a trade-off between over protection and under
protection: better to err slightly on the over protection end,
Process takes into account, to a certain extent, uncertainty
and variability present in biological measurements and
when dealing with biological organisms,
This deterministic process is able to give a defensible range
(better) or a single number (best) for purposes of ecological
risk characterization, risk management, and TSCA
regulatory actions,
Controlled experimental laboratory data can be accurately
extrapolated to the field (well-designed field studies may
be used to "calibrate" the lab to field extrapolation),
Mechanism is in place to revise/update the process as new
information, new analyses, etc. become available.
-------
TESTING SCHEME I
Testing Scheme for Determining
Environmental Effects
Aquatic
Vertebrate
Acute
Toxicity
TIER I
TOXICITY TESTS
Aquatic
Invertebrate
Acute
Toxicity
Terrestrial
Vertebrate
Acute
Toxicity
Aquatic
Plant
Acute
Toxicity
_J
TIER II
TOXICITY TESTS
I
Terrestrial
Plant
Toxicity
i 1 1 1—
Additional Additional Additional Additional
Aquatic Aquatic Terrestrial Aquatic
Vertebrate Invertebrate Vertebrate Plant
Acute Acute Acute Acute
Toxicity Toxicity Toxicity Toxicity
t
TIER HI
TOXICITY TESTS
I 1
Aquatic Aquatic
Vertebrate Invertebrate
Chronic Chronic
Toxicity Toxicity
I I
i
Additional
Terrestrial
Plant
Toxicity
_l
Aquatic Terrestrial Terrestrial
Bioconcen- Vertebrate Plant
tration Reproductive Uptake
TIER IV
FIELD TESTS
From: EPA (1983a, 1983b) Smrchek and Zeeman, (1998), Smrchek et al. (1993),
and Zeeman and Gilford (1993).
-------
CHARACTERISTICS OF OPPT TESTING SCHEMES
[BOTTOM]
Lowest Tier >>>>>>>>>> Highest Tier
Simple Tests
Acute Tests >
>
Complex Tests
Subchronic Tests > Chronic Tests
Mortality
>
Growth, Reproduction
Small portion >
of Life-cycle
Few Species Tested
Partial Life-cycle
>
> Complete Life-
cycle^)
More, Many Species
Tested
Less Resources Expended
>
More Resources Expended
® Testing scheme is more highly developed for aquatic
organisms than for terrestrial organisms,
9 Testing scheme is more highly developed for animals than
for plants,
® The OPPT hazard/risk assessment process is currently
deterministic, eventually it may also be probabilistic, or
perhaps some combination of both.
-------
-Iterations, expansion of Testing Scheme I:
Testing Scheme I (1980s)
Environmental Effects of
Industrial Chemicals On
Organisms
9
v
Testing Scheme V (2000-)
Population/Community/
Ecosystem-Level Effects of
Industrial Chemicals
Testing Scheme IV (1995-)
Plant Effects of Industrial
Chemicals On Organisms
Testing Scheme II (1990s)
Environmental Effects of
Endocrine Disruptors On
Organisms
Testing Scheme III (1987-)
Sediment Effects of
Industrial Chemicals On
Sediment-Dwelling
Organisms (Preliminary)
Microcosm, Model Ecosystem Tests
Mesocosm Tests
Community Tests
Ecosystem-Level Tests
-------
TESTING SCHEME II
Testing Scheme for Determining
Endocrine Disrupter Effects
Modified
Aquatic
Vertebrate
Chronic
Toxicity
TIER I
TOXICITY TESTS
Aquatic Aquatic Terrestrial
Vertebrate Invertebrate Vertebrate
Chronic Chronic Reproductive
Toxicity Toxicity Toxicity
I I
Chironomid
Life-Cycle
Toxicity
Terrestrial
Plant
Toxicity ?
TIER II
TOXICITY TESTS
Additional Additional
Aquatic ' Terrestrial
Invertebrate Vertebrate
Chronic Reproductive
Toxicity Toxicity
Special
Toxicity
Tests
I
Terrestrial
Plant
Toxicity ?
I
Aquatic
Vertebrate
Whole-Life Cycle
Toxicity "
TIER ni
TOXICITY TESTS
T
Aquatic
B ioc on cen-
tra tion 1
1
Special
Toxicity
Tests
I
TIER IV
FIELD TESTS
-------
TESTING SCHEME III
Preliminary Testing Scheme For
Determining Sediment Effects
TIER I
TOXICITY TESTS
Fish Acute
Toxicity
1
Daphnid Acute Algal
(Pore waters/ (Pore waters/
elutriates) elutriates)
Toxicity Toxicity
Amphipod Chironomid
Sediment Sediment
Acute Subchronic
Toxicity Toxicity
TIER II
TOXICITY TESTS
1
Fish
1
}
Daphnid
1
Higher
1
Amphipod
1
Chironomid
Chronic
' Chronic
Plant
Chronic
Chronic
Toxicity
1
Toxicity
1
Sediment
Toxicity
1
(emergence)
1
TIER III
TOXICITY TESTS
I 1
Invertebrate/ Fish Higher Invertebrate
Fish Full Life Bioconcentration Plant Sediment
Cycle Chronic Uptake Chronic
Toxicity' Toxicity
I I L
i
Invertebrate
Lumbriculus,
Macoma, Neanthes,
Sediment
Bioaccumulation
TIER IV
FIELD TESTS
(Modified from Smrchek, J.C. and Zeeman, M.G. 1998. Assessing Risks to Ecological Systems from
Chemicals, Chapter 3, pp.24-90. In, P. Calow (ed). Handbook of Environmental Risk Asessment
and Management. Blackwell Sci., Ltd., Oxford, U.K.
-------
CHARACTERISTICS AND SELECTION OF TEST SPECIES
TO BE USED IN OPPT ENVIRONMENTAL EFFECTS
TESTING (EMPHASIS ON ANIMAL SPECIES)
«> Easily maintained in healthy condition in the laboratory
during acclimation and testing, amenable to laboratory
culture,
® Available in sufficient numbers, of suitable life stage(s), at a
reasonable cost during most of the year, for testing,
® Representative of species likely to be exposed (similar
habitat, life history, and mode of exposure),
® Widely distributed in the particular natural environment of
interest (for example, biome, ecoregion) in the U.S. and
North America,
® Commercially, recreationally, or ecologically important,
e If possible, currently used in testing; with a database
available on the toxicity of various chemicals to the test
species,
@ Must be sensitive to some, but not all classes of chemicals.
-------
USE OF TESTING TRIGGERS OR DECISION CRITERIA IN
THE TESTING SCHEMES AT VARIOUS DETERMINISTIC
TIERS OR PROBABILISTIC LEVELS
• These are specific values, general conclusions or
characteristics developed for a chemical from fate and
effects testing, other information, comparisons with similar
chemicals, and professional judgement,
• These may be a single acute or chronic toxicity value, a fate
parameter, a range of toxicity values, or a subjective,
defined identifier (e.g., significant, high),
® Used in the Testing Scheme at points between any two tests
in a sequence or at points between Tiers, to indicate which
of several further actions or decisions should be taken or are
appropriate to emphasize,
9 Results of a test may trigger another test(s),
0 After testing is completed, decision must be made to stop
testing (e.g., and manage/mitigate risks, end the assessment)
or continue the assessment,
« For example, if chemical is found to be most toxic to marine
organisms: test additional marine organisms in same or
different groups, or both,
-------
9 Verify or confirm high toxicity (and high chemical
sensitivity) in a group of organisms,
@ These are based on literature values, research results, and
are accepted by most ecotoxicologists,
9 Testing Triggers or Decision Criteria reduce unnecessary
testing, promote consistency in the testing scheme, provide
flexibility in designing appropriate tests and increase
transparency of the process.
-------
Table 3.3 Decision criteria used to enter a tiered testing scheme
applied in assessing the ecological hazard or effects of an
industrial chemical or pesticide, or to move from one tier to
another. (Modified from Smrchek et al., 1993; Zeeman and Gilford,
1993.)
Decision criteria or triggers
Appropriate Tier
TIER I
"High" production level
Chemical may be "toxic",
LC(EC)50 <100 mg/L, test(s) valid?
Chemical has "moderate" or "high"
acute toxicity, LC(EC)50 <.100 mg/L
Toxicity information is incomplete,
more test results needed
QSAR (or analog chemical) analyses
indicate "some" toxicity
Chemical is persistent (Half-life
is >4days)
Chemical is bioaccumulative
(Log Kow [or log Pow] >3.5)
Predicted or monitored environmental
concentrations is "high" or
"significant"
Enter Tier I
Enter Tier I
Enter Tier I and/or
Tiers II & III
Enter Tier I, complete
base set testing
Enter Tier I, complete
base set testing
Enter Tier I
Enter Tier I
Enter Tier I
TIER II
Results of acute base testing
(most sensitive test species)
indicate ' high acute toxicity,.
LC(EC)50 <1 mg/L
Chemical will partition to a
specific environment (e.g.,
estuaries)
Confirmation of high acute
toxicity is needed
(e.g., to plants)
Chronic effects are indicated
in chemical or analog
Testing indicates "low" toxicity
LC(EC)50 >100 mg/L
Proceed to Tier II, and/
or conduct chronic
testing (Tier III)
Proceed to Tier II,
conduct additional acute
testing (estuarine spp.)
Proceed to Tier II,
conduct additional acute
testing
Proceed to Tier II or III
Stop- exit Tier I or II
-------
Table 3.3 continued
TIER III
Chemical is persistent (half-life
is >4 days)
Chemical is bioaccumulative
(Log Kow >4.2)
Chemical has high acute toxicity
based on additional acute testing,
LC(EC)50 <1 mg/L)
Chronicity indicated by fish test
(ratio of 24h LC50/96h LC50 >2) or
by invertebrate test (24h EC50/
48h EC50 >2)
Chronic toxicity low (MATC >10 ug/L)
Proceed to Tier III
Proceed to Tier III,
conduct bioaccumulation
testing
Proceed to Tier III,
conduct chronic testing
Proceed to Tier III,
conduct chronic testing
Stop- exit Tier III,
complete hazard
assessment
TIER IV
Chemical is persistent, with
"significant" field exposure
Proceed to Tier IV,
conduct field testing,
complete hazard
assessment, then risk
characterization
Chemical is bioaccumulative
Chemical has moderate or high
chronic toxicity (MATC <.10 ug/L)
Proceed to Tier IV,
conduct field testing,
complete hazard
assessment, then risk
characterization
Conduct additional
chronic- testing or
Proceed to Tier IV,
conduct field testing,
then complete hazard
assessment, and risk
characterization, or
Stop- exit Tier III
complete hazard
assessment, then risk
characterization
-------
Table 3.3 continued
Predicted or monitored concentration
in environment through production,
use, disposal will be "high,"
chemical has high chronic toxicity
or field testing indicates high
toxicity, chemical is persistent
and/or bioaccumulative
Stop, exit Tier IV,
complete risk
characterization
AT ANY TIER
Chemical will partition to1
sediments (based on exposure
and/or physical-chemical
properties)
Enter Sediment Effects
Testing Scheme, Tier
level is dependent on
completed testing
Notes:
The terms vlow, moderate, high, toxic, some, and significant"
should be qualitatively and quantitatively defined in the hazard
assessment process.
Risk management, risk reduction, pollution prevention, or other
regulatory activities such as restrictions on release, a
chemical ban, or renewed regulatory review when certain
production levels are reached, can occur as outcomes from any
Tier level, usually after hazard assessment and risk
characterization are completed-.
FROM: Smrchek, J. & Zeeman, M. (1998). Assessing Risks From Chemicals For
Ecological Systems, Chapter 3, pp. 24-90. In, P. Calow (ed.), Handbook n f
Environmental Risk A ssessment and Management. BlackwellScL Ltd., Oxford, UK.
-------
HAZARD RANKING CRITERIA
® Various hazard criteria have been developed by OPPT to
rank concern (high, medium or moderate, and low) in
assessing industrial chemicals,
® These criteria are based on the results of valid toxicity tests,
® They are similar to several European national efforts. For
example, the Classification Criteria Approach used by the
Nordic Council to determine dangerous chemicals through
hazard classification and labeling of chemicals,
® See detailed table.
-------
Table. Hazard Criteria Used by the USEPA, Office of Pollution
Prevention and Toxics (OPPT), in Assessing Industrial Chemicals.
Criteria Used For Ranking Concern
Hazard Parameters HIGH MEDIUM (MODERATE) LOW
(and Test Endpoints)
1. Aquatic Exposure (Water Column)
-Acute Toxicity <1
LC50, EC50 (ppm or mg/L)
-Chronic Toxicity <0.1
MATC (ppm or mg/L)
-Log KOW >4.2
-Bioconcentration Factor >1000
(BCF)
2. Aquatic Exposure (Sediments)
-Acute Toxicity <1
EC50 (mg/kg dry weight
of sediments)
-Chronic Toxicity <0.1
MATC, EC05, EC10, EC25?
(mg/kg dry weight of soil)
>1<100
>0.1<10
<4.2>3.5
<1000>100
>1<100
>0.1<10
>100
>10
<3.5
<100
>100
>10
3. Aquatic/Terrestrial
(semi-emergent) Exposure
-Acute Toxicity (Water Column) See 1
LC50, EC50 (ppm or mg/L)
-Chronic Toxicity (Water Column) See 1
MATC (ppm or mg/L)
-------
Table. Hazard Criteria (Continued)
-Acute Toxicity (Sediments)
EC25, EC50? (mg/kgdry
weight of sediments)
See 2
-Chronic Toxicity (Sediments)
MATC, EC05, EC10, EC25?
(mg/kg dry weight of soil)
4. Terrestrial Exposure
Avian and Mammalian Wildlife
-Acute Oral Toxicity <100
LD50 (mg/kg) (<50)
LC50 (mg/kg in diet <1000
or food) (<500)
-Chronic Toxicity (<50)
MATC (mg/kg in diet
or food)
Terrestrial Plants and Soil
Organisms
-Acute Toxicity <1
EC50 (mg/kg dry wt of soil)
-Chronic Toxicity <0.1
MATC, EC05, EC10, EC25?
(mg/kg dry wt of soil)
See 2
>100<2000
(>50<500)
>1000<5000
(>500<1000)
(>50<100)
>1<100
>0.1<10
<2000
(>500)
>5000
(>1000)
(>100)
>100
>10
-------
NOTES:
Sediment chronic toxicity endpoints are only tentative at this time.
Aquatic/Terrestrial (semi-emergent) toxicity endpoints are very tentative at
this time and they are derived from both the separate aquatic and terrestrial
endpoints.
Terrestrial acute and chronic values for avian and mammalian wildlife are
tentative at present in OPPT because of limited available test information.
Criteria in parentheses are from the USEPA, Office of Pesticide Programs
and have not been finalized by OPPT.
Terrestrial acute and chronic values (and chronic toxicity endpoints) for
plants and soil organisms are only tentative at this time.
-------
TEST GUIDELINES
Recommended methods and standards to test the toxicity of
industrial chemicals (TSCA Section 4[b]). There are many
valid methods available,
ASTM International, other national standard methods,
OECD, and ISO methods are acceptable,
OPPTS Harmonized Test Guidelines, Series 850.1000 to
850.6800. These replaced previous OPPT environmental
effects tests guidelines (40 CFR part 797), and OPP
guidance. These consist of:
-Special considerations for conducting aquatic
laboratory studies (850.1000)
-Group A, Aquatic Fauna Test Guidelines (850.1010-
850.1730) [Acute, chronic, BCF, sediment,
microcosm tests]
-Group B, Terrestrial Wildlife (850.2100-850.2500)
[Avian, mammalian, microcosm tests]
-Group C, Beneficial Insects and Invertebrates
(850.3020-850.3040) [Honeybee tests]
-Group D, Nontarget Plants (850.4000-850.4800)
[Terrestrial plants, Lemna, plant uptake]
-------
-Group E, Toxicity to Microorganisms (850.5100-
850.5400) [Soil microbial community, algal tests]
-Group F, Chemical-Specific (850.6200-850.6800)
[Earthworm, activated sludge tests]
• These were published as drafts in April 1996; they are
currently being finalized by a contractor, who should be
finished later in 2002.
-------
OPPTS Test Guidelines
Hard Ta^jal Lot—Final
Revised as ot April l, isas
OPPTS
NunOer
B50.1000
Mama
Spactel considanflurid tor conducting aquatic tafcoratary atudes
<3roup A—Aquatic Fauna Teat GuMeinae.
Aquatic Invertebrate acute toxicity, (set, rrasWmalar daphnlds
Gammartd acute Icoadty test
Oydar acuta toxicity test (she! deposBlon)
Mysid acute taxldty tad
Penaett acute toxicity test
Bfralve acuta toxicay test (enferyo larval)
Ftsft acute taxldty test, (restMasar and marine '
Fish acute totfdty miOgaled by tunic add
DepfrnfcJ cftionlc tadc*y test
Myaid chronic taxldty last
Bat) eary-We stage tooddty last
Fist} ((• cycle toxicity
Oyster BCF
FhhBCF
Wtoto sadkneni acute toeddly Invertebrates, (reshwtfar
Whoto sadknent acuta tooadty kiwrtetoratea, marine
CftiwuMrtd wflmerrt Knaaty test
.Ta^ioiafeedmanl sutodvonic leoddy lest
Aquatic food chasi transfer
Generic freshwater miuucoaiii test, laboratory
Sfta iportBr aquatic mtottcusin test, Wraratay
Field testing tor aquatic oigantens
Qroup 8—T»nea trial WUdnte Teat Guideline*.
Avian acute orai toxicity test
Avian dietary toodctty test
Avian raproductton teat
Wild mammal acute toxicity
Terrestrial (sotf-cors) microcosm lest
Held testing (or terraetrial wtfcfflla
Group C Beneficial Ineartc and Inwtafci nee Test
Honey bee acuta contact hndcSy
Honey bee toxlcay or residues on toSaga
Row tasting lof pofflnators .
Group D—Nontarget Plants Test Oiirtitiiw.
Background—NortargM plant tasting
Target area priytoteortcfly
TorrestrtaJ plant taxldty, Tier I (seecfing emergence)
Terrestrial plant taxldty. Tier I (vegetative vigor)
Seed gemrinatforrtoot elongation toxicity test
Seeding emergence. Tier U
Eariy seeding graetft tadcfly test
Vegetative vigor. Tier O
Terrestrial pMWd study. Tier Ul
Aquatic plant ttodcfty test using Lamrja spp. Tien I and 0
Aquatic ptants Md study, Tlsr HI
RMzntwim-teguma tnoddiy
Piart i^Ufce and mslocation test
Qroup E—Toxfctty to Mteroimiariame Taw i
So* mtuufetal communCy tcaddty test
Algal taxidty, Tiets I and n
Group F—Chentcal-Specific Test Guideline*.
Earthworm sutodvonic tmdcty test
MwrtHfa* np^ninrtinn inhibition lev* tnr Ttartnntv •qnflllHW rtllWrtralS-
Exteting Numbers
OTS
none
797.1300
785.120
797.1800
797.1830
797.1970
797.1400
797.1460
797.1330
797.1060
797.1000
787.1830
797.1S20
79S.136
797.1905
797.3060,
.3100
7972100
797.2175
7972060
7972130.
2150
7972775
7972750
7972780
7972800
79727SO
797.1160
7972900
79728S0
TSTJSTOa
797.1050
795.150
7BSl7n
OPP
72-2
72-3
72-3
72r-3
72-3
73—1, 3
none
72-4
72-i
72-4
72-6
72-6
72-6.
185-4
none
none
72-6
none
none
72-7.
165-5
71-1
71-2
71-4
71-3
none
71-S
141-1
141-2
141-6-
12D—1
121—1
122-1
122-1
122-1
123-1
123-1
123-1
124-1
122-2.
123-2
124-2
122-2.
123-2
_DQQB_
OCCD
Kuta
Ef
712-C-
none
none
none
none
203
none
202
210.
305
none
none
none
none
none
205
206
none
none
207
-209.
96-113
96-114
96-130
96-115.
96-136
96-137
96-100
96-118
96-117
96-120
96-166
96-121
96-122
96-127
86-129
. 96-354
96-3S5
96-313
96-132
96-133
96-134
98-173
96-135-
96-140
96-141
96-142
96-143
96-144
96-147
96-146
96-150
96-151
96-152
96-153
96-163
96-154
96-363 '
90-347
96-364
96-155
96-156
96-157
86-156
96-159
96-K
36-1
86-167
-------
OTHER TOOLS USED IN OPPT HAZARD/RISK
ASSESSMENT
QSAR/SAR development and use: over 150 chemical
classes and subclasses have these structure relationships
that can be used to predict ecotoxicity,
New test methods, refinements and new applications of
existing methods, development of new test guidelines,
ECOTOX database (AQUIRE, PHYTOTOX, and
TERRETOX), other references and compilations of toxicity
data,
Surrogate species identification and application, for
example, endangered species considerations,
Uncertainty or Assessment extrapolation factors,
-------
UNCERTAINTY IN HAZARD/RISK ASSESSMENT
$ Uncertainty will always be present: cannot test all or most
species of concern (or the most sensitive), never will be
able to determine "safe" levels or truly assess the hazard of
industrial chemicals to all or even most aquatic and
terrestrial organisms. Also, it is difficult to estimate field
concern levels,
® Uncertainty is an integral part of risk assessment,
® This is because we are dealing with variable biological
systems, which sometimes act randomly, and we use
imperfect, variable test methods, and work with limited
amounts of data,
© Based on two concepts: addressing uncertainty due to
variability in testing and extrapolating from testing, and
addressing uncertainty due to the amount of available
ecotoxicity testing,
e Strategy: devise UNCERTAINTY FACTORS (also called
assessment factors, extrapolation factors, or safety factors,
to help reduce uncertainty and variability, so as to avoid or
minimize false negative results (Type II error in hypothesis
testing), or consider a chemical to be "safe" when in reality
it will be toxic, and to estimate field concern levels. It is a
number used to adjust toxicity values to arrive at a level or
concentration of concern,
-------
® Historically based on concepts mainly from water pollution
biology: application factors (AFs), acute to chronic ratios
(ACRs), and uncertainty factors as used in mammalian
regulatory toxicology,
« Sources of uncertainty (variability):
-Variability due to the range of sensitivities of species to
chemicals (intra- and intertaxa variations), including
variability due to life stage or in test conditions,
-Variability due to estimation of chronic effect levels from
acute test data,
-Variability due to extrapolations from the laboratory to
the field or to natural ecosystems,
e Handling uncertainty is important because the ultimate goal
of OPPT TSCA regulatory activities is to obtain a simple,
conservative, bottom-line single number, or a range of
numbers, which can be used as an indicator of hazard or
risk and as the basis for making scientifically credible,
defensible risk management or policy decisions,
• Uncertainty should decease as more and more tests are
completed, with more and more test species, at higher Tiers,
and with increased testing it should become "easier" to
determine a no-effect concentration,
-------
® Acute tests (at lower testing Tiers) should have more
uncertainty than chronic tests (at higher Tiers). Thus, if
there is a choice, emphasize chronic tests with sublethal
(more sensitive) endpoints (e.g., growth, reproduction),
t Uncertainty factors are based on multiples of 10 and range
from 1000 down to 1. They decrease as additional, valid,
test results are obtained,
@ Uncertainty factors represent an expansion of the
assessment factors (AFs) concept. Latter was developed by
OPPT in the early-1980s. An AF is a number used to adjust
toxicity endpoints to estimate concern levels for
concentrations of new (PMN) chemicals which, if exceeded,
could result in adverse environmental effects. AFs account
for limited available toxicity test data for PMN chemicals
and identify those for testing or additional testing,
® QSAR-calculated or Acute LC50/EC50 (actual chemical or
an analog) with one or two species: use 1000.
$ Acute LC50/EC50 with two or three species (use most
sensitive tested species), or QSAR LC50/EC50 based on 2-
3 species: use 1000-100.
® Acute LC50/EC50 with several species (use most sensitive
tested species), or QSAR LC50/EC50 based on several
species: use 100.
-------
• Chronic toxicity value (MATC) from one tested species: use
100-10.
• Chronic toxicity value (MATC) from 2-3 tested species (use
most sensitive): use W.,
$ Toxicity value derived from microcosm or mesocosm tests:
use 10-1.
$ Toxicity value derived from field test(s): use 1.
e Uncertainty factors provide a consistent policy to assign a
level of concern that is appropriate for the amounts and
types of data,
e Uncertainty factors are not considered as "safety factors "
or "margins of safety" as it is impossible to reliably define
and determine "safety" to most organisms. The first term
is an overused, value-laden term, and gives a false sense of
security.
-------
Ecotoxicology Hazard Assessment/Identification
Diagram (Aquatic and Terrestrial)
Problem
(Risk Assessment Division, Office of Pollution Prevention and Toxics;
US EPA Ecological Risk Assessment Framework, 1992,1998)
-------
PUTTING IT ALL TOGETHER: DETERMINING THE
HAZARD OF AN INDUSTRIAL CHEMICAL
Ecological Effects Characterization:
1. Review existing information: physicochemical
properties, QSAR calculations, previous ecotoxicity tests,
2. Recommend testing or additional testing,
3. Conduct testing (Tiers I, II, III),
4. Validate all test methods (before testing begins) and test
results (after testing concludes),
5. Concentrate on most sensitive tested group, select most
sensitive tested species, verify with further testing;
determine lowest (worst-case) effect value,
6. Use hazard ranking criteria (Table) to determine level of
concern. Low concern: stop; medium or moderate
concern: continue review or stop; high concern: continue
review,
7. Complete toxicity testing; for all high and medium
concern chemicals, use uncertainty factors to determine
concern concentration (CC) or a predicted no-effect
concentration (PNEC); these are concentrations which, if
exceeded, will most likely result in adverse effects to
organisms occurring in the environment,
-------
8. Complete deterministic hazard assessment/identification or
ecological effects characterization. Compare predicted
environmental concentration (PEC) or expected
environmental concentration (EEC) to the PNEC,
Risk Characterization:
9. Begin Risk Characterization (Risk Estimation). Determine
magnitude, probability of occurrence, and ratio (Quotient
Method):
• If PEC is greater than the PNEC, begin risk management
activities (e.g., risk reduction, pollution prevention),
# If PEC is "near" the PNEC, risk management activities
may or may not be needed. Other factors come into
play: for example, how near is the ratio, annual
production volume, distribution of the chemical in the
environment,
9 If PEC is lower than the PNEC, few or no risk
management activities are needed.
Communication With Risk Managers:
-------
CURRENT OPPT DETERMINISTIC RISK ASSESSMENT
PROCESS COMPARED TO PROBABILISTIC RISK
ASSESSMENT (PRA) IN OPP
e For those few industrial chemicals with a large number of
toxicity tests, PRA may be useful, applicable, and an
"improvement" over deterministic risk assessment,
# Adoption of PRA would require a modification of the
Tiered Testing Scheme (compressing of Tiers into several
levels, with more initial testing), or only partially using the
PRA approach within the existing Testing Scheme and
keeping the Tiered organization relatively unchanged,
o Both approaches can be compared by evaluating a chemical
using and comparing each approach. How similar will be
the derived ecologically significant concentrations of the
chemical, or the conclusions on hazard and risk?
« Currently, the chemical industry is often reluctant to
conduct even the limited testing for new chemicals. How
will they react when much more testing may be required?
• For most new TSCA chemicals reviewed by OPPT it is not
known beforehand if the chemical is toxic to
aquatic/terrestrial organisms. The intended purpose of the
new chemical is usually not to directly eliminate or control
pest organisms. Testing is recommended and conducted.
The results of this testing determines the regulatory
action(s) to be taken or not taken by OPPT.
-------
HAZARD/RISK ASSESSMENT OF TOXIC
(INDUSTRIAL) CHEMICALS ON AQUATIC AND
TERRESTRIAL PLANTS
-------
9 Importance of plants under TSCA and basis of plant
hazard/risk assessment
-The development of test guidelines (standards) is
mandated by TSCA and the need to evaluate certain areas
of environmental concern; testing scheme was developed
to provide guidance on how the test guidelines could be
used, and on the sequencing of tests,
-OPPT, is interested in the adverse effects of concern
that can occur as the result of the interaction of chemicals
with organisms found in aquatic and terrestrial
environments
-There are three broad concern categories: 1) result in an
undesirable reduction in or loss of a component of the
environment that is, valued for its economic or societal
worth, 2) could result in an undesirable reduction in or
loss of a component of the environment that is valued for
its future economic or societal worth, and 3) may cause
indirect effects on human health.
-------
-The test guidelines are used to develop data to evaluate: 1)
if a chemical causes or can cause an undesirable shift in the
structure of plant and animal populations through lethal
effects, 2) if a chemical causes or can cause an undesirable
reproductive effect in plants and animals, 3) if a chemical
causes or can cause an undesirable effect in development
and growth of plants and animals, 4) if the chemical causes
or can cause secondary toxic effects in organisms through
the accumulation of the chemical or its transformation
products in other organisms that serve as food sources, and
5) if the chemical affects or can affect the marketability or
usability of plants and animals through the accumulation of
substances in these plants and animals.
-Plants are important: primary producers in aquatic
environments and the main or only food source of many
aquatic organisms, a major source of oxygen in the world,
contributor to eutrophication of freshwater habitats,
directly or indirectly the prime source of food and energy
for terrestrial organisms, a food and fiber source for
humans.
This process used by OPPT differs somewhat from that used by
OPP, yet the similarities are still very close, even though TSCA
is 25 years old.
-------
• Non-target plants are treated indirectly because test species
are used to directly measure adverse effects, and these
effects are extrapolated to other plants,
e Hazard assessments are usually not site-specific (examples
of exceptions : presence of an endangered species, or when
the chemical is produced, used or discharged in a specific
identified area),
• Process for the most part, currently is deterministic not
probabilistic,
e The plan is to move from organism/population level tests to
later to the community and ecosystem levels,
e Effects on native and endangered plants should be
considered, if possible,
• Valid test results and valid test methods are used wherever
possible. Evaluate by using well-established criteria or
parameters (e.g., GLP, quality control/assurance, OPPT
validity criteria),
• Aquatic Plant Test Species: See previous discussion.
Includes Pseudokirchneriella subcapitata (Selenastrum
capricornutum), Skeletonema, and Lemna.
-------
-Plants are economically important as food, forage, or
ornamentals, and constitute a major cash crop,
-Their distribution, abundance, and taxonomic
representation suggest broad coverage of the plant
kingdom,
-They are sensitive or exhibit high toxicity to many
industrial chemicals and chemical classes, and have
been used to some extent in previous toxicity testing.
Their use in herbicide tests, heavy metal screening,
salinity and mineral stress tests, and allelopathic
studies indicates sensitivity to a wide variety of
stressors,
-An additional criterion, if more non-crop (wild) plant
species are added later, is that the species should be
ecologically important, a keystone species in one or
more ecoregions, it is an important component of a
food web, or serves in an important ecosystem
process (e.g., nitrogen fixation),
-------
Terrestrial Plant Test Species: Ten agronomic species (see
below)
A few crop plant species (in laboratory toxicity tests) serve
as surrogates for the larger universe of plants,
The main Tiers I and II terrestrial plant test is the early
seedling growth test,
The list of 10 recommended test species collectively meets
certain selection criteria (in addition to those already listed),
applicable especially to plants.
-Ease of rearing in the testing laboratory; compatible
with the environmental growth conditions and time
constraints of the test method,
-Seeds of the test species require no special
pretreatment such as soaking, chilling, prewashing,
light, or scarification,
-Results reproducible within a testing laboratory
("minimal" intra laboratory variability),
-Uniformity in test plants; plants amenable to testing
in the laboratory,
-Reproducibility between testing laboratories
("minimal" inter laboratory variability),
-------
& List of the current 10 terrestrial test species:
-Lycopersicon esculentum (tomato)
-Cucumis sativus (cucumber)
-Lactuca sativa (lettuce)
-Glycine max (soybean)
-Brassica oleracea (cabbage)
-Avena sativa (oat)
-Lolium perenne (perennial ryegrass)
-Allium cepa (common onion)
-Daucus carota (carrot)
-Zea mays (corn)
-------
TESTING SCHEME I
Testing Scheme for Determining
Environmental Effects
Expanded Testing Scheme IV (Plants) Is Being Developed
TIER I | j I
TOXICITY TESTS 1 1
Aquatic
Vertebrate
Acute
Toxicity
I
Aquatic
Invertebrate
Acute
Toxicity
Terrestrial
Vertebrate
Acute
Toxicity
0^1
Aquatic
Plant
Acute
Toxicity
Terrestrial
Plant
Toxicity
TIER II
TOXICITY TESTS
Additional
Aquatic
Vertebrate
Acute
Toxicity
1
Additional
Aquatic
Invertebrate
Acute
Toxicity
I
Additional
Terrestrial
Vertebrate
Acute
Toxicity
Additional
Aquatic
Plant
Acute
Toxicity
Additional
Terrestrial
Plant
Toxicity
' I
1 r I—
TIERIU
TOXICITY TESTS
r
Aquatic
T
Aquatic
Vertebrae Invertebrate Aquatic Terrestrial | Terrestrial
Chronic Chronic Bioconcen- Vertebrate Plant
tration Reproductive j Uptake
Toxicity
I
Toxicity
_l
1
tzL _J
TIER IV
FIELD TESTS
FROM:
EPA (1983a, 1983b), Smrchek ei (1993), and Zetman & Gilford (1993).
-------
OPPT PROBLEM AREAS AND RESEARCH NEEDS IN
PLANT ECOTOXICOLOGY
-------
Test Guidelines
• Are the current 850 plant test guidelines adequate for
hazard/risk assessment? If not, why? Formulate directed
research to correct deficiencies,
0 What new plant test guidelines are needed? Chronic
(partial to complete life cycle) test methods are needed to
evaluate chronic low-dose exposures. Assess possible
test species with longer life-cycles (greater than 30 days).
What additional acute test methods are needed?
• In the mid-1980s EPA developed an Arabidopsis thaliana
growth inhibition life-cycle test. This method was
subsequently published in Environmental Toxicology &
Chemistry, vol. 5: 55-60 (1986). It was intended to become
an 850 OPPTS test guideline but for unknown reasons this
never happened. There may be problems with this method,
however. After the small seeds are produced they dehisce
and are hard to collect and count. Also, it has been
mentioned that control reproducibility is too variable.
OPPT wishes to re-open investigation of this species and
method as a possible plant life-cycle test method to be
added to the OPPTS 850 test guidelines. Perhaps the
limitations of using this plant can be eventually solved.
-------
© ASTM International has developed a 36-d Brassica rapa
life cycle test, based on the work described in Shimabuku et
al. (1991), appearing in the ASTM publication Plants for
Toxicity Assessment: Second Volume. However, this plant
is not self-pollinating but requires pollination by hand or
the presence of bees. Can this method serve as the basis
for a new OPPTS harmonized chronic plant TG?
@ OPPTS Seed Germination/Root Elongation Toxicity Test,
850.4200. Because of test variability, lack of sensitivity,
and procedural problems this TG is to be deleted from the
final 850 guideline package. Should this decision be
reconsidered? Are there any important advantages for not
deleting this TG?
9 OPPTS Plant Uptake and Translocation test, 850.4800.
This TG prescribes tests using commercially important
terrestrial plants to develop data on the quantity of
chemical substances incorporated in plant tissues and the
potential for entry into food chains with resultant indirect
human exposure. This test method was written in the early
1980s, but has been rarely conducted. The method should
be reviewed and updated.
-------
Test Species
• Are the current 10 recommended agronomic or crop test
species for the OPPT unique early seedling growth test
(850.4230) adequate to assess the effects of industrial
chemicals? If not what other species should be added and
why? What is the scientific rationale for recommending
these 10 species? Are these species adequately sensitive to
a variety of industrial chemicals? Also, provide
recommendations on relative number and proportion of
monocots vs. dicots to be tested (may not be relevant if all
10 are tested each time). Many important plant families are
not represented by the 10. Is this absence important in
terms of species selection and chemical sensitivity, and if so
what should be done? Can the range of chemical
sensitivities in a variety of plant families be covered by
using only 10 surrogate species? What species from other
families should be added? Should wild or non-crop species
be added, and if so, which ones and why?
• Possible Additional Test Species (Long Term Wish List!)
Research would consist of development of new test
methods or updating of existing methods, validation of the
method through round-robin testing, and eventual adoption
as an 850 TG.
Terrestrial species:
& Additional annual monocot/dicot agronomic or wild
species, or both,
• Biennial monocot/dicot agronomic or wild species, or both,
9> Perennial monocot/dicot wild or non-crop species, or both,
-------
® Gymnospenns and other woody species
® Ferns
® Club mosses and ground pines
• Bryophytes: Mosses, liverworts, and horn worts
Aquatic species:
® Additional freshwater, estuarine, and marine microalgae
and macroalgae
$ Higher aquatic vascular plants: Floating monocots/dicots,
submergent non-rooted and rooted monocots/dicots, and
emergent rooted monocots/dicots
Plant Test Endpoints and Decision Criteria (Testing triggers^
9 Are the current test endpoints (mainly survival, growth and
reproduction) adequate to assess the effects due to
industrial chemicals? Should other new possible endpoints
be evaluated- pysiological or biochemical process
endpoints? How can these endpoints be used in plant
hazard/risk assessment? Which calculated test endpoint(s)
should be used: EC50, EC25, EC 10, EC05? Which kinds
of endpoints are most sensitive to some or most chemicals
(e.g., growth, reproduction, change in some physiological
process or value)?
-------
§ A variety of endpoints, some uncommon or new, could be
considered. These include decreased seed production,
reduced vegetative growth (stunting), chlorosis, necrosis,
extremely dark green leaves, failure of leaves to elongate,
internodes failing to extend, collapsed stems, leaf
deformities, leaf abscission, color changes, limited root
growth, effects on photosynthesis, respiration, protein
synthesis, and RNA synthesis. Based on EPA research,
should some of these endpoints be added to the 850 test
guidelines? If so, how and why?
• Are the decision criteria in the testing scheme adequate?
What is missing? Which new ones are especially relevant
to plants and should be developed and added?
Substrate or Support Medium
• Use of completely artificial substrates or support media
(quartz sand, glass beads) in OPPT early seedling growth
test as contrasted with using reformulated, partially or
completely "natural" substrates. Artificial substrates were
recommended by EPA researchers in the mid-1980s, in
order to reduce variability in physical and chemical
characteristics (that may in turn affect availability of the test
chemical) seen in "natural" soils and to minimize reaction
with the test chemical, carrier, or both. Possible influences
by microbial populations indigenous to native soils would
thus be avoided. Native or completely natural soils are
undesirable for testing industrial chemicals because of the
varying clay, sand, and humus components, the types and
proportions of which vary within the same soil type and
between soil types. Microbial populations also vary
-------
between soil types. Quartz sand and glass beads
minimally sorb substances, ensuring that the chemical will
be maximally available to the seedling via root uptake.
There are support media that are intermediate between
being totally artificial (e.g., quartz sand, glass beads) and
those obtained from natural environments and brought back
to the laboratory. These intermediate media consist of
reformulated mixes made from natural components such as
sphagnum peat moss, vermiculite, perlite, dolomite, and
sand. What are the advantages and disadvantages to using
such intermediate media or artificial formulated mixes? As
compared with completely artificial or completely natural
soils?
What is the current thinking on the use of artificial
support media? In spite of the advantages of using
artificial substrates, should there be provision for testing
industrial chemicals in "natural" or artificially formulated
soils to get an idea of their behavior in these soils as
compared to artificial substrates, and to allow more
realistic comparisons with natural or near-natural
environments? How much and what variability should be
allowed in "natural" and artificially formulated soils?
-------
Uncertainty Factors and Hazard Ranking Criteria
« Are the current uncertainty factors (1-1000) used for
aquatic organisms (including algae and Lemna), applicable
to, or adequate for other non-vascular and vascular aquatic
plants and terrestrial plants?
9 Are uncertainty factors developed for animals different from
plant uncertainty factors?
o Are these factors adequate to deal with variability in
plants? As a means of extrapolating to higher levels, to
other species, from lab to field?
9 Should specific, new factors be developed? Are additional
factors needed to account for variability and to extrapolate,
for example, from crop species to non-crop species, from
terrestrial vascular plants to rooted emergent vascular
plants, from non-woody to woody plants, and from
seedlings to mature plants? From aquatic animals to
terrestrial plants (NO!)?
• Is taxonomic distance between plants an important source
of variability and uncertainty?
# Hazard ranking criteria need to be developed for plants.
Can the aquatic criteria simply be converted and used for
terrestrial criteria? Or should new criteria, specific to
plants be developed? If so, how will these unique plant
criteria differ from those for aquatic organisms? Will the
values be smaller, or larger, and will the ranges be
narrower or wider?
-------
Other
® How does one deal with difficult to test substances, those
chemicals with low or poor water solubility, high volatility,
high photodegradation or biodegradation, and ones that
adsorb easily? What modifications need to be made to the
TGs?
$ Development of readily useable plant toxicity test data
bases should proceed. What is the current status of
PHYTOTOX?
® Efforts should be expended to identify keystone plant
species in U.S. ecoregions. These species could be
potential new test species.
0 How do plants respond (mode of action) to different classes
of industrial chemicals of interest to OPPT?
® How does the mechanism of chemical uptake by plants
differ between different classes of industrial chemicals? Is
there a pattern or chemical-specific relationship?
There should be a means available in EPA for completing
short-term, program-specific, small-scale applied plant
ecotoxicological projects for OPPTS. At present OPPT contracts
out work of this nature, but we would rather fund this work in-
house if possible. For example, these projects are important
because they can provide information and scientific support for
positions taken by OPPTS on issues at the international level.
-------
CONCLUSIONS AND FUTURE DIRECTIONS
-------
• The OPPT hazard/risk assessment process for industrial
chemicals (under TSCA) represents over 20 years of
development, gradual refinement, and "improvement,"
9> It applies more to aquatic organisms than terrestrial
organisms, and more to animals than to plants,
® The Agency should remain in the forefront of plant
ecotoxicology. A committment should be made to conduct
plant research of interest to OPPTS,
9 Much remains to be done in OPPT hazard/risk assessment:
-Continue to gather aquatic/terrestrial test data and
perform valid studies,
-Develop population/community/ecosystem-level tests
and assessment schemes,
-Better characterize uncertainty,
-Continue with development of new test guidelines and
updating/revision of existing test guidelines (OECD,
ASTM International, ISO),
-Continue research on development of new test methods
with new test species, for example, emergent and
submergent aquatic vascular plants, non-crop plants, and
keystone plant species,
-------
-Investigate new methods of assessment beyond the
quotient method for their applicability to industrial
chemicals, for example, probabilistic risk assessment
methods, modeling,
-Revisit question on what to test: a few species from
many groups, or many species from few groups, or some
combination,
-Revisit the question of how much toxicity information to
require in initial review steps, a large amount, a moderate
amount, or a low amount gradually progressing to a
greater amount,
-Continue to refine plant hazard/risk assessments in
relation to the Agency Guidelines for Ecological Risk
Assessment.
^ ^ ^ ^ ^
-------
NONTARGET PLANT RISK ASSESSMENT
WORKSHOP FOR REGULATORS
Holiday Inn, Old Town Alexandria
January 15,16,17th, 2002
-------
non-target plant risk assessments
Richard C. Petrie, Senior Biologist
EPA/Office of Pesticide Programs
Environmental Fate and Effects Division
-------
History of Non-target Plant
Testing - Pesticides
¦ First FIFRA/SAP review of non-target plant tests in
1978.
¦ Subdivision J Guideline published in 1982.
¦ ORD workshop to improve testing in 1990.
¦ Second FIFRA/SAP, review of non-target plant data
requirements, 40 CFR 158 in 1994.
. Third FIFRA/SAP, review of EPA/OPPTS harmonized
ecotoxicology guidelines in 1996.
> EPA sponsored ILSI workshop on low dosage, high
toxicity herbicides in 1999.
¦ Fourth FIFRA/SAP, review test tiers and adequacy of
existing aquatic and terrestrial plant tests, 2001.
-------
Ongoing Harmonization
¦ Internal harmonization of toxicity tests in EPA
¦ U.S., Canada, and Mexico Harmonization
Process under NAFTA
¦ OECD Guideline Development
¦ TO: Agree on test methods, data
requirements.
¦ TO: Gain efficiencies by sharing data reviews
and risk assessments, improve scientist
communication, and improve research focus.
-------
Phytotoxicant Use Is High In
1 The U.S.
¦ 580 Million Pounds of Herbicides used in
the U.S. annually.
¦ 480 Million Pounds in agriculture
¦ 50 Million Pounds in Industrial/Commercial/Govt.
¦ 50 Million Pounds in home and garden.
¦ High potential for non-target plant
exposure.
-------
^ Herbicide Use (cont.)
¦ Herbicides are used on > 90% of all corn and
soybean acres.
¦ Herbicides, desiccants, and defoliants are
designed to kill plants, therefore, any
movement off-target has potential to
adversely affect individual plants, plant
communities, plant populations, and
ecosystem function and structure.
¦ Many non-herbicides can also be toxic to
plants (benomyl and azoxystroben fungicides,
toxaphene insecticide).
-------
Environmental Monitoring and
Incident Complaints
¦ USGS environmental monitoring data.
¦ FIFRA 6(a)(2) reporting.
¦ American Association of Pest Control
Officials (AAPCO) reports.
¦ Reports from growers, citizens, and by-
standers.
-------
Uses Of Plant Toxicity Data
¦ Risk assessments - To screen pesticides for
their phytotoxic potential to move from the
treated field to non-target plants, including
crops; recommend restrictions on use as
needed.
¦ Post-registration incident report analysis.
¦ State/regional EPA enforcement offices -
spray tank cross-contamination, spray drift.
-------
Problem Formulation: Current Emphasis Is On
Single Species Protection
(Valued Resources)
¦ Adjacent crops, orchards, and
ornamentals of high economic value.
¦ Endangered and Threatened Species.
(706 plants listed as endangered/threatened by FWS in 1998;
672 flowering plants, 26 ferns and allies, 6 conifers, 2 lichens)
-------
Benefits Of Non-target Plant
Protection
¦ Reduce soil erosion
¦ Reduce nutrient and pesticide runoff
¦ Maintain habitat (food/shelter) for
terrestrial and aquatic animals (ie. aigae-
primary energy production in aquatic ecosystems,
submerged aquatic plants - direct relationship to
health of Chesapeake Bay clams, oysters, and fish)
¦ Maintain genetic and ecological
diversity.
-------
What Are Plant Resources Worth ($)?
¦ Quantifiable plant resources - resources used
by humans for fishing, hunting, bird
watching, recreation, food, structures, paper
products.
¦ Plant resources difficult to quantify -
aesthetics, endangered species, pollution
filters, primary energy producers, climate
stability, oxygen production.
-------
Current EPA/OPP Testing
Scheme for Non-target Plants
¦ Tier I - Maximum label dosage is applied.
(850.4100 - SE, 850.4150 - W, 850.4400 -
850.5400 - algal tox.)
¦ Tier II - Dose response testing for those
species triggering Tier II testing. (850.4225 - se,
850.4250 - W, 850.4400 - Lemn850.5400 - algal tox.)
¦ Tier III - Field testing. (850.4300 - terrestrial plants,
850.4450 - aquatic plants)
-------
AQUATIC PLANT TOXICITY
^ TESTS
¦ AQUATIC PLANTS TESTED
¦ Green algae - Pseudokerschne (Se/enastrum
Capricorn utum)
¦ Blue-green cyanobacteria -
¦ A Freshwater diatom - Navicuta pelliculosa (current
preferred)
¦ Marine diatom - Ske/etonema costatum
• Floating aquatic macrophyte -
-------
TERRESTRIAL PLANT
X TOXICITY TESTS
¦ TYPICAL PLANTS TESTED
¦ Corn
¦ Soybean
¦ Carrot, Radish, or Sugarbeet
¦ Oat
¦ Wheat or Ryegrass
¦ Tomato
¦ Onion
¦ Cabbage, Cauliflower, or Brussels Sprout
¦ Lettuce
¦ Cucumber
-------
TEST ENDPOINTS
• AQUATIC PLANT TESTS
¦ Algae tests - 96 or 120 hour test. Endpoints
include: Environmental Concentration (EC) 50 (cell
count) and No Observed Adverse Effect
Concentration (NOAEC or EC05).
- Floating Aquatic Macrophyte - 7 or
14 day. Endpoints include: EC50 (frond growth)
and NOAEC or EC05.
-------
X TEST ENDPOINTS
¦ TERRESTRIAL PLANT TESTS: (Seedling
Emergence and Vegetative Vigor)
¦ Typically greenhouse/growth chamber test of 14 to
28days duration.
¦ Endpoints include: Environmental Concentration
(EC)25 for % emergence (SE test only), plant height,
plant dry weight, and % visual phytotoxicity, plus
NOAEC or EC05.
-------
X TOXICITY VALUES
¦ EC25 - Derived from terrestrial plant tests,
used for non-endangered vascular plants
inhabiting terrestrial and semi-aquatic areas.
¦ EC50 - Derived from aquatic plant tests, used
for non-endangered aquatic algae and
macrophytes.
¦ NOAEC or EC05 - Derived from terrestrial and
aquatic plant tests, used for endangered
species assessments.
-------
Current Deterministic Risk Assessments
for Plants
¦ Risk Quotient (R/Q) = EEC/Toxicity Value.
¦ EEC Value = Expected Environmental Concentration
(Jim Carleton will explain)
¦ Level of Concern (LOC) if the R/Q is 1.0 or greater.
¦ The most sensitive toxicity endpoint is used in the
risk quotient.
> Aerial fixed wing, aerial helicopter, airblast (forced
air), or chemigation applications are high drift
methods. The AgDrift model will help us assess
buffer zones.
¦ No uncertainty (assessment) factor(s) currently used.
¦ Long-range transport not currently assessed (unless
literature available).
-------
Risk Characterization
¦ Determine extent and methods of use (where, when, why, how
much, how often used, how applied).
¦ Are non-target organisms at risk? (Crops or other high value
plants/endangered species, sensitive habitats, organic farms).
¦ Are environmental monitoring data available? (stream, river,
rainfall, reservoirs)
¦ Other data sources (literature, University research, adverse
effects reports (6a2 data base)?
¦ Analysis of fate properties in soil, water, on plant surfaces
(persistent, volatile, etc.) and organism toxicity for use
sites/areas.
¦ Are there any active degradates (toxic, persistent)?
-------
Uncertainties - Non-target Plant Risk
Characterizations
Are the proper test species tested?
Do early growth tests predict adverse reproductive
effects?
Are annual plants surrogates for perennial/woody
spp.?
Do greenhouse tests predict field effects?
Long-range residue transport and resulting low level
plant effects are poorly understood (acute, chronic)!
Residue levels causing plant injury (irreversible) are
poorly understood!
Are terrestrial vascular plants surrogates for rooted
aquatic vascular plants?
-------
ORGANISMS NOT CURRENTLY
CONSIDERED IN RISK ASSESSMENTS
¦ Lichens
¦ Fungi and Actinomyces
> Microbes (Some OECD work)
¦ Protozoa
¦ Soil Invertebrates (Some OECD work)
-------
Concerns With Current
Process
Current plant toxicity tests on early plant growth are
inadequate.
Low-dosage, highly phytotoxic herbicides cannot be
detected analytically in the field at levels that harm
plants, tend to move great distances, present unique
cross contamination problems (manufacturing and
spray tank).
Baseline plant toxicity research has not kept pace
with animal toxicity research.
Approach to endangered/threatened species
protection.
-------
Concerns With Current
X Process (cont.)
¦ OPP, SAP issues from 1994 and 1996
briefings:
¦ Early life stage test are not adequate - life cycle
testing needed.
- Test species inadequate - No native plants, no
perennial or woody plants.
¦ Lab to field relationship poorly understood.
¦ Improvements needed at the national and
international levels (NAFTA, OECD, and
ASTM).
-------
Review Recommendations to
X Risk Managers
¦ Precautionary label statements regarding
drift, surface runoff, and groundwater
contamination.
¦ Restrictions on dosage, number of
applications per crop or year, and total active
ingredient per crop/year.
¦ Incident report summary.
¦ Geographic and/or soil type restrictions on
product use.
-------
Post-registration Field and Monitoring
Studies Derived From Incident
Reports
¦ Isoxaflutole herbicide - The Agency requested field
evaluation of spray drift to adjacent crops, plus a
study to monitor surface runoff to native plants
adjacent to treated fields (ground applications only).
¦ Quinclorac herbicide - A voluntary high volume air
sampler and bio-assay plant (tomato) field
monitoring study was conducted for 2 years by the
registrant with University assistance.
¦ Azoxystroben fungicide - A voluntary survey,
research study, and analysis of spray drift injury to
sensitive tree fruits (apple) was conducted by the
registrant with state guidance.
-------
iiiierdLiion wiui uit: era -
Office of Research and
X Development
¦ Scientist to scientist meetings/briefings.
¦ Prioritization of non-target plant research
needs.
¦ Corvallis, WED laboratory - terrestrial plant
research and test methods development for
pesticides, industrial chemicals, and air
pollutants.
¦ Duluth laboratory - Aquatic plant research
and test methods development for pesticides
and industrial chemicals.
¦ Gulf Breeze laboratory - Estuarine/marine
nrnanicm rocoarrh inrl focf mafhnHc
-------
Some Research Needs Identified
> Verify that our existing test battery is representative
of all plant species.
¦ Determine appropriate species for reproductive tests.
¦ Determine the most sensitive and predictive
endpoints.
> Determine the laboratory to field ratio.
> Determine how to conduct monitoring studies.
¦ Determine linkages between plant loss and effects on
fish and wildlife.
¦ Determine causes of long-range transport, impacts
on plants exposed.
-------
X Research Needs (cont.)
¦ Assess feasibility and applicability of new
technologies: GIS mapping, Satellite infrared
imaging, mRNA.
¦ Assess plant residue collection efficiency (geometry,
surface texture, etc.).
¦ Determine number of droplets and size of droplets
needed to be phytotoxic, plant age and stage of
growth factors.
¦ Evaluate phytotoxic effects to plants from exposure
to surface runoff or contaminated irrigation water.
-------
FEST METHODS NEEDED
¦ TERRESTRIAL PLANTS:
. IMPROVEMENT:
Existing Test Species - Laboratory and Field.
Life Cycle Tests - Laboratory and Field.
. DEVELOPMENT:
Foliar Soil Toxicity Test,
Rainfall Simulation Test,
Field Monitoring.
-------
rEST METHODS NEEDED (cont.)
. TEST METHOD IMPROVEMENT AND DEVELOPMENT
FOR AQUATIC PLANTS:
¦ Classes of algae/cyanobacteria (only 4 currently
tested)
¦ Submerged Aquatic Vasculars
¦ Emerged Aquatic Vasculars (Rooted)
¦ Floating Aquatic Vasculars (one grass tested now)
¦ Foliar Exposures To Aquatic Plants
¦ Review/revise/update marine species.
-------
REFINED RISK ASSESSMENTS
FOR PLANTS
The OPP is actively engaged in the
development of a tiered risk assessment
that includes probabilistic approaches.
Atrazine assessment (Solomon, k.r. et ai.,
1996. Envir. Tox. And Chem. Vol.15(1): 31-76.)
Uncertainty factors must be determined
and assessment tools developed.
(Including Range/Confidence Limits)
-------
Interactions With U.S. Groups
¦ United States Department of Agriculture.
¦ United States Department of the Interior/Fish and
Wildlife Service.
¦ United States Food and Drug Administration
¦ United States Geological Survey
¦ United States National Oceanic and Atmospheric
Administration
¦ American Society of Testing and Materials
¦ Society for Environmental Toxicology and Chemistry
¦ Crop Life (American Crop Protection Association)
-------
INFORMATION SOURCES
¦ EPA Subdivision J Guideline
¦ EPA 850.4000 draft Guideline Series (EPA
Web)
¦ Standard Evaluation Procedures (SEP's)
¦ American Society for Testing and Materials
(ASTM) guidelines
¦ Rejection Rate Analysis Document (Crop
Life/EPA)
¦ ILSI Workshop Report (In publication)
¦ June 2001 SAP report to be available soon
(EPA Web)
-------
^ Questions
¦ Telephone:
- FAX:
. E-MAIL:
or Comments?
(703) 305-7358
(703) 305-6309
Petrie. Rick@epa .gov
-------
EFED Models in Plant
Exposure Assessments
James N. Carleton
« EPA
United St3tes
Environmental Protection
/igency
Office of Prevention, Pesticides,
and Toxic Substances
-------
Terrestrial Exposure
(screening method)
• Runoff and drift assumed from one cropped
acre to one adjacent acre.
• Drift: 1% for ground; for aerial
application
• Runoff is 1,2, or 5%of application rate,
depending on solubility (<10 ppm, 10-100
ppm, or >100 ppm). For aerial and
chemigation, efficiency = 60%.
-------
Semi-Aquatic Exposure
(screening method)
• Runoff and drift assumed from ten cropped
acres to one adjacent acre.
• Drift: 1% for ground; 5%for aerial
application
• Runoff is 1,2, or 5% of application rate,
depending on solubility (<10 ppm, 10-100
ppm, or >100 ppm). For aerial and
chemigation, efficiency = 60%.
-------
Comparison with endpoints
• Terrestrial and semi-aquatic
"concentrations" given in units of
mass/area, using screening method.
- Compared against terrestrial plant endpoints
given in same units.
• Exposures to aquatic plants (e.g. Lemna) are
true concentrations (mass/volume),
generated with aquatic models.
-------
Spray Drift Modeling
-------
Spray Drift Field Studies
Required for pesticides in some instances under
FIFRA
Movement of spray particles is not dependent on
the active ingredient
Application parameters, meteorology, and
application site affect downwind deposition
The Spray Drift Task Force conducted studies to
collectively address drift requirements
-------
TREATMENTS
Aerial View of
Test Site
2/) 00 feet
Spray Application Area
180 feet
Spray Collectors
25 feet
• • •
Wind
I
figure 1
• •
2,600 feet
BPHAY DRIFT
TASK POM!
-------
GROUND EQUIPMENT
9 combinations of:
• Application volume: 2.6 to 27 gal/acre
• Pressure: 20, 40, or 55 psi
• Tractor speed: 5 or 15 mph
~ Boom height: 20 or 50 inches
~ Four types of nozzles (spray coarseness)
-------
10
1
0.1
0.01
Fine Spray
High & Low Boom
high boom
low boom
100 200 300 400
downwind distance (m)
-------
Medium-Coarse Spray
High & Low Boom
100
10
1
0.1
0.01
0.001
0 100 200 300 400
downwind distance (m)
low boom
high boom
-------
EVALUATING AgDRIFT
Aerial AgDRIFT (AGDISP) predictions
were compared to field study results
The model compares very favorably at close
distances
The model over predicts at greater distances
but this may be a problem with the field
data
-------
EXTERNAL PEER REVIEWS
Peer Review Workshops
(Academia, US Department of Agriculture, PMRA Canada,
EPA/Office of Research and Development,
California EPA/Department of Pesticide Regulation)
Scientific Advisory Panels
-------
Spray Drift Modeling for Plants
• EFED still using screening assumptions (1
and 5%).
• AgDRIFT being modified to read
meteorological data file, for variable drift.
• Incorporate variable drift into aquatic
models (PRZM/EXAMS and 'puddle'
model).
-------
Aquatic Assessment
-------
Aquatic Pesticide Concentrations
Show Seasonal, Yearly Variations
USGS analysis of Heidelberg College data
-------
Why Not Use Monitoring Data
Instead of Models?
• Not available for all pesticides
• Generally limited in time, frequency of samples
• Data are variable in quality, quantity, and
relevance
-------
Surface Water Modeling Tiers
• 1: single site, single storm
- point estimate in time
- GENEEC
• 2: one or a few sites, many years
- Semi-probabilistic in time
- PRZM and EXAMS
• 3: many sites, many years
- probabilistic in time and space
- MUSCRAT (?)
-------
Surface Water Modeling
25 Acre Field
100% Treated
2.5 Acre I
x 2 m Pond
-------
GENEEC
Generic Estimated Environmental
Concentration
• Requires a minimum number of input parameters
• Uses conservative assumptions
- Provides estimates in vulnerable surface water
• Mimics higher tier model (PRZM-EXAMS) for a
vulnerable site
• If compound does not pass screen more
sophisticated modeling/assessment performed
-------
GENEEC
Application
•Rate (lbs a.I./acre)
•Total Number
•Frequency (days)
•Method
— ~ Spray Drift
Simulated Rain Event
Field
•Metabolism
•Adsorption
>
Runoff
• dissolved
• adsorbed
Water Body
•Hydrolysis
•Photolysis
•Metabolism
•Adsorption
-------
GENEEC Implementation
• Essentially a single event model (one runoff
event).
— Provides an upper-bound estimate on a high
exposure site for all possible use patterns.
- 10 ha field draining into a 1 ha pond, 6 m deep
with no outlet.
• Maximum runoff fixed at 10% of applied
• Spray Drift fixed at 5% of applied for aerial, 1%
for ground spray
-------
GENEEC Implementation, cont.
• Effects of pesticide chemistry accounted for
by 'modeling' PRZM-EXAMS output:
- Two day delay between application and runoff event.
- Maximum, 4 Day, 21 Day, and 56 Day EEC's
calculated.
- Effects of multiple applications accounted for only
partially.
-------
GENEEC Algorithms
1
• Incorporation - Runoff reduced 1 / Incorporation
Depth (cm)
• Degradations - First order
• Photolysis - Divided by 122.7 to account for light
attenuation with depth.
-------
Tier II: PRZM3-EXAMS
• Pesticide Root
EXposure Analys
System
Y 1
-------
PRZM - EXAMS Scenario
Identify Entire Usage
Area
Select Location With
High Expected Rainfall
Select Soil With High
Expected Runoff and
Soil Erosion (usda soil
data)
¦ V ¦;
-------
Selecting Weather for Modeling
Use actual measured weather
data from U.S. National
Weather Service
Select nearest weather
station to pesticide usage
area and selected soil
Use multiple years (36) of
weather for temporal
assessment at the site
'/"T r ¦'
¦ W'Mf. ' W! ( '«
-------
•Rate (lbs a.I./acre)
•Total Number
•Frequency (days)
•Method
PRZM-EXAMS
Weather Data
Field
•Soils Data
•Crop Info.
•Hydrolysis
•Photolysis
•Metabolism
Spiay Drift
Atmosphere
//
Volatility
w
Runoff
• dissolved
• adsorbed
Water Body
•Hydrolysis
•Photolysis
•Metabolism
•Adsorption
•Adsorption
-------
Essential Algorithms
• SCS curve number method for runoff
hydrology (water).
• Modified mixing-cell model for pesticide
extraction into runoff.
• Batch reactor kinetics (2 compartment) for
the pond.
-------
Some Representative Scenarios
• Corn: Ohio on a Cardington soil
• Cotton: Mississippi on a Loring silt loam
• Soybeans: Georgia on a Lynchburg loam
sand
• Walnuts: California on a Kimberlina silt
loam (irrigated)
• Wheat: North Dakota on a Fargo silt loam
-------
PRZM - EXAMS Results
• Models Report
- Maximum Annual Daily Concentration Values
• Concentration Values Are Sorted and
Ranked
- 1 -In-10 Year Return Period Values are Calculated For
Use in Risk Assessment
• Single day value (aeute)
• Annual average (chronic)
-------
-------
A
25 "
0 -
0
Maximum
Annual Mean
-------
Future Improvements: 'Puddle'
Model
• Stripped-down version of PRZM/EXAMS
- SCS Curve Number method with mixing cell extraction
of pesticide.
• Generates real concentrations (mass/volume) for
semi-aquatic exposure, and equivalent terrestrial
EECs (mass/area) for comparison with terrestrial
plant endpoints.
• Incorporate use info, soil, weather, fate properties
and spray drift.
• Temporal variability accounted for.
-------
USE OF INCIDENT DATA IN
PLANT RISK ASSESSMENTS
Nick Mastrota
Environmental Fate and Effects Branch
January 15, 2002
-------
The Ecological Incident
Information System (EIIS)
• A database of incidents of adverse effects to
nontarget plants and animals.
• Available to OPP staff on the LAN.
• Generates reports of individual incidents
and summary reports of pesticide AI's.
• Paper copy of source report is kept on file.
-------
Sources of Incident Data
• Reports submitted by pesticide registrants
for compliance with FIFRA section 6(a)(2).
- CFR 159.184 now allows minor incidents to be
reported in aggregate form.
• Reports voluntarily submitted by state and
federal agencies.
• Other information on incidents (published
papers, newspaper articles, web sites, etc.)
-------
Incidents of Plant Effects
• 1989 plant incidents
(41% of all incidents
Cin database).
1989 • Mostly complaints of
crop damage.
• Mostly submitted
under 6(a)(2).
• Includes on treated site
and offsite effects.
~ Plants ¦ Other
-------
Types of Plants Affected
• Field Crops
• Orchards
• Trees and Shrubs
• Lawns and Gardens
• Except for trees in residential areas, effects
on natural vegetation not reported
-------
Pesticides with Greatest Number
of Reported Plant Incidents
250 '
200 '
150 '
100 '
50 '
0 ^
~ Glyphosate
~ Flumetsulam
~ 2,4-D
~ Clopyralid
~ Isoxatlutole
¦ Metalochlor
-------
Information from Plant Incidents
• Pesticide use
• Species affected
• Response
• Magnitude
• Route of exposure
• Distance from application
• Symptoms
• Measured residues (rare)
-------
Information from Plant Incidents
• Pesticide use
• Species affected
• Response
• Magnitude
• Route of exposure
• Distance from application
• Symptoms
• Measured residues (rare)
-------
Information from Plant Incidents
• Pesticide use
• Species affected
• Response
• Magnitude
• Route of exposure
• Distance from application
• Symptoms
• Measured residues (rare)
-------
Information from Plant Incidents
• Pesticide use
• Species affected
• Response
• Magnitude
• Route of exposure
• Distance from application
• Symptoms
• Measured residues (rare)
-------
Information from Plant Incidents
• Pesticide use
• Species affected
• Response
• Magnitude
• Route of exposure
• Distance from application
• Symptoms
• Measured residues (rare)
-------
Information from Plant Incidents
• Pesticide use
• Species affected
• Response
• Magnitude
• Route of exposure
• Distance from application
• Symptoms
• Measured residues (rare)
-------
Information from Plant Incidents
• Pesticide use
• Species affected
• Response
• Magnitude
• Route of exposure
• Distance from application
• Symptoms
• Measured residues (rare)
-------
Information from Plant Incidents
• Pesticide use
• Species affected
• Response
• Magnitude
• Route of exposure
• Distance from application
• Symptoms
• Measured residues (rare)
-------
Considerations for Risk Assessment
• How many incidents are there?
• What are the magnitude of effects?
• Registered use, misuse, or abuse?
• What is the certainty of the cause?
•Was the label changed since the incident?
• Why incidents might not be reported?
• Is the data accurate? (Check source report)
-------
Lack of Incident Reports Does not
Mean Lack of Adverse Effects
• Incident may not be recognized or reported.
• Incident may not be investigated.
• Investigators may not include pesticide in screen
for residues.
• Residues may be too low to be detected.
• No standards for when registrants are required to
report incidents under 6(a)(2).
• Effects on natural vegetation not obvious, usually
not noticed or reported.
-------
CONCLUSION
Plant incident data are useful as
evidence of the pesticides, application
methods, and/or conditions that can
harm nontarget plants, but limitations
of data must be considered.
Mastrota.nicholas@epa.gov
-------
Implementing Refined Risk
Assessments for Pesticides in
the U.S. Environmental
Protection Agency
I.M. Sunzenauer,
Office of Pesticide Programs,
U.S. Environmental Protection Agency
5/3/2002 EPA Plant Workshop 1
-------
Goal
Provide regulators and the public with improved
risk assessments and characterizations by
beginning to address the following questions:
• What is the magnitude and probability?
• How certain are the assessment predictions?
• Are the effects seen across different species?
• Will density and diversity be affected?
• Are there population or community impacts?
5/3/2002 EPA Plant Workshop 2
-------
Overview of Initiative
1. Background
2. New Tiered Refined Risk Assessment
Scheme
3. Level II Refined Risk Assessment
Model
4. Outreach, Training, and Research
5. Future Directions
5/3/2002
EPA Plant Workshop
3
-------
1. Background
• Beqan initiative to investigate new methods
(1997)
• Formed stakeholder workgroup to provide
recommendations (1999)
• Held peer review workshops to review
recommendations (1999)
• Formed EFED Implementation Team to
develop an implementation plan (1999)
• Developed 4 level refined risk assessment
scheme; reviewed and supported by SAP
(2000)
5/3/2002
EPA Plant Workshop
4
-------
2. New Tiered Refined Risk
Assessment Scheme
• Level I: Screening level assessment
• Level II: Provides preliminary assessment of
probability and magnitude of effects
• Levels III and IV: More complex assessments
representing increasingly realistic biological
and exposure scenarios
• Guidance will be developed for moving to
higher levels
5/3/2002
EPA Plant Workshop
5
-------
3. Level II RRA Model
• Spatial scale is at the field level
• Effects distributions based on D/R curve to
estimate intra-species sensitivity and address
inter-species toxicity uncertainty
• Exposure distributions:
- Terrestrial based on measured concentrations of
pesticide on avian food items and avian behavior
- Aquatic based on exposure simulation model or
monitoring data
• Combined using Monte Carlo
• Aquatic chronic risk based on frequency of
NOEC and LOEC exceedences
5/3/2002
EPA Plant Workshop
6
-------
3. Level II RRA Model: Status
• Conducted a case study on carbofuran
which was used in regulatory document
• Presented as generic case study for
scientific peer review and strongly
supported as "state-of-the-art" (March
2001)
5/3/2002
EPA Plant Workshop
7
-------
3. Level II RRA Model:
Next Steps
• Finish Level 2 terrestrial and aquatic models for beta-
testing: Spring 2002
• Develop model documentation and training
materials:
• Training:
• SAP Review: Fall 2002
• Implementation: FallA/Vinter 2002
• Further Improvements:
5/3/2002 EPA Plant Workshop 8
-------
4. Outreach, Training, and
Research
• Outreach:
- Agency regulators: Workshops and
briefings
- Stakeholders and the Public: peer review
meetings, web site, and participation in
national and international workshops, etc.
• Training: Courses to provide foundation for
RRAs
• Research: Contracts, grants and inter-
agency agreements
5/3/2002 EPA Plant Workshop 9
-------
6. Future Directions
Next 2-3 Years:
• Address plants and other taxa
• Further develop approach for chronic effects
• Develop Level 3 RRA models
• Continue research support, training, and outreach
Long-term:
• Complete Level 3 and 4 RRA models
• Develop new data requirements for Levels 3 and 4
• Continue research support, training, and outreach
5/3/2002 EPA Plant Workshop 10
-------
Closing Remarks
• Work in progress
• Have made a great strides forward from
proposing pilot models to conducting
case studies
• Will result in more informed
environmental decision-making
5/3/2002
EPA Plant Workshop
11
-------
RRA Implementation Team
Chair
Ingrid Sunzenauer, M.S.
Terrestrial Team
Ed Fite, M.S. (Lead)
Ed Odenkirchen, Ph.D.
Tim Barry, Sc.D. (OPEI/OA)
Kathryn Gallagher, Ph.D.
(Lead)
James Lin, Ph.D.
Tim Barry, Sc.D. (OPEI/OA)
Other Team Members
Stephanie Irene, Ph.D.
Douglas Urban, M.S.
David Farrar, M.S.
5/3/2002
EPA Plant Workshop
12
-------
Division Contributions
Terrestrial Exposure
Tech Team
4 Scientists
3 Projects
Terrestrial Biology
Tech Team
4 Scientists
4 Projects
Aquatic Biology
Tech Team
3 Scientists
2 Projects
Water Quality
Tech Team
4 Scientists
3 Projects
GIS Workgroup
5 Scientists
3 Projects
5/3/2002
Exposure Model Development
Effects Testing Protocols
Effects Extrapolation Methods
Setting Parameters for Model Variables
EPA Plant Workshop
-------
Web Site
http://
[Www.epa.gov/oppefed1/econsk/index.htm1
5/3/2002
EPA Plant Workshop
-------
A Risk Manager's Perspective
Non-Target Plant Risk Assessment Workshop For
Regulators
January 15 - 17, 2002
-------
Today's Herbicides
Higher potency
¦ Lower use rate - g/A
¦ Potential Problem for both target and nontarget
crops
-------
Phytotoxicity to target crops
Corrected By Market Forces
¦ Isoxaflutole on corn
-------
Phytotoxicity to nontarget crops
Workshop
¦ Sulfometuron methyl (Oust) - roughly 250 sq.
miles of damage
barley, sugarbeets, corn
¦ Quinclorac (Facet) - use on rice
phytotoxic effects on tomatoes in Arkansas
¦ Clopyralid - compost
phytotoxic effects on garden plants
-------
Closer to registration - the more
difficult it is to deal with
environmental concerns
-------
Timing of nontarget plant concerns
¦ Difficult to take action based on environmental
concerns once registered
¦ A number of factors could be involved, such as
weather conditions, misuse, other pesticides, lack
of sensitive analytical method at phytotoxic levels
¦ Prior to registration
¦ Earlier response from scientist and their
management
-------
Conditions of Registration
Mechanism of last resort
¦ Field perimeter monitoring
¦ Simulated pond studies
¦ Drift studies
¦ Tile drain studies
m Littoral zone monitoring (aquatic)
-------
Conditions of Registration
Time intensive for Agency scientists
¦ Meet with Registrants
¦ Review protocols - negotiate what is acceptable
¦ Review reports and provide timely response
¦ Glean results for future use
-------
Outside support
Registrant
¦ Registrant enlist aid of grower groups to speed up
Agency response
-------
Outside support
States, Environmentalists, Private Citizens, Adversely Effected
Growers
¦Need to involve other interested parties sooner in
the process
¦ Input - State FIFRA Issues Research and
Evaluation Group (SFIREG)
¦Notice of Registration for new chemical - Eco
risk assessment could be put in Federal Register
Docket
¦Office of Pesticide Programs Web Site -
http://www.epa.gov/pesticides
-------
Risk Managers And Risk Assessors
As Partners
¦ Scientist has important support role
¦ Interaction between Registration Team Members
and Scientist
¦ Explain the Numbers - EC25, EC50, RQ, TER
¦ Reiterate Triggers - Support Label
Recommendations
¦ Help Risk Manager to understand science
decision
-------
Non-Target Plant Risk Assessment:
Issues and Opinions
Jane Staveley
ARCADIS
ARCADIS
I
-------
What is the best way to conduct a
non-target plant risk assessment?
A: Realistically - what is currently in the "tool
box" or what tools can be added with little
difficulty?
¦ B: Ideally
How do we move from A B in a pragmatic
way?
ARCADIS
2
-------
Which pesticides/uses require
assessment?
¦ Determine during problem formulation (if there is no
potential for exposure, risk is minimal or non-existent)
¦ Exposure considerations
- Use information (how and when applied)
- Fate and transport information (physical and chemical
properties)
¦ Effects considerations
- Mode of action information
- QSAR evaluation
¦ Non-herbicides should be evaluated
ARCADIS
-------
Framework for ecological risk
assessment (U.S. EPA)
Ecological Risk Assessment
Planning
(Risk Assessor/
Risk Manager/ -
Interested Parties
Dialogue)
PROBLEM FORMULATION
~~T~
CO
<2 Characterization
_3 °f
¦< Exposure
Z
<
~rr
Characterization
of
Ecological
Effects
RISK CHARACTERIZATION
>
w)
ra"
0
1
w
in
cu
I*
5. xj
o E.
CD O
0)
% ?
5T
0
EU
o
r>
ARCADIS
Communicating Results
to the Risk Manager
i
Risk Management and
Communicating Results to
Interested Parties
-------
Problem formulation
¦Steps:
- develop goals
- select assessment endpoints
- develop conceptual model
- prepare analysis plan
•May need a "base set" of required information to
facilitate comparative risk assessments
¦A properly-done problem formulation will be
specific to the pesticide
¦Trade-off:
- generic approach allows more predictability
- specific approach allows more realism
ARCADIS
-------
Conceptual model for generic
pesticide application
SOURCES
RELEASE
MECHANISMS1
MIGRATION
PATHWAYS /
EXPOSURE
MEDIA
RECEPTORS
ARCADIS
m
6
-------
The use of a "tiered" approach
¦"Tiers" or "levels" of assessment are generally
considered useful (conserves resources)
¦Need careful evaluation of "triggers" between levels
(e.g., avoid reliance on "most sensitive species")
¦Objective of first level: to easily make decisions
about pesticides that do not pose a risk and need no
further consideration. The first level should thus be
conservative and protective
¦ Low risk and high risk: easily managed
¦Medium risk requires most of the effort
¦ Insufficient information to make an informed
decision? Requires more analysis!
-------
Chemical properties
¦Information to refine risk assessment: fate
properties, MOA, MOU
¦Use of TGAI (at lower levels) vs. formulated
product (at higher levels)
- Practicality: want to use study on TGAI to
support various potential formulations
- Realism: formulated product is actually used
- Possible compromise: include two extra
treatments in TGAI study, formulation blank and
highest concentration (formulated)
ARCADIS
8
-------
Exposure assessment
¦Should be linked to the effects assessment
¦Generic assumptions at lower levels of assessment
¦More refined evaluations of exposure at higher
levels of assessment
- Spatial and temporal variability
- Scale
ARCADIS
9
-------
Effects assessment
¦ Factors useful in selecting test species
- Geographic occurrence
- Representativeness
- Previous use as a test organism
- Has several responses that can be observed, with low
variation in controls
- Sensitive to at least some chemicals
- Cover a spectrum of taxonomy
¦ How many test species? Which species to test?
- Answer may be variable, depending on use pattern,
MOA, MOU, other factors
¦ Are important groups currently missed?
10
-------
Effects assessment
¦ Can efficacy data be used?
¦ What endpoints should be generated?
- Measuring population effect (EC50 for algae) vs.
individual effect (EC25 for terrestrial plants)
¦ Need for reproductive endpoints?
- difference between reproductive and growth
endpoints for a species vs. the variation in growth
endpoints among species
¦ Other effects (physiological, biochemical)
¦ Multi-species tests
¦ Extrapolation between species
ARCADIS
-------
Risk characterization
¦ Integrate exposure and effects information, describe
uncertainties
- At least one input variable is a distribution
- Output is usually expressed as a distribution
¦ Expressing uncertainty
¦ Presentation of results
- Deterministic more easily understood
- Probabilistic more difficult to present
¦ Recovery
¦ Mitigation
¦ Monitoring
¦ Deterministic
Probabilistic
ARCAuib
12
-------
What is the best way to conduct a
non-target plant risk assessment?
»Follow EPA risk assessment framework
¦ Problem formulation
- Develop goals
- Select assessment endpoints
- Develop conceptual model
- Prepare analysis plan
¦ Characterize exposure and effects
¦ Characterize risk
¦ More data and analysis at higher levels of assessment
¦ Move from deterministic approaches at lower levels to
probabilistic approaches at higher levels
¦ Integrate risk assessment with risk management in an
iterative manner
ARCADIS
-------
Research needs and action items
¦ Mine existing data to look at differences in species
sensitivity and explore correlations
- Review registrants' raw data for efficacy tests
- Develop acceptance criteria for efficacy data
¦ Additional development and validation of test
methods for other species (e.g., submersed and
emergent aquatic plants) prior to requiring their use
¦ Extend ECOFRAM to non-target plants
¦ Develop generic approaches and examples of
specific "case studies" for non-target plant
ecological risk assessment
-------
Facet (Quinclorac)
Herbicide Application
on Rice in Arkansas
¦ A Case Study-
Dick A. Watkins
US Environmental Protection Agency
Region 6 - Dallas, Texas
-------
Facet (Quinclorac)
a Systemic herbicide for ann grasses
¦ Tank mix w/ propanil or Arrosolo
for preemergence
¦ Tanked w/ Prowl after dry plant
¦ Surface rate .033 and .67 lbs/acre
¦ Solanaceae sensitivity
¦ AR labeled Facet herbicide for
Rice in 1993
-------
mMmmJlimti
Specialty crop - unique crop Man. practices
44% of US rice production in AR
Facet used by 85% of AR farmers
1.37 million acres - $766 million value in 1997
Replaced two applications of Propanil on
resistant barn yard grass.
Facet applied by air in Arkansas
Mono-culture crop in Northeastern Ark
-------
*
-------
Arkansas Tomatoes
¦ $7.9 Million in tomato production
(most outside rice areas)
¦ 10 Acres of tomatoes in Poinsett
¦ Tomato production located along
Crowleys Ridge
¦ Gardeners affected
-------
-------
Gathering Storm
¦ Region began receiving complaints
¦ FIFRA delegation
¦ Referrals to Ark State Plant Board
¦ Plant Board Investigations
¦ Facet vs 2,4-D
¦ Tomato Farmers continuing loss
¦ Rice vs Tomatoes
-------
-------
-------
-------
-------
-------
-------
-------
- tM -
v « %» « ~ -
V, IT,ai
«• • V* cr-r
.W. ?£. .
&j S
¦ *
/ ,
«¦' "" -,.»*¦
1r*3*
' ?
9mh
-------
-------
Stage is Set
¦ Political firestorm
¦ Gov./Plant Board
¦ Press interest increased
¦ EPA Registration Division
. BASF
¦ State/Federal Congressional
Interest
¦ U of A/Extension Service
-------
What do We Need to
Know About Facet?
¦ Is it moving from the application
site
¦ Is it deposing on tomato production
¦ Is it affecting tomatoes
-------
Two Pronged Attack
One study was conducted by the
ACES to determine tomato
sensitivity to facet exposure
Another study was conducted by the
U of Arkansas to assess the
possibility of facet intrusion at
growing sites.
-------
Facet Study
¦ Region 6 provides grant for study
¦ EPA Houston Lab analyze air samples
¦ BASF contributes lab capability
¦ U of Ark provides manpower
¦ Coop Extension contributes plots/research
¦ Plant Board provides grant admin, and
supplies
¦ Farmer provides housing
¦ Tomato growers provide sampling source
-------
-------
-------
-------
-------
Field Study Findings
¦ All five sites experienced abnormal
growth.
¦ Quinclorac detected from the beginning
of sampling - highest with applic window
a Ambient air contained detect. Quinclorac
on many days
¦ Repeated low level exposure
¦ No soil sample detection
¦ Detection on days when no application
Within 5 miles.
-------
Grow-out results...
(0.001,0.01,0.1 X)
¦ Injury increased with rate increase
¦ Multiple applications resulted in more
injury than single applications at all rates.
¦ As rates increased, fresh weights
decreased significantly
¦ Reduced flowering followed injury
pattern
¦ More sensative than the analytical
procedures used to chem. detect
quinclorac
-------
Facet Status in AR
¦ Other products on line - New path
¦ Biotech - Clear Field
¦ Liablitiy - 3 civil lawsuits
¦ Weed resistance
¦ Facet restrictions
-------
Implications for Non-
Target Plant Risk
Assessment
¦ Registration
¦ Enforcement
¦ Laboratory Capability
¦ Economic Implications
¦ Legal Issues - Law suits/trespass
> Health Ramifications
¦ Political Inequities
¦ Ethical Considerations
-------
Non-target Crop
Herbicide Spray Drift Mitigation
in California
Terrell Barry, Ph.D.
Environmental Monitoring Branch
California Department of Pesticide Regulation
California Environmental Protection Agency
-------
Selected California Rice Herbicides
Past, Present, Future
• 2,4-D
• Super Wham, Stam EDF - Propanil
• Shark - Carfentrazone ethyl
• Regiment - Bispyribac-sodium
• Clincher - Cyhalofop butyl
-------
Case Study: Propanil
-------
History of Propanil Use
• 1957 - First synthesized
• 1961 - First use on rice
• 1968 - Rapid expansion of use in California until
1968 when 210,000 acres were treated
by air. Widespread damage and
defoliation of sensitive crops occurred
• 1969 - CDFA issued use prohibition north of
Sankey Road
-------
Propanil History - Continued
• 1982 - The Westside Propanil Use Area formed
- compliance monitoring
- prune foliage threshold = 0.1 ppm
- Helicopter applications only
• less than 60 mph air speed
• 600 to 800 |xm VMD
• 500 acres/day
• 10 gal/acre
-------
Propanil History Continued
• 1997 - Expanded Use Area
- emergency regulation package
- ground applications allowed valley wide
- aerial applications in Butte County Study Area
- buffer zones
• air - 4 miles to sensitive crops
• ground - 1 mile to sensitive crops
- compliance monitoring
-------
Progressive Off-site Problems since 1997
• 1996 - Westside use only
- residues on prune foliage at the detection limit
• 1997 - Westside and Expanded Use
- residues above the detection limit in several samples
• 1998 - Westside and Expanded Use
- a few residues above the threshold of 0.1 ppm
• 1999 - Westside and Expanded Use
- many residues above the threshold of 0.1 ppm
- spotting observed
• 2000 - Westside and Expanded Use
- many residues above the threshold of 0.1 ppm
• 2001 - Westside and Expanded Use
- many residues above the threshold of 0.1 ppm
-------
IN
A
®
Monitoring Locations
WM
Vines
¦i
Prunes
¦¦
Rice
10
0 10 20 Miles
Map created by Johanna Walters
January 2002
-------
Propanil Permit Conditions
• Propanil is a California restricted material
• Compliance monitoring is required by regulation
• California Propanil Task Force
- formed to conduct studies to provide
information to refine the California Propanil
regulations
• Studies conducted since 1999
- Large scale air study
- Ground studies
-------
Existing Propanil Ground Buffer Zones in
California Regulation
Note: Ag. Commissioner may modify
• Prunes, Grapes, Pistachios
- 2 miles
• Cotton
- 1/2 mile
• Modification guidance by DPR policy letter
- 1/2 mile with wind away to all crops
- based upon empirical fit of deposition field
study results
-------
Existing Propanil Aerial Buffer Zones in
California Regulation
Note: Ag. Commissioner may modify buffer size
• Aerial applications are allowed in only two
areas
- The Propanil Aerial Use Area (West side)
• Glenn County
• Colusa County
- The Butte County Study Area
• 4 mile buffer zone to all sensitive crops
-------
Analysis Basic Approach
• Prune foliage spotting
- dose/response data
• Prune foliage vs. horizontal deposition
- co-located prune saplings and mylar
• Downwind horizontal deposition characterization
- Field data
- Empirical modeling - ground
- ISC - far field (>300m) - aerial
-------
Dose-Response Data
California Propanil Task Force (CPTF) 1999 and 2000 field study results.
Plot shows, for 99 co-located pairs of potted prunes, percent of total leaves
on a potted prune sapling showing any symptoms versus prune foliage
propanil residue (ppm).
100
—
90
—
80
—
70
—
60
—
CO
-
>
50
—
SO
-
40
—
30
—
20
—
10
—
0
—
PPm
-------
Prune Foliage Concentration
vs.
Horizontal Deposition
E
Q.
CL
c
o
-4—'
ca
-*—'
c
0
O
c
o
O
o
CD
03
O
LL
10.00 -
1.00 -
0.10
0.01 -I
0.0001 0.0010 0.0100 0.1000 1.0000
Horizontal Deposition (ug/cm**2)
-------
Ground Application
Downwind Horizontal Propanil Deposition Characterization
XRTeeJet 8004 / 36" boom height
4 lbs/acre app. rate 3.3 Acre, 2 pass (2400 ft x 60 ft)
CM
¦K
¦K
E
c
o
CO
o
Q.
CD
Q
"CD
C
o
.N
0
1
0.0100 -
• •
0.0010 —
0.0001 —I
i 1 1 n 1—1 n~~
1 10 100
1000
distance (m)
-------
Large Scale Aerial Studies
CPTF
• 1500 acres treated on a single morning within
an area 2 miles by 5 miles (6400 acres)
• Standard aerial application practices
• Sampling sites out to 4 miles
- Mylar horizontal deposition sheets
- potted prune saplings
• Meteorological data collected on-site
-------
Samplers located downwind
1/4 mile to 4 miles
Two potted prunes and a horizontal
Mylar sheet at each station
2 section by 5 section
application area
-------
2 section by 5 section area
(6400 acres)
1500 acres treated on one
morning
Standard grower practices
All applications in the
surrounding areas were
halted
PROPANIL STUDY A-1
JUNE 11 1999
¦ ~r; TREATED FIELDS
-------
Large Scale Aerial Studies
CPTF
Potted prune foliage propanil concentration (ppm)
E
Q.
Q.
C
o
cc
o
e
o
o
CD
O)
03
10.0 -
1.0 -
0.1 -I
1000 10000
distance from downwind edge of test area (m)
-------
Horizontal Deposition Modeling
• Gaussian Plume model algorithm (ISC model)
- straight-line
- steady-state conditions
- area source
- Dry Deposition option
- deposition by:
• turbulent diffusion
• gravitational settling
-------
Results from Large Scale Aerial Study
ISC Computer Simulation
C\J
*
¦*
E
o
o>
c
o
CO
o
Q.
Cl>
"D
TJ
CD
3
CO
co
a>
E
1.000 H
0.100 -\
0.010 H
0.001 H
0.001
0.010
0.100
1.000
modeled deposition (ug/cm**2)
-------
Shark® Herbicide Labeling
• Section 18 label
-1998
-1999
• Section 3 label
-2001
-------
Shark - Carfentrazone ethyl
• Shark field studies by FMC
- Prune Dose-Response
• prune foliage concentration and response
• co-located horizontal deposition
- Ground Application
• in-swath horizontal deposition
• downwind cotton and tomato bioassay
-------
AgDrift Modeling
by
FMC
• Ground Tier I
- Wind speed =10 mph
- coarse droplet spectra
- lOgal/A finished spray volume
-------
AgDrift Modeling
by
FMC
• Air Tier I
- wind speed =10 mph
- coarse droplet spectra
- default application conditions
- 10 gal/A finished spray
-------
Application of Analysis and Modeling
to
Shark Label Language
• Suggested Ground Buffer Zone = 2640 ft
• No Aerial Application of liquid spray
-------
science for a changing world
Use of Chemical Benchmarks for
Decision-Making in the Risk
Assessment Process
By
James F. Fairchild
U.S. Geological Survey
Columbia, MO
¦¦¦¦i
-------
Personal Research Background
• Research aquatic ecologist
- Development of aquatic plant toxicity tests
- Chemical specific risk assessments
- Laboratory:field validation of risks
• No regulatory function
-------
¦¦¦¦¦¦
Emerging Needs for Risk Assessment
• Selection of minimum number species
• Selection of appropriate test species
• Development of test methods
• Selection of appropriate endpoints
• Developing tiered testing process
USGS
¦
-------
Developing tiered testing process
- Uniformity necessary at earliest tiers
- Additional data required as risk is indicated
- Problem formulation increasingly important
during tier progression to reduce time/cost
- Decision-making must be science-based
- Use of chemical benchmarks is sound
approach for data prioritization
1USGS
-------
What is a chemical benchmark?
• A dataset with extensive fate/effects
information that can be used prioritize
data collection efforts for new chemicals
- Chemical
- Chemical class/mode of action
- Use pattern
1USGS
-------
Advantages of chemical benchmarks
• Benefits from existing data
• Narrows data collection efforts
• Saves time and money
• Science-based
• Amenable to problem formulation effort
1USGS
mmmm
-------
Atrazine as a chemical benchmark
• Widely used
• High use rates
• Uniformly toxic as photosynthetic inhibitor
• Water soluble
• Highly persistent in environment
• Extensive database
• Old chemical with few proprietary concerns
• Worst-case example for fate and effects
IUSGS
-------
Examples for presentation
• Triazines
- Atrazine
- Metribuzin
• Acetanilides
- Alachlor
- Metolachlor
IUSGS
-------
Community response profiles
(/)
<1)
¦
o
0)
Q.
(/>
100
SUSGS
0
Atrazine
Alachlor
Metribuzin
Metolachlor
1000 2000
Concentration (|ag / L)
3000
-------
Lemna response to 4 herbicides
O100
C
80
o
o
60
o
xO
40
20
0
Atrazine
Alachlor
Metribuzin
Metolachlor
o
500 1000 1500 2000 2500
Concentration (jag / L)
3000
USGS
-------
Se/enastrum response to 4 herbicides
—¦¦— Atrazine
Metribuzin
- Alachlor
Metolachlor
150
0
iUSGS
100 200
Concentration (^g / L)
300
-------
Use of benchmarks for chemical fate
• Currently data required for hydrolysis
and photolysis only; no field fate
• Fate is modeled in early tiers
• Model data frequently inaccurate
• Actual aquatic fate data scarce for
many chemicals
IUSGS
-------
Modeled Herbicide Risk Ratio
Selenastrum Risk
Lemna Risk
Herbicide
(EEC/EC50)
(EEC/EC50)
1% runoff
10% runoff
1% runoff
10% runoff
Atrazine
0.3
3.1
0.5
4.8
Metribuzin
0.5
4.6
0.5
5.5
Alachlor
24.4
244.8
0.7
7.4
Metolachlor
1.8
18.2
0.4
4.1
Assuming 10:1 watershed:wetland ratio with 15 cm depth.
-------
Risk Estimate Compared to
Measured Environmental Exposure
Herbicide
Median
Plant
response
(ug/L)
1%
Exposure
Scenario
Maximum
Exposure
(ug/L)
95th Percentile
Exposure
(ug/L)
Atrazine
92
73
7 - 69
1 -1 1
Metribuzin
36
20
1 - 25
<2
Alachlor
584
147
1 - 65
0 - 4
Metolachlor
1138
140
1 -9
1 - 9
Toxicity values from eleven species comparison (Fairchild et al. 1998)
Exposures in 7 Great Lakes Tributaries (Richards and Baker 1993)
iusgs
-------
Comparison of triazine fate in water
Atrazine
- Hydrolysis: stable
- Photolysis: stable
- Biological: NR
Metribuzin
- Hydrolysis: stable
- Photolysis: 4.2 h
- Biological: NR
EEC (1% loss): 73
EEC (1% loss): 20 ug/L
Aqueous Half-life: 47 d
Aqueous Half-life: 5d
IUSGS
-------
Fate characteristics of herbicides
Herbicide
Water
Solubility
(mg/L)
Koc
Soil
half-life
(d)
Aqueous
Half-life
(d)1
Atrazine
33
160
60
47-193
Metribuzin
1220
41
30
4-16
Alachlor
242
190
14
5-21
Metoiachlor
530
200
20
19-75
Lower limit from 0.1 ha mesocosm (Fairchild et. al. 1994;2002)
Upper limit from 550 ha reservoir (Spalding et al. 1994)
a m
-------
Merging Fate and Effects Data
Half Life vs. Risk Ratio (EEC/EC50)
~0
O)
"ro
c/)
o
-------
Merging Fate and Effects Data
Aqueous
Half Life 25
(d)
EM
Uyanazine
153
166EC50
(ug/L)
200
734
869
1537
EEC (ug / L)
-------
Conclusions
• Chemical benchmark approach may be
defensible way to prioritize data needs
for registration
- Plant sensitivity profiles frequently similar
within and across chemical classes
- Fate data inaccurate from current models
- Empirical determination of fate imperative
• Microcosm approach with benchmark chemical
1USGS
-------
Proposed Research for the
Protection of Nontarget Plants
Thomas Pfleeger
U.S. EPA, Office of Research and Development
Western Ecology Division
Corvallis, OR
-------
Objective
To improve OPP risk assessments used in
the registration and re-registration of
pesticides.
-------
Existing Plant Tests
mergence
• Vegetative Vigor
-------
What test species represent the flora of the
United States?
Is the response of a seedling similar to a
mature plant?
Do the existing plant tests protect
reproductive or other developmental
processes?
What is the relationship between greenhouse
tests and reality?
-------
Conclusion from Previous Work
Low concentration can be disruptive to
reproductive process with minimal leaf damage.
SU's are more active at below field application
rates than other herbicides at similar below
field application rates.
• Current guidelines under FIFRA for the
protection of non-target plants are inadequate
-------
Physiological Studies
Alternative modes of actions
Physiological / biochemical endpoints
Pesticide effects on plant physiology
(i.e.,+/- sensitive to pathogens, herbivory,
transpiration, peroxidase activity)
Taxonomic differences in toxicant uptake and
response
Physiological basis for differential sensitivity to
toxicants - degradation processes
-------
Terrestrial Endpoints
Natural Vegetation
(Wildlife Habitat Quality)
Community Composition
Succession
Community Structure
Cover for Wildlife
Food for Wildlife
Endangered Species
-------
Select Test Species
Lists of
Regional
Plants for
Testing
Screening
Evaluation
Phytotoxicity
Variability
Seed Availability
•.•i-i&MMaa
-------
GIS Approach
® Select test species of ecological and
economic importance in relation to
the anticipated release area.
-------
A system of maps including current
pesticide use patterns, anticipated
pesticide use patterns, climate (especially
r r j
wind speed), crops, natural vegetation,
endangered plant/animal species, water
resources, wildlife ranges
-------
O Row Crops (52.37%)
] Pasture, hay(23.13%)
I Deciduous forest (11.12%)
] Grasslands, herbaceous (4.01%)
¦ VVfeter (1.55%)
~ V\foody wetland (1.53%)
] Low intensity residential (1.19%)
| Emergent herbaceous wetlands (1.11%)
I Commercial, industrial, trans. (1.03%)
n Small grains (0.94%)
~ Urban, recreation grasses (0.58%)
Major Classes
Projection; ialfceis equal aiea, NAD83
SHOME/arc/5 aiptsjl rT/hi ip iflnr I 3/do_m rt t a m I J. S. Kern, 0 9 J an 0 2
-------
Kuchler Potential Natural Vegetation
Central Feed Grains and Livestock Land Resource Region
0 240
Projection: Albeis equil area. NAD83
SHOME/ircrtsaifrtsdrr/hi aps/lrrI ITdoJfuch.aml, J.S. K«rn, 0 9 Jan 02
Major Classes
] Bluestem prairie & oak-hickory forest (28.66%)
I Oak-hickory forest (24.88%)
D Bluestem prairie (24.4C%)
~ Beech-maple forest (8.99%)
~ Maple-basswood forest (4.24%)
~ Oak savanna (3.88%)
| Northeastern spruce-fir forest (1.91%)
~ Cross timbers (1.03%)
] Wheatgrass-bluestem-needlegrass (0.94%)
~ Southern floodplain forest (0.42%)
~ Nebraska Sandhills prairie (0.27%)
-------
I Oak-hickory forest (33.61%)
~ Bluestem prairie & oak-hickory forest (26.92%)
~ Bluestem prairie (12.57%)
~ Beech-maple forest (7.62%)
~ Maple-basswood forest (6.47%)
~ Oak savanna (5.54%)
] Northern floodplain forest (2.69%)
~ Cross timbers (2.54%)
~ Southern floodplain forest (0.60%)
I] Wheatgrass-bluestem-needlegrass (0.42%)
~ Nebraska Sandhills prairie (0.38%)
Major Classes
?HOME/ircteorifrtsJrr/m ips/lir I 2/-Jn
-------
Greenhouse Tests
Culture New
Species in
Greenhouse
Greenhouse
Tests
I
-------
Developmental Test
Both annual and perennial plants
Different agro-ecoregions
-------
Questions
Physiological mechanisms leading to
inhibition of reproduction
Inhibition of photosynthate transport?
Inhibition of membrane transport?
Ecological significance of selective
reproductive failure
Plant community alteration
Rare and endangered species
Food chain consequences - invertebrates and
vertebrates
-------
Plant Life Cycle Test
• Seed to seed
-------
Field Tests
Multi-species tests
Interactions
Ecological relevant end points
-------
-------
Open-top Chamber Modification
mMm
I9K
Root Barrier ~
Gravel
iSroxw^wlWInr? .<• i
Plastic Ring
Ground Level
Plastic Ring
Water Proof
Pond Liner
Pipe
Sffiffi'A' Perforations in
Sump Pump
Sand r.
mm
Greenhouse Bench Top
Cinder Block
Pipe
-------
Experimental
Literature
Risk
Methodology
Data Base Information
-------
Monitoring
Incident reports
Test effectiveness
Site location by GIS
-------
Scientific Understanding
Physiology
Ecology
Molecular
Biology
Practical Tools
Test
Development
GIS
Molecular
Markers
7mmasr
Improved Risk Assessments
Assistance to OPPTS
-------
The
end
-------
Herbicides and Nontarget Plant Loss
John Fletcher
U.S. EPA Laboratory, Corvallis, OR
-------
Herbicides and Nontarget Plant Loss
FACTS
• Herbicides are designed to kill plants
• Applied herbicides do not always land or stay on "target" plants
• Low level exposure of "nontarget" plants can reduce yields and
farm income
QUESTIONS
• Does the registration process prevent herbicides from going
off target? NO A Adverse weather
B. Contaminated spray equipment
C. etc.
• Does the registration process:
A. Ensure no nontarget plant losses? NO
B. Provide data necessary for farmers to reclaim financial
losses?
-------
Idaho Oust Episode Spring/Summer 2001
Land Characterization Classes
National Land Characterization Database (NLCD)
Idaho
J 0e«id«jous forest
J E vergr eer> forest
m Mxedforest
~ Shrub land
| Grasslands. herbaceous
^ Woody wetland
¦ E merger* hcibaeeous wert
| Pasture. hay
Row Crops
¦ 5 MaM grains
| | Fallow
~1 Uiban. recreation grass*
| Orchard, vineyards
| Low int ervirty r esi derrfia I
kilometers
z c=
0 240
Ptojirton: AlD«rs equal area. NAOI5
High irrt entity r esident i a I
J| Co mm ercial, industrial. tran^-oit mon
| B ar e r o« k. s anHOMDirc*criptvtaric/map&rmi Jt_s«i/il-2i imi. J.J K«n, 12 Jul 01
-------
¦Minidoka
F39.259.1I
Aberdeei
jBannocld
F 2.859 48j
Uerome,
^370_a
Estimated Acres/County Affected by OUST
Map Created August 27, 2001
Data Provided By: Idaho Dept of Water Resources, TIGER
Map Created By: Land View Systems
Idaho Land Ownership
E B.L.M
I Open water
¦ Private
OUST Spray Area
| Region
Affected Area x County
Region 615,356
Acres
K_|
B-TT"
MILES
~ 'Lincoln®
B36,563.27|
Hi
pingham|
234.449 35
-------
Time Schedule of Oust Episode
2000
Summer
(drought)
• Grass fire on
BLM
property
from natural
causes
2000
Fall
(drought)
• Oust
herbicide
applied to
17,000
acres of
burned
grasslands
2000
Fall
(drought)
• Wind erosion
• Dust storms
2001
Spring-Summer
(drought)
• Crop failures
-Sugar beets
-Grains
-Potatoes
-------
-------
-------
-------
-------
-------
MULTIPLE KNOBS
CRACKS &
ELEPHANT HIDE
-------
-------
Damage estimates by commodity for
Oust drift from BLM lands in Idaho for
the 2001 crop year
Crop
$ Amount
1. Sugar beets
16,000,000
2. Barley, wheat corn, alfalfa
2,000,000
3. Potatoes
60,000,000
TOTAL
78,000,000
-------
Farmer Proof Necessary to Secure Court Settlement
for Crop Loss from Nontarget Herbicide Drift
Needed Proof
1. Source of herbicide
2. Path followed by herbicide
from target to nontarget
3. Visible evidence of crop
damage due to drifted
herbicide (unique
symptom)
4. Presence and
concentration of herbicide
at nontarget site
5. Data showing crop yield
reduction at concentration
of herbicide at nontarget
site
Status of Idaho Farmers
1. Have spray dates and locations
2. Dust storm pictures
3. No unique leaf damage
4. 0.5 ppb detection (throughout
investigation), 0.1 ppb
detection (starting Fall 2001)
5. Response data of sugar beet
and potato at 0.1 ppb
exposures are not available
(Research in progress)
-------
Exposure-Response Curve
100
Response
(% yield
reduction) 50
0
Exposure concentration
Research Efforts
Industry herbicide development
-------
Exposure-Response Curve
Response
(% yield
reduction) 50
Exposure concentration
Research Efforts
Industry herbicide development
EPA, FDA, etc. registration process
Remediation/restoration companies
Academic ecological research
Estimate of total data
contributed
90%
10%
<1%
<1%
-------
Do surrogate species work?
Table 8 Comparison of EC50 Values at Four Taxonomic Levels
Taxon 1
Taxon 2
Species in a genus
Ipomoea lacunosa
Ipomoea lacunosa
Ipomoea quamoclit
Ipomoea quamoclit
Ipomoea purpurea
Ipomoea purpurea
Ipomoea lacunosa
Ipomoea quamoclit
Ipomoea purpurea
Ipomoea wrightii
Amaranthus retroflexus
Amaranthus retroflexus
Amaranthus retroflexus
Setaria viridis
Setaria viridis
Mean
Genera in a family
Digitaria
Festuca
Panicum
Echinochloa
Phalaris
Mean
Families in an order
Cyperaceae
Amaranthaceae
Caryophyllaceae
Rosaceae
Mean
Orders in a class
Asterales
Asterales
Liliales
Papaverates
Rosales
Rosales
Mean
Ipomoea quamoclit 4
Ipomoea wrightii 3
Ipomoea wrightii 3
Ipomoea purpurea 3
Ipomoea lacunosa 4
Ipomoea hederacea 5
Ipomoea hederacea 4
Ipomoea hederacea 3
Ipomoea wrightii 3
Ipomoea hederacea 3
Amaranthus hybridus 5
Amaranthus palmeri 5
Amaranthus spinosus 8
Setaria faberi 3
Setaria glauca 3
Echinochloa 3
Echinochloa 4
Setaria 4
Eleusine 3
Sorghum 3
Poaceae 35
Chenopodiaceae 11
Portulacaceae 7
Leguminoseae 10
Chenopodiales 24
Polemoniales 26
Poales 9
Polemoniales 13
Euphorbiales 16
Scrophulariales 9
0.984
0 988
0.998
0.998
0.919
0.997
0.931
0.996
0.994
0.997
0.971
0.484
0.841
0.752
0.236
0.868
0.645
0.146
0.481
0.886
0.841
0.559
0.183
0.132
0.065
0.013
0.134
0.024
0.001
0.003
0.361
0.126
0.060
0.081
• The number of chemicals tested on each pair of taxa.
6 The coefficient of determination log-transformed EC50 values.
From Fletcher, J.S., F.L. Johnson, and J.C. McFarlane. Environ. Tox-
icol. Chem., 7, 615-622, 1988 With permission.
-------
Are seedling tests adequate?
Figure 2 Plant life cycle.
-------
A New Approach to Registration Testing
The Regional Approach
Test-plants and tests by region
-------
Canada's Perspective on
Non-Target Plant Risk
Assessment
Pest Management Regulatory Agency
(PMRA)
-------
Risk Assessment
» I ^ I* i, * I r
* • j* *
Background
Current approach
Deficiencies
Preferred approach
-------
B ackground
Mandate
protect non-target plants and wildlife habitat
(not crops)
improve risk assessment
provide mitigative measures
-------
Background
Proposed Canadian Guidelines
developed by Environment Canada (1993)
no. of publications
recognised the inadequacy of current
requirements
recommended additional aquatic and terrestrial
species and testing
advancement in plant testing
never fully adopted by PMRA
-------
Canadian Data Requirements
- Pre-1998
# ¦" * * • ¦" •
£jC i ¦#>.'" ' f * ' ' f~> ™~
Generally equivalent to current EPA
Additional data were occasionally
submitted prior to harmonization
Data required for all pesticides
-------
Canadian Data Requirements
- Post -1998
* . - # * ,. * * • *¦
Harmonization with EPA in 1998
avoid additional data generation by registrants
interim until NAFTA guideline (U.S.- Canada)
Identical data requirements to EPA
three tiered structure
Data required for herbicides only
-------
Current Risk Assessment
Tier 1
single concentration testing of required species
Tier 2
multiple concentration definitive testing
Tier 3
field trials
-------
Current Risk Assessment
> » i y t % t y i % i ) ~ » i ^ ~
• ' » * * »
Deterministic assessment
risk quotient
Expected environmental concentration
(EEC)
Toxicity endpoint
most sensitive species
-------
Current Risk Assessment
- Risk Quotient (RQ)
* * . * *
Ratio of EEC to toxicity endpoint
EEC resulting from 100% application
Toxicity endpoint
NOEC
ec25
Level of risk if RQ exceeds 1
-------
Current Risk Assessment
-EEC, ;v A •
* * # • #
Aquatic
agriculture
30 cm deep pond
forestry
15 cm deep pond
Terrestrial
maximum seasonal application
-------
Current Risk Assessment
- Aquatic Toxicity Endpoint
# * • * *
Acute NOEC
Used as tests measure effects on population
(e.g., cell count)
In absence of NOEC, use an approximation
¦ - 1/10 ec50
-------
Current Risk Assessment
- Terrestrial Toxicity Endpoints
Acute EC25
Assumption that plants recover from EC25
-------
Current Risk Assessment
- Testing Endpoints
* * * *
Terrestrial plants
percent emergence
seedling height
seedling weight
visual symptoms
Aquatic plants
biomass
growth rate
-------
Current Risk Assessment
- Route of Exposure
Terrestrial plants
soil
foliar
Aquatic plants
through water column
-------
Current Risk Assessment
- Terrestrial Species
' * *
10 crops
Extrapolation to non-crops
Received some data on non-crop
e.g., shelterbelt species
-------
Current Risk Assessment
- Aquatic Species
Algae
green (Selenastrum
blue-green {Anabae
diatom (Navicula
diatom (Skeletonema
Floating vascular species
duckweed {Lemnasp.)
-------
Current Risk Assessment
- Summary
* "* 'b *ik
Screening assessment
risk quotient
EEC
maximum rate
Toxicity endpoints
NOEC
ec25
-------
Current Risk Assessment
- Summary
# • * * w
Standard testing endpoints
Most sensitive species
10 crop species
5 aquatic species
-------
Deficiencies
- Assessment Level
Screening assessment
Single toxicity endpoint
EEC represents a "worst-case"
No range of effects
No probability of effects
-------
Deficiencies
- Toxicity Endpoints
Use of EC25 for terrestrial species
recovery is assumed
Use of acute NOEC
Other endpoints are not considered
ECg, ECjQ
-------
Deficiencies
- Testing Endpoints
* • •' * * " »
Alternative endpoints are not considered
C02 fixation
02 evolution
Could act as early warning of toxicity
-------
Deficiencies
- Route of exposure
• # '* * ' +
No foliar exposure in vascular aquatic
plants
-------
Deficiencies
- Terrestrial Species
* '* * . » *
Crop species may be inappropriate for risk
to habitats
Conflicting data on sensitivity distribution
in crops and weeds
-------
Deficiencies
- Terrestrial Species
Data for only a limited number of plant
families
6
Choosing most sensitive crop species could
possibly overestimate the risk
Mitigation of habitats based on crop effects
-------
Deficiencies
- Aquatic Species
* *j- * * ' •
Rooted plants not represented
wetland species
provide food and shelter for waterfowl, fish
regulate flow of water
major contributors to primary productivity and
oxygen production
-------
Deficiencies
- Aquatic Species
Uncertainty in extrapolation from algae and
duckweed to rooted plants
Limited data on sensitivity distribution of
rooted plants
Choosing most sensitive species could
possibly overestimate the risk
-------
Preferred Approach
- Assessment Level
Need levels beyond screening assessment
More realistic exposure scenarios
fate models (PRZM/EXAMS)
Utilize dose-response curve
Incorporate probabilistic risk assessment
-------
Preferred Approach
- Toxicity Endpoints
Alternative to NOEC
e.g., EC^, EC10
Examine the use of EC25
Is it appropriate for all pesticides?
Is it appropriate for all species?
-------
Preferred Approach
- Testing Endpoints
• » # '*
Reproduction
flowering, seed and fruit formation
effects of ALS-inhibiting herbicides
Alternative endpoints
C02 fixation, 02 evolution, etc
early indicators of toxicity
-------
Preferred Approach
- Route of Exposure
* * " .« * r
Foliar exposure for aquatic species
spray drift
Lemnasp.
Emergent vascular
i
-------
Preferred Approach
- Terrestrial
'* # » * *
Habitats to protect
shelterbelts, riparian areas, grasslands, etc.
Select relevant species
non-crops (weeds and non-weeds)
annuals, herbaceous and woody perennials
Not necessarily the most sensitive
-------
Preferred Approach
- Terrestrial
'# ^ * '» ¦" • *
26 species proposed in NAFTA guideline
17 plant families
Crops
Retained to provide continuity
Non-crops
-------
Preferred Approach
- Aquatic Species
'# * ' * " * »
Protect habitats
wetlands, ponds, streams, marine/estuarine
systems, etc.
Select relevant species
Not necessarily the most sensitive
-------
Preferred Approach
- Aquatic Species
* . ». * . * *
m Protect habitats
wetlands, ponds, streams, marine/estuarine
systems, etc.
Select relevant species
Not necessarily the most sensitive
-------
Preferred Approach
- Aquatic Species
Algae: Dinoflagellate
(Gonyaulax or
(.Pyrocystis lumula)
Red algae
(Champia parvula)
Golden-brown algae
(Phaeodact
-------
Summary
- Preferred Approach
'*¦ * * * *
Go beyond screening assessment
Introduce probabilistic risk methods
Select appropriate number of test species
Select appropriate test species
-------
Summary
- Preferred Approach
• * '* » *• '*• ^
Consider the route of exposure in testing
Select appropriate toxicity endpoints
Consider alternative testing endpoints
-------
NON-TARGET PLANTS: A UK
perspective
Paul Ashby
Environment Branch (Ecotoxicology)
Pesticides Safety Directorate
DEFRA UK (formerly MAFF)
(www.pesticides.gov.uk)
-------
The role of Pesticides Safety
Directorate
• UK Regulatory Authority for pesticides
• Evaluate data and assess risks to humans and
environment
• Make draft recommendations
• 2 stage Committee process (new actives and old review
compounds)
• Inter-departmental (IDS) and Advisory Committee (ACP)
• Final recommendation signed by Ministers
-------
UK: Current position
Aquatic and Terrestrial ecosystems assessed separately
(common to EU)
"No spray zones" used as a risk management tool to
protect aquatic plants but NOT terrestrial plants
For terrestrial plants advisory labelling for most
herbicidally active compounds (i.e. sulfonylureas)
"Take extreme care to avoid drift onto nearby plants"
Recently refused approval for a specific formulation of a
isoazolidinone class herbicide due to impacts on non
target terrestrial plants associated with post application
vapour movement
-------
Aquatic plants: UK position
routes of exposure as per EU, but with greater emphasis
on drainflow for mobile compounds
Lemnaconsidered to be a representative species
If TER >10 based on Lemna toxicity, then risk is
acceptable
breaches of TER trigger raise concerns, principally
relevance of recovery in Lemna
need for test protocols for wider range of species
-------
-------
Terrestrial Plants: UK position
Historically based on need to ensure high levels of weed
control
No quantitative risk assessment scheme for non-crop
plants (extensive adjacent/following crop safety data)
Concerns:
- Much publicised declines in arable bird populations
and some arable weed species
- DEFRA High Level Quality of Life Indicators (reverse
decline in farmland bird populations)
- Advent of GMHT crops raised profile of indirect effects
m
-------
Weeds and birds
Percentage declines in the UK Common Bird Census
farmland index, 1969-1994
Species % decline, 1969-1994
Tree sparrow
Grey partridge
Corn bunting
Turtle dove
Lapwing
Skylark
Linnet
89
82
80
77
62
58
52
-------
Weeds and herbivorous insects
Potential ecological effects of herbicides on invertebrates.
(after Breeze et al. 1999).
direct
herbicide
toxic
indirect
Arable weed
Invertebrate
community
food
resource
species composition
structure / architecture
phenology
nectar/pollen
foliage/flowers
Habitat
modification
predators
decomposers
IPSI
F
-------
Problem solution
• Evidence for significant impacts of herbicides on
flora on arable areas?
• Which species are commonly associated with
arable areas?
• Are lower levels of control economically viable?
• Assess ecological importance of common
species
• DEFRA Desk study PN0940
-------
Results from DEFRA Desk
Study PN0940
• "The Impact of herbicides on weed abundance and biodiversity"
• Floristic composition of arable areas is the result of
complex interactions (nutrient status, crop type, status of
seedbank, cultivation techniques, methods and timing of
weed control)
• Evidence of long term reductions in arable flora is
limited
• Short term impacts may have important consequences
for biodiversity
-------
Results cntd
Certain plant species are always likely to be targets for
control (Avena fatua, Alopecurus myosuroides, Elytrigia
repens and Galium aparine)
These species are of limited ecological value
Other species are important for invertebrates and birds
(Rumex obtusifolius, Stellaria med
album, Senecio vulgaris)
These species are generally of lower competitive ability
PSD
-------
Some selected weeds
Common name Latin name
Value for
seed-eating
birds
Value for
insects
Annual
Meadow-grass
Poa annua
Competitive
index
0.10
% fields
infested
Black-grass
Alopccurus
myosuroides
0.40
Common Field Veronica persica
Speedwell
0.08
Common
Chickweed
Stellaria media
0.20
Fat-hen
Chenopodium ***
album
0.20
Knotgrass
Polygonum
aviculare
0.10
No data
-------
Conclusions from PN0940
• Further research required
- competitive ability of wider range of "weed" species
- confirmation of trends in desk study (field work)
- assessment of biodiversity importance of other
common weed species not covered in study
- quantification of importance of weeds for
inverts/birds (i.e. food depletion modelling)
- development of weed management systems that
encourage biodiversity without incurring high
economic cost
-------
Overall Conclusions 1
Floristic composition of arable areas is the result of
complex interactions
Arable areas are not "ecological deserts", fields are not
"green factories"
Greater consideration of potential impacts of pesticides
on non-target plants is needed
More collaborative research required to establish if
'within crop' risk assessments are necessary and, if so,
how to manage.
-------
Overall Conclusions 2
• DEFINE OVERALL PROTECTION AIMS
• REALISTIC ABOUT THE EXTENT TO
WHICH THE PESTICIDE REGULATORY
PROCESS CAN HELP TO ACHIEVE
GOALS
-------
Activities related to non-target
plant risk assessment in Denmark
Niels Elmegaard
National Environmental Research
Institute
-------
Back ground 1
• test with terrestrial plants not a part of
pesticide legislation
• more than 60% of the land is cultivated
• in year 2000 every ha of agricultural land
was on average subjected to a spray
intensity corresponding to two treatments
with full label rate
-------
Back ground 2
• most uncultivated terrestrial habitats are
situated next to a crop field
• most lakes and streams are bordered with
fields to some extent
-------
Research projects involving non-
target-plants and pesticides 1
The reproduction of hawthorn as
indicator of changing biodiversity
following pesticide drift in arable land
hedgerows
-------
Research projects 2
Effects of herbicides and ammonia on
hedgerows
Effect of herbicide drift on uncultivated
habitats neighbouring sprayed fields
Development of a field-test for plants
exposed to pesticides
-------
Research needs 1
• Effect of herbicide drift on terrestrial plant
communities in uncultivated habitats
• Effect of herbicide on freshwater plant
communities
• Effect of atmospheric deposition of
herbicides on plant populations and
communities
-------
Research needs 2
• The research should clarify to what extent,
if at all, there is a significant environmental
impact
• risk should be addressed by risk assessment
• mitigation by use of restrictions (deposition)
or buffer zones (drift)
-------
Niels Elmegaard 08.01.02. EPA workshop on non-target plants.
Research projects at NERI, Denmark, related to non-target plant risk
assessment for pesticides.
(Projects dealing with faith like atmospheric, water or soil transport of pesticides are not
included)
Effect of herbicide drift on uncultivated habitats neighbouring sprayed fields
PM: Niels Elmegaard , . _nn„
° Start of project: 2002-01-01
Dept: Terrestrial Ecology End of project: 2002-12-31
Data on spray drift, species sensitivity distribution, location of uncultivated habitats and crop of
neighbouring fields, pesticide use in different crops, spray timing, wind directions, are linked
with in a GIS-frame. The project aims at identifying spots with the highest risk of herbicide
impact. Further studies will have to reveal if any unacceptable effects on plant communities in
uncultivated habitats are occurring and if so what width of unsprayed border zones is needed
to protect the plant community.
The reproduction of hawthorn as indicator of changing biodiversity following pesticide drift
in arable land hedgerows
PM: Christian Kjaer
' Start of project: 2001 -01-01
Dept: Terrestrial Ecology End of project: 2003-12-31
Programme Area: Risk Analysis Chemicals/Biotech
Project Area: Terrestrial and Aquatic Ecoloxicology
This pilot project measures variables related to growth and reproduction of hawthorn Crataegus
monogyna. The purpose is to assess variation in these variables and to calculate the sample size
needed to analyse for sensitivity to pesticide spray drift in a following project. In this context
within and between site variation are to be considered. In 2002 experiments with replication
dimensioned according to power analysis of the results of the pilot study are carried out.
In addition to sampling of hawthorn a preliminary test of effects of herbicides on hawthorn will
be carried out at one occasion (flowering) in one hedgerow. Sampling of insects will take place
at least in one hedgerow regularly throughout the season. Consequences to bird life are
considered.
-------
Effects of herbicides and ammonia on hedgerows
PM: KrmdTybirk Start of project: 2001-01-01
Dept: Landscape Ecology End of project: 2002-12-31
Objectives: The project aims to document and quantify the separate and combined effects of
herbicides and ammonia on vegetation and arthropods in uncultivated biotopes near fields
simulating drift of herbicide and ammonia. Methods: Through a controlled experiment on
standardised soil with sown vegetation (simulating grassland) the project will analyse the
separate and combined effects of herbicide (Glyphosat) and nitrogen. The study will focus on
the flora, vegetation and selected arthropod groups important as food items for birds. The
experimental plots will receive applications of 0,1, 5, and 25% of Glyphosat field dose and 0. 25
and 100 kg N/ha/yr of ammonium nitrate in randomised block design with 6 replicates. In
addition, 20 hedgerows (10-15 years old) will be selected (10 on organic farms, 10 on
conventional farms) and the vegetation and arthropods of the hedgerows will be collected using
similar methodologies. Data will be analysed using traditional analysis of variance and
multivariate statistics.
Development of a field-test for plants exposed to pesticides
PM: Helle Weber Ravn .
Start of project: 2000-01-01
Dept: Terrestrial Ecology End of project: 2002-12-31
Programme Area: Risk Analysis Chemicals/Biotech
Project Area: Terrestrial and Aquatic Ecotoxicology
The effect of spraying-free zones as a mean to protect water-streams and lakes is proved using
experimental investigations in agricultural fields according to Action Plan for Pesticides II. The
investigations include evaluation of spray drift to water-streams and its impact on the aquatic
environment. The spray drift is measured as accumulated biochemical responses in terrestrial
and aquatic plants sampled along a transect from the field through the unsprayed buffer zone
into the stream. In the water-streams effects on the fauna also is assessed. Streams with no
buffer zones are used as reference.
-------
Non-Target Plant Risk Assessment Workshop
Assessing the Effects of
Plant Protection Products on
Non-target Terrestrial Higher Plants
The German Approach
(BBA, UBA)
BBA Federal Biological Research Centre for Agriculture and Forestry
UBA Federal Environmental Agency
Federal Biological Research Centre for Agriculture and Forestry (BBA)
Biology Division
-------
Content
• Background, documents
• Protection Aim, definition
• Effect Assessment
• Exposure Assessment
• Risk Assessment and Evaluation
• Experience so far
Federal Biological Research Centre for Agriculture and Forestry (BBA)
$ Biology Division
-------
Background
EU Directive 91/414/EEC. Annexes II. Ill and VI
„ ... Summary of available data ... used to assess
biological activity and dose range finding ... which
may provide information with respect to ... non-target
species, both flora and fauna ... "
member states shall ensure that use ... does not
have any long-term repercussions for the abundance
and diversity of non-target species."
^ ^ No details, no procedure for risk assessment
? Federal Biological Research Centre for Agriculture and Forestry (BBA)
Biology Division
-------
Documents, activities
OECD: OECD Guideline for the Testing of Chemicals,
Proposal for updating Guideline 208,
Terrestrial (non-target) Plant Test:
208 A: Seedling Emergence and Seedling
Growth Test
208 B: Vegetative Vigour Test
EPPO: Environmental Risk Assessment Scheme
for Plant Protection Products (Chapter 12,
Terrestrial Non-target Higher Plants, draft)
jp Federal Biological Research Centre for Agriculture and Forestry (BBA)
y Biology Division
-------
Protection Aim, definition
Protection aim: No unacceptable risk to
populations of non-target
plants
Definition: Non-target Terrestrial Plants:
„non-crop plants outside the
treatment area" (EPPO)
Federal Biological Research Centre for Agriculture and Forestry (BBA)
Biology Division
-------
What needs testing?
Negligible exposure -> No testing
e.g. wound treatment products, rodenticides,
granular products, seed dressings, single plant
treatment, indoor uses
Exposure -» Testing required
Federal Biological Research Centre for Agriculture and Forestry (BBA)
J Biology Division
-------
Assessment - Step 1
Evaluation of the phvtotoxic potential
(mainly for non-herbicides, certain standards apply)
- available data (screening data, data on crop
tolerance from efficacy testing)
- limit-test according to OECD (vegetative vigour,
application on top); 6 species from 4 to 6 families;
always recommended Avena sp., Brassica sp.,
leguminose; highest single application rate
Federal Biological Research Centre for Agriculture and Forestry (BBA)
Biology Division
-------
Assessment - Step 1
Assessment of phytotoxic symptoms or
fresh/dry weight of the above-soil biomass
Effects > 50 % in one species
or
Effects > 25 % in two or more species
step 2
Federal Biological Research Centre for Agriculture and Forestry (BBA)
> Biology Division
-------
Assessment - Step 2
Pesticides with a phvtotoxic potential
(always relevant for herbicides and growth regulators)
- Dose-response test according to OECD 208
- On top application, formulated product
- Type of testing (seedling emergence/vegetative
vigour) chosen according to the route of uptake
- 6 species, 4 to 6 families
- Always recommended: Avena sp., Brassica sp.,
one species of Leguminosae
- Select other species accord, to the efficacy spectrum
Federal Biological Research Centre for Agriculture and Forestry (BBA)
Biology Division
-------
Assessment - Step 2
- Fresh/dry weight of the above-soil biomass, plant
height, any phytotoxic symptoms
- Additional for the seedling emergence test:
number of seedlings emerged
^ EDS0 of the most sensitive species based
on above-soil biomass and plant height
(for seedling emergence additionally
number of seedlings)
I (probabilistic risk assessment if appropriate)
J Federal Biological Research Centre for Agriculture and Forestry (BBA)
^ Biology Division
-------
Assessment - Step 3
Experience with:
- Increase of species number
(species sensitivity distribution)
- Field test (small plot tests with sensitive
species from the laboratory)
Federal Biological Research Centre for Agriculture and Forestry (BBA)
Biology Division
-------
Assessment - Step 3
- Fresh/dry weight of the above-soil biomass,
plant height, any phytotoxic symptoms
- Additional for the seedling emergence test:
number of seedlings emerged
ED50 of the most sensitive species based
on above-soil biomass and plant height
(for seedling emergence additionally number
of seedlings)
(probabilistic risk assessment if appropriate)
Federal Biological Research Centre for Agriculture and Forestry (BBA)
p Biology Division
-------
Exposure Assessment
The off-crop scenario is considered
State of the art:
Spray drift events: yes
Gaseous transport events: to consider (under
development)
Run-off events: is not considered
as main route of
exposure
Federal Biological Research Centre for Agriculture and Forestry (BBA)
Biology Division
-------
Exposure Assessment - Drift
Drift events: German drift data (BBA, 2000)
Interception:
- for uptake via the roots currently 50 % interception
by the vegetation of the adjacent biotopes
- if a buffer zone to an adjacent biotope is required,
the crop itself may reduce the amount of drift;
up to now no reliable data on the amount of
reduction are available
jp Federal Biological Research Centre for Agriculture and Forestry (BBA)
J Biology Division
-------
Toxicity Exposure Ratio (TER)
TER = ED50/PEC of the most sensitive species
step 2: uncertainty factor of 10 (may be reduced
depending on the number of species
tested)
TER < 10 -» step 3
step 3: uncertainty factor reduced
Federal Biological Research Centre for Agriculture and Forestry (BBA)
Biology Division
-------
Risk Mitigation
A risk mitigation system is under development and
partly implemented
Principles:
- prescribe drift reducing technique in the 20 m
area from the field margin
- no-spray zone of 5 m to certain adjacent biotopes
(width > 3 m)
- if a region has a certain amount of off-crop biotopes
(under consideration: 5-20 %) exceptions are planned
Federal Biological Research Centre for Agriculture and Forestry (BBA)
J Biology Division
-------
Risk Mitigation - Example
On a 20 m strip to adjacent areas
(except agriculturally or horticulturally used areas,
roads, etc.)
the product must be applied using drift reducing
technique which is listed in index ...
under at least drift reducing class ... (e.g. 50 %).
(In addition an untreated buffer zone of at least 5 m is
to be kept from adjacent areas (except, see above ...))
... (next sheet)
Federal Biological Research Centre for Agriculture and Forestry (BBA)
Biology Division
-------
Risk Mitigation - Example
Neither drift reducing technique nor a buffer zone of at
least 5 m is required if
- the product is applied with portable spraying
equipment, or
- the adjacent areas (field boundaries, etc.) are less
than 3 m wide, or
- the product is used in a region where (as listed in
index ...) a sufficient amount of off-crop biotopes is
I available.
y Federal Biological Research Centre for Agriculture and Forestry (BBA)
J Biology Division
-------
Experience
So far:
Non-herbicides: nearly no effects (screening data/limit-
tests), step 2 seldom required, no risk
mitigation according to step 2 needed
Herbicides: always definitive tests (step 2) required,
always risk mitigation according to step 2
necessary, if step 3 data submitted, risk
mitigation category may change
Federal Biological Research Centre for Agriculture and Forestry (BBA)
Biology Division
-------
Open questions/Outlook
Is the protection aim reached?
Risk refinement options (step 3) for terrestrial
plants needed (guidelines, risk assessment)
Effects on crop species versus wild species
Effects on annual species versus perennial species
Effects on single species versus communities
¦4 Federal Biological Research Centre for Agriculture and Forestry (BBA)
J Biology Division
-------
Terrestrial non-target plants
View from the EPPO
Robert Luttik
Centre for Substances and Risk assessment
National Institute for Public Health and the Environment
Bilthoven, The Netherlands
frWffffi)
-------
What is the EPPO
i> EPPO = European and Mediterranean Plant Protection Organization
(not only the EU but also eastern Europe and north Africa)
• In 1989 the EPPO and the CoE established a Joint Panel with the
task to develop guidelines for environmental risk assessment of
plant protection products.
• In 1993 to 1995 schemes were published on nine different areas of
environmental concern: soil, ground water, surface water, aquatic
organisms, soil microflora, earthworms, arthropod natural enemies,
honeybees, and terrestrial vertebrates.
• In 1999 a revision of the nine schemes was started and in addition
two new schemes should be developed: air and higher non-target
plants.
iHiwmi
-------
How is the EPPO organized?
Higher plants
Vertebrates
Earthworms
I
Soil
Air
Water
Environ-
mental
Panel
Council
of EPPO
Min. of Agriculture country 1
Min. of Agriculture country 2
Min. of Agriculture etc
Working
party
fFWlffiil
-------
Aims of the EPPO schemes
• Guide assessors on the questions that should be addressed, and the
data that may need to be requested from registrants.
• Provide information on the test methods and approaches that are
suitable in each case.
• Indicate how the data should be interpreted in a consistent manner,
involving expert judgement where appropriate.
• Produce a reliable assessment of environmental risk, that is suitable
to aid risk management, although it will not provide all the information
necessary for decisions about the acceptability of plant protection
products.
-------
EPPO schemes - general aspects
• Exposure assessment (emission, fate models)
• Effect assessment (toxicity data, extrapolation factors (methods))
• Quotient driven (exposure over toxicity ratios)
• Risk characterisation
• Analysis of uncertainty
• Risk management
• Focussed on first tier
• There are 4 categories of risk (high, medium, low and negligible)
-------
Vision of vertebrate group on high and low risk
The line of yes or no
Realistic
worst case
Most
likely case
C
0
m
P
1
e
x
i
t
y
field i
probabilistic
/ special
/
studies
modelling
studies \
Area of low risk
Area of
straight
concern
forward
. Area of high risk
**•...
Risk quotient
ffWlfifil
-------
Non-target higher plant scheme (1)
1. Collect data
Go to 2
2. Can the possibility that higher plants will be exposed to the active
substance, directly or indirectly, be ruled out (see Note 1)?
If yes Classify as negligible risk, go to 17
If no Go to 3
3. To which group does the active ingredient belong (see Note 2)?
Herbicide or growth regulator Go to 7
Other Go to 4
4. Obtain screening data (in addition efficacy data can also be used)
from at least 6 plant species from 6 different families including
both mono- and dicotyledons (see Note 3). Go to 5
iHivififfi
-------
Non-target higher plant scheme (2)
5. Are the results in one or more of the screening or efficacy tests with
vascular plant species showing > 50% phytotoxic effects (see Note 4)
at the maximum recommended application rate (MRR) or higher (e.g.
growth, chlorosis; Note 5)?
If no Classify as low risk, goto 16
If yes Go to 6
6. Obtain concentration/response tests on the affected species (Note 6)
If MRR/EC50 is < 1 for all affected species
Classify as low risk, go to 16
If MRR/EC50 is > 1 for one or more affected species Go to 7
-------
Non-target higher plant scheme (3)
7. Obtain concentration/response tests on at least 6 species
representing families for which significant herbicidal action has
been found (Note 7) and determine the EC50 value for each of
the tests.
Go to 8
N
8. Using the EC50 values for 6 or more different species, generate the
5th percentile of the log-logistic or the log-normal distribution
according to Aldenberg and Slob (1993) or Wagner and Lokke (1991)
or the final acute value (triangular distribution) according to Stephan
et al (1985) depending on the most likely shape of the toxicity
distribution curve (see Note 8) and use this calculated toxicity value
in the risk assessment.
Go to 9
fFWffiR)
-------
Extrapolation methods
5th percentile = 10(AVG *E *STD) /
in which:
AVG = the mean of the log10 transformed EC50 values
STD = the standard deviation of the log10 transformed EC50 values
E = Extrapolation factor dependent on sample size
Although this procedure formally works for n = 2, 3 in practice the
method is applied for n = 4, 5 only.
For higher plants n = 6 proposed.
iFWffffil
-------
Non-target higher plant scheme (4) Drift events
9. Obtain from the air chapter 12 information about the drift concentration
of the active ingredient (a.i.) that can be expected at distances of 1 and 5
metres from the treated area or 3 and 7 meters for orchards (see Note 9).
Go to 10
10. Calculate the exposure over toxicity ratio.
If PEC(1 m distance)/calculated toxicity value < 1
Classify as low risk for drift events, go to 11
If PEC(5 m distance)/calculated toxicity value < 1
Classify as medium risk for drift events, go to 11
If PEC(5 m distance)/calculated toxicity value > 1
Classify as high risk for drift events, go to 11
-------
Non-target higher plant scheme (5) Gaseous transport
11. Obtain from air chapter 12 information about the concentration of the
a.i. that can be expected at a certain location.
If PEC(gaseous transport) < PEC(drift at 1 m distance) Go to 12
If PEC(gaseous transport) > PEC(drift at 1 m distance) Go to 13
12. No further risk assessment for gaseous transport events is necessary,
because the outcome of this risk assessment will always be lower than
that one based on drift events. Go to 14
13. Calculate the exposure over toxicity ratio (see note 10).
If PEC(gaseous transport)/calculated toxicity value is < 1
Classify as low risk for gaseous transport, go to 14
If PEC(gaseous transport)/calculated toxicity value is >1 and <10
Classify as medium risk for gaseous transport, go to 14
If PEC(gaseous transport)/calculated toxicity value is >10
Classify as high risk for gaseous transport, go to 14
-------
Non-target higher plant scheme (6) Run-off events
14. Is the uptake of the active substance mainly via the roots (in general soil
applied compounds/formulations)
If yes Go to 15
If no Classify as low risk for run-off events, go to 16
15. Obtain from soil chapter xx information about the concentration of the
active ingredient that can be expected at the scenario location (note 10).
If PEC(run-off scenario)/calculated toxicity value is < 1
Classify as low risk for run-off events, go to 16
If PEC(run-off scenario)/calculated toxicity value is >1 and <10
Classify as medium risk for run-off events, go to 16
If PEC(run-off scenario)/calculated toxicity value is >10
Classify as high risk for run-off events, go to 16
frimm
-------
Non-target higher plant scheme (7)
Analysis of uncertainty
16. Review the data that led to high, medium or low risk category and
check whether the conclusion are correct
if yes, confirm assessment
Go to 17
if no, obtain more information as needed
Go to 2
17. Risk management
— frWfffifl
-------
Some thoughts about the process (1)
• Is it difficult?
Yes processes dealing with consensus are very often difficult.
Major discussion points were:
- Where do we carry out the risk assessment?
- Inside/outside the crop
- At which distance (meters)
- Do we use fixed extrapolation factors or more statistical methods?
- How many toxicity tests do we need/ask?
- What can we use from the screening procedures (GLP or no GLP)?
- Do we have to test wild species or are crop species enough?
- Do we include other exposure routes like gaseous transport and run-off
events?
fHivffifii —
-------
Some thoughts about the process (2)
• Was the development of this particular scheme difficult?
Yes because most of the members of the group did not have much
experience with hazard/risk assessment for non-target higher plants.
no standard scenarios available
only one other scheme could be used as a guide
• Did it make fun?
-------
INIA's PROPOSAL FOR NON-TARGET [PLANT] RISK ASSESSMENT
By Jose Luis Alonso Prados
In 1997, the Spanish Ministry of Agriculture assigned INIA (Instituto Nacional de I
nvestigaci6n Agricola) as the accredited entity responsible for evaluating monographs for active
ingredients. These monographs are to be included in Appendix I of Directive 91/414/CEE. In this
talk, I am going to explain our proposal for evaluating the risk of crop protection products to
non-target plants. This proposal was prepared de by INIA's researchers from the plant protection
and environmental departments.
Slide 2: Directive 91/414/CEE- Environmental Risk Assessment
Directive 91/414/CEE, and concretely the Uniform Principles in Annex VI, establishes the
[data] requirements and methodology to perform environmental risk assessments. It specifies the
following groups of organisms:
• Terrestrial vertebrates (mammals and birds)
• Aquatic organisms (fish, arthropods, algae, and aquatic plants)
• Bees and beneficial arthropods)
• Soil macro- and microorganisms
However, there are no requirements for non-target plants nor a established methodology
for performing risk assessments for them. There is only one paragraph in the Directive that says "
Effects on other organisms, flora and fauna expected to be at risk".
In evaluating the risks of crop protection chemicals to non-target plants, one must realize
that, by their own phytotoxic nature, herbicides ought to be considered as the worst case.
Slide 3: Which plants are non-target plants? And which type of exposure?
To define the problem, one must start considering the AGRONOMIC USE of crop
protection products, particularly of herbicides.
• The group formed by those plants growing outside the treated area which,
at the time of application, will be exposed to spray drift of the product.
• The group formed by those plants that will be exposed to soil residues once
the
crop has been harvested. Within this group, one can distinguish two
subgroups:
-------
~ The next rotated crop
~ Crops and wild species that could grow in the treated area.
From these groups, one can define three different scenarios:
° Spray drift scenario
o Scenario for the next rotational crop
° Long-term scenario, which intends to cover the risk posed by persistent
chemicals
Once the problem and the scenarios are defined, one moves next to compare exposure
and toxicity as is done for any risk assessment. To do that, we propose the following.
Slide 4: Spray drift scenario
For the spray drift scenario, our proposal to evaluate exposure considers Ganzelmaier's
accepted spray drift values use to evaluate drift to surface water at a community level and which
establish a 4% spray drift and a buffer zone of 1 m. These values have been refined and the
current values will be used. For chemicals that are absorbed through the root system, we propose
calculating the expected concentration in soil by considering a homogeneous distribution in the
top 5 cm layer and a soil density of 1.5 km/dm^.
To evaluate toxicity, we propose using a probabilistic approach from phytotoxicity data for
12 different species and establishing the Maximum Acceptable Concentration as that will protect
95% of the distribution using the EC50 as the endpoint.
The proposed safety factor is 1.
Slide 5: Long term scenario
With this scenario, we intend to cover those chemicals that are persistent and have the
tendency to accumulate.
To evaluate exposure, we propose calculating the maximum expected concentration after
3 years of continuous use for annual crops and 10 for perennial crops.
To evaluate effects, we propose using a probabilistic methodology using phytotoxicity data
for 12 spicies and establishing the Maximum Acceptable Concentration with the 95th percentile
and the EC10 or NOEC as the endpoit.
Slide 6:
Scenario for next rotational crop
-------
To evaluate the exposure for this case, we propose calculating the expected residue left in
the soil at the time of planting the next crop (90-150-180 days). To do this, we will utilize an
accepted generic scenario in which the kinetics of degradation in soil is first order in a soil of 1.5
g/cm3 and a homogenous distribution at the top 20 cm of a plot prepared for cropping.
To evaluate effects, we propose a deterministic methodology u?ing phytotoxicitv data for
the most sensitive crop and the EC10 or NOEC as endpoint.
Slide 7: Species to be tested
One of the main topics of discussion is the number of species that should be looked at to
evaluate risk to non-target plants. We have proposed 12 species, taking into account that they
should be easy to handle in the laboratory, that germination is homogeneous, and that they are
representative of the main families, but the with selection leaning more towards cultivate than
non-cultivated plants.
Slide 8: Methodology proposal (Species to be tested)
For selecting the species to be tested to evaluate herbicide selectivity, we consider their
importance as crops and recommend that they be representative of different families. It is also
recommended that an equilibrium between monocotyledons and dicotyledons be maintained. We
propose conductive tests on:
• Wheat, barley, ryegrass, and oat (Proaceae family)
• "Veza" (Leguminous family)
• Tomato (Solanaceous family)
• Melon (Cucurbits
• Sugar beet (Chenopodiaceas familiy)
• Sunflower (composite family)
We have also included two non-crop species that are easy to handle. One is a
monocotyledon (Bromus) and the other a dicotyledon for the cruciferous family (Lepidium)
Slide 9: Selective Herbicides
There are some herbicides that are selective. Aloxidim, cycloxidim, and sethoxydim are
examples of herbicides that are selective towards monocotyledons. Examples of herbicides
selective towards dicotyledons are 2,4-D, MCPA and triasulfuron. Selectivity is an important point
to consider when evaluating risk, as additional data may be required for monocotyledons or
dicotyledons to selectivity.
Slide 10: Methodology proposal for the spray drift scenario
The idea is to use the data from laboratory studies to perform risk assessments for the
three scenarios. To do that, phytotoxicity data are needed for foliar and root exposure.
For the spray drift case, exposure would be foliar. A test at a single dose 2X the proposed
-------
dose will be used to determine phytotoxicity. If this test shows evidence of phytoxicity, then a
higher tier test will be performed . This higher tier consists of generating dose-response curves to
determine the EC50 followed by a probabilistic approach to evaluate exposure from spray drift
using accepted spray drift values.
Slide 11: Methodology proposal for the next rotated crop and long term scenarios
Phytotoxicity data from studies involving root exposure will be used to evaluate these two
scenarios. A single dose at 5X the PEC and 50X the PEC will be performed first. To calculate "PEC
", one must take into account the use of the product and the rotated crop that could be
considered as the worst case. The test will be conducted in 2 types of soil. If there are no signs of
phytotoxicity, then it can be said that there is no risk. But, if there is evidence of phytotoxicity,
then dose-response curves will be required to determine EC10 or NOEC to evaluate the risk. If
there is still risk, we will require field tests to determine the phytotoxicity of the actual residue
[residue concentration] left in the field after normal application. The data would serve to establish
rotational crop intervals for planting the next sensitive crop or for future uses of the treated area.
Slide 12: Examples of spray drift scenarios
Let's consider several examples. As I mentioned earlier, INIA is the entity that evaluates
the monographs for [active ingredients]. From the data contained in the monographs, we have
generated a database of phytotoxicity data for tested biological parameters (germination, %
emergence, % survival, plant height, dry weight of the plant, and root length) and the NOEC,
EC10, EC20, EC30, EC25, and EC50 endpoints. We have a total of 38 chemicals and 594 data. Of
the data, 349 come from soil exposure, 160 from foliar exposure, 109 hydroponic and 36 for
which the type of exposure was not indicated.
Slide 13 Examples of spray drift scenario
Using the database, we have performed a goodness-of-fit analysis of normal or log normal
distributions for a total of 9 herbicides for which enough data were available.
For root exposure, the number of herbicides we looked at was 8. We performed 21
statistical analyses using the NOEC or EC50 as the endpoint and 9 biological parameters. The
mean number of data for each [statistical] analysis was 10. Of the 21 analyses 10 fell within a
log/normal distribution, but 11 did not.
For foliar exposure, the number of herbicides we looked at was 6. We performed 13
statistical analyses using the NOEC or EC50 as the endpoint and 7 biological parameters. The
mean number of data for each [statistical] analysis was 10. Of the 21 analyses 4 fell within a
log/normal distribution, but 9 did not.
For hydroponic exposure the number of herbicides we looked at was 2. We performed 4
statistical analyses using the EC10 and the EC50 as the endpoints and only one biological
parameter. The mean number of data for each [statistical] analysis was 18. The 4 fell within a
log/normal distribution.
Slide 14:
Examptes of spray drift scenario- Foliar exposure
-------
In this slide, I am showing an actual case. The blue curve represents the log of the
concentration expected at the edge of the field of the treated crop, as a function of distance [from
the site of treatment]. The accepted, new spray drift values we used. The straight line and the red
dots represent the cumulative relative frequency of effects using the dry weight of the stem as the
biological parameter and the EC50 as the endpoint. The species analyzed were onion, corn,
wheat, sorghum, beet, soybean, pea, tomato, rape, and cucumber.
Considering the 95^ percentile, which corresponds to a 5% acceptable risk, the no risk
distance is 40 m.
Slide 15: Examples for successive crops
We have carried in our laboratory several bioassays with herbicide A, for uses on wheat at
an application rate of 20 g a.i./ha. Herbicide A has a half-life in soil of 54 days. In Spain,
sunflower is a typical rotational crop after wheat. Sunflower is very sensitive to residues of
herbicide A.
We obtained in our laboratory dose-response curves for sunflower and herbicideA. The
EC10 ranged from 0.38 to 0.67 parts per billion, ppb (yg/kg of soil). Considering a first-order
dgradation kinetics, the expected residue in soil at the time of planting would be 0.096 ppb. By
performing a deterministic risk assessment, we expect a risk to sunflower since in many cases the
EC50 is less than 5 times the PEC.
Bioassays performed in the laboratory using soil collected from treated field, have shown
that the actual residue left after a normal application do not pose a risk for sunflower, in spite
that there were significant differences between soil treated at a double dose and control soil.
Slide 16: Examples for successive crops
Triasulfuron is a second example. Triasulfuron is applied to cereals at an application rate
of 7.5 g a.i./ha, The half-life of this herbicide in soil is 38 days. In Spain, sunflower is the typical
crop that follows wheat.
We have obtained in our laboratory triasulfuron dose-response curves for sunflower in two
soils. The NOEC or EC50 would fall within 0.60 to 0.78 ppb (pg/kg of soil). Comparing this range
of concentration with the expected concentration in soil at the time of planting, we found that 330
days after application of triasulfuron, it is larger than the safety factor of 5 and concluding that
there is no risk for sunflower.
Slide 17: Examples for successive crops
The last example involves Herbicide B for uses on [sugar] beets at 200 g a.i./ha. Herbicide
B has a half-life in soil of 34 days. It is a herbicide used to control grasses in beets. From the
curves of dose- response of wheat to this herbicide, the NOEC or EC50 was found to be 2.89 ppb (
pg/kg of soil). When compared to the expected residue in soil, one can expect that there is a risk
for wheat planted after treated beets.
-------
Environment Australia's Experiences in Assessing Risk to
Non-Target Plants
OVERVIEW
Australia is a small market with limited resources.
Consequently, we do not have the capacity to
undertake significant work in:
• modelling
• testing
• refining methodologies
We tend to rely heavily on international
methodologies and data.
-------
ustralia's Experiences in Assessing Risk to
Non-Target Plants
Methodology
Australia seldom requests phytotoxicity data for chemicals other
than herbicides where these data do not accompany a submission.
Where these data are not available, other factors are considered to
address likely exposure. These include:
•Physico-chemical properties, particularly solubility and
adsorption characteristics; and
•data available from efficacy trials and field dissipation studies.
These data are used to determine whether non-target vegetation are
likely to be impacted on through exposure.
-------
Australia
Australia's Experiences in Assessing Risk to
Non-Target Plants
Methodology cont...
Where exposure is considered significant, phytotoxicity data
may be requested.
• •••••••
-------
ustralia's Experiences in Assessing Risk to
Non-Target Plants
Non-Target Terrestrial Plants
• Non target terrestrial plants are generally considered to only be
those in the natural environment.
• This may include native vegetation bordering fields, and
riparian vegetation along stream and river banks.
• There have been instances where severe non-target impacts
have been seen on ornamental plants (roses; home fruit trees).
• These have been considered in the assessment report, but only
as far as similar impacts may be expected on native vegetation.
-------
ustralia's Experiences in Assessing Risk to
Non-Target Plants
Non-Target Aquatic Plants
• Generally, aquatic plants are always considered to be non-
target.
• Recently, high profile examples of adverse impacts to aquatic
plants have been seen with mangroves and sea grass in the
Great Barrier Reef Marine Park.
• These impacts have been attributed in the press to runoff of
herbicides, specifically diuron, from the neighbouring sugar
cane farming areas.
-------
ustralia's Experiences in Assessing Risk to
Non-Target Plants
CASE STUDY 1 - Sulfonyl urea herbicide
Physical/Chemical Properties
Solubility: 15 ppm (pH5); 1657 (pH 7); 7470 (pH 9)
LogKoc: 28-178
Half life soil; 6-11 days
Half life water: <2 days
Ecotoxicitv
Algae: EC50 = 5.3 ppb
Aquatic plants: IC50 = 0.04 ppb
NOEC = 0.02 ppb
-------
ustralia's Experiences in Assessing Risk to
Non-Target Plants
CASE STUDY 1 - Cont...
Spray Drift Scenario:
• Direct overspray not relevant as aerial application is prohibited.
• Method of application by shielded sprayer or hand-gun/knapsack
sprayer - expected to result in negligible off-target drift.
• Main risk to aquatic plants expected to occur through run-off,
including being carried in irrigation water in the case of cottonH
-------
Ji'.U Environment
-CxF Australia
Australia's Experiences in Assessing Risk to
Non-Target Plants
CASE STUDY 1 - Cont...
Runoff Scenario:
ASSUMPTIONS (with an application rate of 50 g/ha):
• 1 hectare treated area available for runoff into a shallow water
body, 15 cm deep with a 1 ha surface area;
• Heavy rain even of 100 mL immediately after application with
10% of water running off;
• Total dilution water therefore is 2.6 ML accounting for original
water plus 100 mL rain plus runoff water.
• 4% of applied chemical runs off (based on literature report);
• The Estimated Environmental Concentration was 0.77 ppb
• The resulting Q value was 19.2 indicating an unacceptable risk.
-------
1
•
•
... Environment Australia's Experiences in Assessing Risk to
: A"Vtrolio J Non-Target Plants
I
1
CASE STUDY 1 - Cont...
Mitigation of Risk:
1
• Restriction on irrigation period following application;
¦
• More detailed analysis of field dissipation studies;
• Restriction on application coinciding with forecast of heavy rain;
and
l
• More detailed analysis of Lemna study.
• •••••••
I
-------
Australia
.ustralia's Experiences in Assessing Risk to
Non-Target Plants
CASE STUDY 2 - Pigment Inhibitor herbicide
Chemical Physical/Chemical Properties
Profile: Solubility: 1100 ppm (20-27°C)
Vapour Pressure: 0.02 Pa
Henry's Law Constant: 4.18X103 Pa nv'.mole"1
Half life soil: 28-56 days
Half life water: 16-22 days
Ecotoxicitv
Algae:
Terrestrial:
plants
EC50 = 3.7 ppm
Phytotoxic. Tier II tests provided.
-------
ustralia's Experiences in Assessing Risk to
Non-Target Plants
CASE STUDY 2 - Cont...
Spray Drift Scenario:
• This assessment was pre AgDrift and Ganzelmeier .
• Urban and Cook (1986) methodology initially used with 10%
spray drift, and applying an assessment factor of 100 to algae
study, showed aquatic plants to be at risk from exposure.
• Literature available showed 1% spray drift 5 m downwind was
realistic, and this essentially mitigated the hazard.
• Runoff was not considered in this assessment.
-------
ustralia's Experiences in Assessing Risk to
Non-Target Plants
CASE STUDY 2 - Cont...
Volatilisation Scenario:
• Impacts on non-target plants was evidenced from use during a
permit trial in Australia, and from many international reports.
• The proponent maintained the problem was due to spray drift, but
Environment Australia considered volatilisation the major cause
due certain factors:
there was no particular pattern to the off-target whitening;
effects were seen at large distances, and in areas unlikely to
result from drift, eg, over hills.
the problem often did not occur until several days after
application
-------
ustralia's Experiences in Assessing Risk to
Non-Target Plants
CASE STUDY 2 - Cont...
Mitigation measures:
• The use of assessment factors, while unusual, was done in this
case;
• Inclusion of many restrictive label statements to ensure limited
off target movement;
• Requirement for monitoring of non-target plant impacts during
further trialing under permit - results were reported;
• Intense product stewardship program.
Environment
Australia
-------
ustralia's Experiences in Assessing Risk to
Non-Target Plants
Do We Need Further Non-Target Species Tested?
• More data are always considered better than less.
• Increasing the number of species tested will lead to a better
determination of hazard to non-target plants.
• In doing so, it will reduce the level of uncertainty and if
assessment factors are to be used, lower the factor required.
• Increasing the number of tests may simply confirm already
known results.
• This may not help in the actual risk assessment.
-------
Appendix E
Non-target Plant Risk Assessment Workshop Agenda
-------
Non-target Plant Risk Assessment Workshop
January 15 - 17, 2002
8:00 AM
9:00 AM
9:15 AM
9:30 AM
9:50 AM
10:20 AM
10:30 AM
10:45 AM
11:00 AM
11:20 AM
Holiday Inn Select
Old Town Alexandria
480 King Street
Alexandria, VA 22314
703-549-6080
January 15
Registration
Brent Room, Holiday Inn Select
Old Town Alexandria, Virginia
Introduction and Welcome
Daniel Rieder, Facilitator
Elizabeth Leovey, Acting Director
Environmental Fate & Effects Division
Office of Pesticide Programs, U.S. EPA
Non-target Plant Risk Assessment
for Toxic Substances in U.S. EPA
Jerry Smrchek,
Office of Pollution Prevention and Toxics
U.S. EPA
Non-target Plant Risk Assessment
for Pesticides in U.S. EPA
Aquatic and Terrestrial Pesticide
Exposure Models
Break
Rick Petrie
Office of Pesticide Programs, U.S. EPA
Jim Carleton
Office of Pesticide Programs, U.S. EPA
Use of the Ecological Incident
Information System for Plants
Implementing Refined Risk
Assessments for Pesticides in
the U.S. EPA
Nick Mastrota
Office of Pesticide Programs, U.S. EPA
Ingrid Sunzenauer
Office of Pesticide Programs, U.S. EPA
A Pesticide Risk Manager's
Perspective
Non-target Plant Risk
Assessment: Issues and
Opinions
Phil Errico
Office of Pesticide Programs, U.S. EPA
Jane Staveley
Arcadis G&M, Inc.
-------
[ :40 AM
1:45 AM
00 PM
20 PM
40 PM
:00 PM
:20 PM
:40 PM
:50 PM
: 10 PM
:30 PM
:50 PM
:10 PM
2
Announcements
Lunch
Region 6 Topic
Use of Incident Data in
Assessing Risks to Non-target
Plants
State Perspective
Use of Chemical Benchmarks
for Decision-Making in the
Risk Assessment Process
Proposed Research in Support of
Non-target Plant Test
Requirements
Break
Plant Incident Case Study
Plant Research in
Gulf Breeze, Florida
Canada's Perspective on Non-
target Plant Risk Assessment
Non-target Plants: A U.K.
Perspective
Closing Remarks
Dan Rieder
Dick Watkins
U.S. EPA Region 6
Dallas, Texas
Karl Arne
U.S. EPA Region 10
Seattle, Washington
Terri Barry (presenter to be determined)
California Department of Pesticide
Regulation
Jim Fairchild
U.S. Geological Survey
Colombia, MO
Thomas Pfleeger
Office of Research and Development
U.S. EPA Corvallis, Oregon
John Fletcher
Office of Research and Development
U.S. EPA Norman, Oklahoma
Michael Lewis
Office of Research and Development
U.S. EPA Gulf Breeze, Florida
Ted Kuchniki, Derek Francois
Pest Management Regulatory Agency
Canada
Paul Ashby
Pesticides Safety Directorate
United Kingdom
Dan Rieder, Facilitator
-------
3
9:00 AM
9:10 AM
9:30 AM
9:50 AM
10:10 AM
10:30 AM
10:40 AM
11:00 AM
11:10 AM
11:30 AM
12:30 PM
January 16
Introduction Dan Rieder, Facilitator
Activities Related to Non-target Niels Elmegaard
Plant Risk Assessment in National Environmental Research Institute
Denmark Denmark
Assessing the Effects of Plant
Protection Products on Non-target
Terrestrial Higher Plants:
The German Approach
Terrestrial Non-target Plants:
View from the EPPO
INIA Proposal for Non-target
Plant Risk Assessment
(Low Dose Herbicides:
A Case Study)
Christine Kula
Federal Biological Research Centre
Germany
Robert Luttik
RIVM/CSR
The Netherlands
Jose Luis Alonso-Prados
INIA - Pesticide Group
Spain
Break
Australia's Experiences in
Assessing Risk to Non-target
Plants
Announcements
Poster Session
Lunch
Chris Lee-Steere
Environment Australia
Australia
Dan Rieder
Captain Piercy Room, Old Town
Holiday Inn Select
Break Out Group A (Topic 1) Jane Staveley, Facilitator
Risk Assessment Problem Formulation and Framework: See Attachment
Break Out Group B (Topic 1) Dan Rieder, Facilitator
Risk Assessment Problem Formulation and Framework: See Attachment
2:45 PM
Break
-------
4
3:00 PM Break Out Group (Topic 2) Jane Staveley, Facilitator
Exposure Assessment: See Attachment
Break Out Group (Topic 2) Dan Rieder, Facilitator
Exposure Assessment: See Attachment
4:50 PM Closing Remarks Dan Rieder, Facilitator
January 17
9:00 AM Introduction Dan Rieder
9:10 AM Reports from Break Out Groups
9:50 AM Break Out Group A (Topic 3) Jane Staveley, Facilitator
Effects Assessment: See Attachment
Break Out Group B (Topic 3) Dan Rieder, Facilitator
Effects Assessment: See Attachment
10:30 AM Break
10:40 AM Continue Break Out Group Discussions (Topic 3)
12:00 AM Lunch
1:00 PM Break Out Group A (Topic 4) Jane Staveley, Facilitator
Integration and Risk Characterization
Break Out Group B (Topic 4) Dan Rieder, Facilitator
Integration and Risk Characterization
3:00 PM
3:10PM
3:40 PM
Break
Reports from Break Out Groups
Closing Remarks Dan Rieder, Facilitator
-------
Attachment to Agenda
Topics for Discussion at
Non-Target Plant Risk Assessment Workshop
1. Risk Assessment Problem Formulation and Framework:
A. Which pesticides/uses require a plant risk assessment? What criteria can be developed and
how can they be used to reduce phytotoxicity as a concern? What factors are considered before
developing a problem formulation? What are the limitations of the risk assessment in helping us
understand the potential for adverse effects to non-target plants?
B. What sources of information, e.g., open literature, laboratory tests, etc., should be considered
and how can this information be used in a generic problem formulation for a non-target plant risk
assessment? What types of other information, e.g., efficacy, residue chemistry, etc., should be
used in a risk assessment? Are the elements of a non-target plant risk assessment similar to those
for animals?
C. How do stakeholders affect the risk assessment? What role do they play? How do
stakeholders' values affect the risk assessment? How do risk managers affect the direction of the
risk assessment?
D. What tiers or levels are currently used in a risk assessment framework? What data are
required for each tier or level? What are the triggers for moving to a higher tier or level? Is the
current framework adequate? What, if any, changes should be considered to the framework?
2. Exposure Assessment:
A. How is the exposure assessment performed at each level or tier? What kinds of data are
needed to assess exposure?
B. What chemical properties are important in assessment of exposure? If formulations are tested,
which ones and how many? How does the type of test material affect the need for additional
data, the assessment, and any further refinements? What important chemical or fate properties
are considered in testing/assessment?
C. What types of models and scenarios should be used?
D. What research is currently being conducted for assessing exposure to non-target plants?
What additional research needs to be conducted?
E. How is uncertainty expressed?
-------
2
3. Effects Assessment:
A. What are the current tests and endpoints used for assessing effects? What are the merits of
each? What types of effects should be measured and what kinds of tests are needed to measure
these effects? What characteristics are important in selecting a toxicity test species?
B. What is tested, e.g. active ingredient, formulated product, etc., at each level and why? If
formulations are tested, which ones and how many? How does the type of test material affect the
need for additional data, the assessment, and any further refinements? What important chemical
or fate properties are considered in testing/assessment?
C. How are species selected for testing: taxonomic or ecological factors?
D. What research is currently being conducted for assessing effects to non-target plants? What
additional research needs to be conducted?
E. How is incident data used or followed? How can regulatory agencies improve incident
reporting?
F. How is uncertainty expressed? How can uncertainty factors be used and what do they
represent, e.g., species variability, lab-to-field extrapolation, etc.?
4. Risk Characterization
A. Besides the risk quotient, what other information should be used in making risk
characterization decisions, e.g., consideration of the pattern of toxicity?
B. What are the appropriate ways to express uncertainty? Should we express uncertainty for
endpoints or expected effects that are not studied?
C. What is the best way to present the results of the risk characterization to risk managers?
What is the nature of the interface between risk characterization and risk management?
D. Should the concept of recovery be incorporated into risk characterization? If so, how?
E. What is the role of mitigation and risk reduction measures? How can the risk characterization
be used to provide input to the development of mitigation options? What is the role of post-
registration monitoring? What are the advantages and disadvantages of obtaining monitoring
data?
-------
------- |