A Comparative Analysis of
Ecological Risks from Pesticides
and Their Uses: Background,
Methodology & Case Study
Environmental Fate & Effects Division
Office of Pesticide Programs
U. S. Environmental Protection Agency
Washington, D.C.
November, 1998
-------
PROJECT TEAM
Environmental Fate and Effects Division
Douglas Urban, primary author
Tom Steeger
Ed Odenkirchen
-------
CONTENTS
Page
I. INTRODUCTION
A. Purpose
B. Background
C. Scope & Method
D. Approach - Case Study Using Selected Insecticides & Use Sites
II. AVIAN EFFECTS AND EXPOSURE ASSESSMENT
A. Effects
1. Acute Toxicity to Birds
2. Chronic Toxicity to Birds
B. Exposure
1. Acute Exposure via Dose (mg a.i./ ft2 available)
2. Acute Exposure via Diet (ppm available in diet)
3. Chronic Exposure via Diet (ppm available in diet)
III. AVIAN RISK QUOTIENTS AND LEVELS OF CONCERN
A. Calculation of the Acute Avian Risk Quotients
1. Avian Acute Dose Risk via Ingestion - Granular Formulations
2. Avian Dietary Risk - Spray Formulations
3. Avian Acute Bird per Day Risk - Spray Formulations
B. Calculation of the Chronic Avian Risk Quotients
-------
1. Avian Chronic risk - Spray Formulations
2. Number of Days to reach Avian Chronic LOC (RQ=1) - Spray
Formulations
C. The Avian Risk Column - % Pesticide Contribution to RQ Sum for
Each Endpoint on Each Site (Crop)
D. Frequency of LOC Exceedance
IV. AQUATIC EFFECTS AND EXPOSURE CHARACTERIZATION
A. Effects
1. Acute Toxicity to Freshwater Fish
2. Acute Toxicity to Marine/Estuarine Fish
3. Acute Toxicity to Freshwater Invertebrates
4. Acute Toxicity to Marine/Estuarine Crustaceans
5. Acute Toxicity to Marine/Estuarine Molluscs
6. Chronic Toxicity to Freshwater Fish
7. Chronic Toxicity to Freshwater Invertebrates
B. Exposure
1. Acute and Chronic Exposure Using GENEEC Model
V. AQUATIC RISK QUOTIENTS AND LEVELS OF CONCERN
A. Calculation of the Acute Fish & Aquatic Invertebrate Risk Quotients
1. Acute Freshwater Fish Risk
-------
2. Acute Marine/Estuarine Fish Risk
3. Acute Freshwater Invertebrate Risk
4. Acute Marine/Estuarine Crustacean Risk
5. Acute Marine/Estuarine Mollusc Risk
B. Calculation of the Chronic Fish & Aquatic Invertebrate Risk Quotients
1. Chronic Freshwater Fish Risk
2. Chronic Freshwater Invertebrate Risk
C. The Aquatic Risk Column - % Pesticide Contribution to RQ Sum for
Each Endpoint on Each Site (Crop)
D. Frequency of LOC Exceedance
VI. COMPARATIVE ECOLOGICAL RISK ANALYSIS
A. Calculation of Potential Risk (% of Risk)
B. Percentage (%) Acres Treated
C. Incident Reports - Bird and Fish Kills
D. Comparison of Potential Risk by Crop Site
1. Alfalfa - Spray Formulations Only
2. Corn - Granular and Spray Formulations
3. Cotton - Granular and Spray Formulations
4. Peanuts - Granular and Spray Formulations
5. Overall Summary
VII. DECISION SUPPORT ANALYSIS TO AID DECISION-MAKING
-------
A. Decision Support Software
B. Baseline Scenario - Alfalfa
C. Ecological Risk Characterization Information
D. Scenario #1 - Baseline Plus Ecological Risk Characterization Information
E. Additional Scenarios - Changes in Criteria Importance
VIII. LIMITATIONS OF THIS ANALYSIS
IX.
REFERENCES
-------
TABLES
Table 1 -
Chemicals & Uses for Comparative Ecological Risk Analysis (17
Chemicals & Uses)
Table 2 -
Linear Regression Parameter Estimates
Table 3 -
Environmental Fate Parameters
Table 4 -
GENEEC Model Report Format
Table 5 -
Data for Figure 3
Table 6 -
Data for Figure 4
Table 7 -
Data for Figure 5
Table 8 -
Data for Figure 6
Table 9 -
Data fro Figure 7
Table 10-
Data for Figure 8
Table 11 -
Data for Figure 9
Table 12 -
Data for Scenario #1
FIGURES
Figure 1 - Avian Risk, Acute Bird per Day Risk on Alfalfa
Figure 2 - Aquatic Risk, Acute Freshwater Fish Risk on Alfalfa
Figure 3 - Combined Charts - Comparative Risk for Pesticides Used on Alfalfa as
PostEmergent Sprays
Figure 3a - Comparative Avian Risk for Pesticides Used on Alfalfa as PostEmergent
Sprays
Figure 3b - Comparative Aquatic Risk for Pesticides Used on Alfalfa as PostEmergent
Sprays
Figure 3c - Incident Reports for Pesticides Used on Alfalfa
Figure 4 - Combined Charts - Comparative Risk for Pesticides Used on Corn as
Granular At-Plant
Figure 4a - Comparative Avian Risk for Pesticides Used on Corn as Granular At-Plant
Figure 4b - Comparative Aquatic Risk for Pesticides Used on Corn as Granular At-
Plant
Figure 4c - Incident Reports for Pesticides Used on Corn
Figure 5 - Combined Charts - Comparative Risk for Pesticides Used on Corn as
PostEmergent Sprays
Figure 5a - Comparative Avian Risk for Pesticides Used on Corn as PostEmergent
Sprays
Figure 5b - Comparative Aquatic Risk for Pesticides Used on Corn as PostEmergent
Sprays
Figure 6 - Combined Charts - Comparative Risk for Pesticides Used on Cotton as
Granular At-Plant
Figure 6a - Comparative Avian Risk for Pesticides Used on Cotton as Granular At-
Plant
Figure 6b - Comparative Aquatic Risk for Pesticides Used on Cotton as Granular At-
Plant
-------
Combined Charts - Comparative Risk for Pesticides Used on Cotton as
PostEmergent Sprays
Comparative Risk for Pesticides Used on Cotton as PostEmergent
Sprays
Comparative Risk for Pesticides Used on Cotton as PostEmergent
Sprays
Incident Reports for Pesticides Used on Cotton
Combined Charts - Comparative Risk for Pesticides Used on Peanuts as
Granular At-Plant
Comparative Avian Risk for Pesticides Used on Peanuts as Granular At-
Plant
Comparative Aquatic Risk for Pesticides Used on Peanuts as Granular At-
Plant
Combined Charts - Comparative Risk for Pesticides Used on Peanuts as
PostEmergent Sprays
Comparative Avian Risk for Pesticides Used on Peanuts as
PostEmergent Sprays
Comparative Aquatic Risk for Pesticides Used on Peanuts as
PostEmergent Sprays
Baseline: Decision Table for Which of the Insecticides Used on Alfalfa are
Less Risky
The Relative Strengths fo the 10 Choices Considering the 11 Risk
Endpoints
Baseline: Decision Table for Which of the Insecticides Used on Alfalfa are
Less Risky
Summary of Scenarios
APPENDICES
Appendix 1 - Pesticide Usage Information
Appendix 2 - Pesticide Label Information
Appendix 3 - Acute Oral Toxicity of Insecticides to Birds
Appendix 4 - Dietary Toxicity of Insecticides to Birds
Appendix 5 - Chronic Toxicity of Insecticides to Birds
Appendix 6 - Estimated Environmental Concentrations (mg a.i./sq ft available to birds)
Appendix 7 - Estimated Environmental Concentrations (ppm in avian diets))
Appendix 8 - Input for Avian Chronic Time to RQ=1
Appendix 9 - Avian Acute Risk (LD50s/ft2)
Appendix 9a -% Pesticide Contribution to RQ Sum for Avian Acute Risk
Appendix 9b -Frequency of Exceeding the LOC for Avian Acute Dose Risk
Appendix 10 -Avian Dietary Risk (EEC/LC50)
Appendix 10a -% Pesticide Contribution to RQ Sum for Avian Dietary Risk
Appendix 10b -Frequency of Exceeding the LOC for Avian Dietary Risk
Appendix 11 - Avian Acute Bird per Day Risk ((EEC x %Food lngestion/day)/LD50)
Appendix 11a -% Pesticide Contribution to RQ Sum for Avian Acute Bird per Day Risk
Figure
7 -
Figure
7a-
Figure
7b-
Figure
7c-
Figure
8-
Figure
8a -
Figure
8b-
Figure
9-
Figure
9a -
Figure
9b-
Figure
10-
Figure
11 -
Figure
12 -
Figure
13-
-------
Appendix 11b -Frequency of Exceeding the LOC for Avian Acute Bird per Day Risk
Appendix 12 - Avian Chronic Risk (EEC/NOAEC)
Appendix 12a -% Pesticide Contribution to RQ Sum for Avian Chronic Risk
Appendix 12b -Frequency of Exceeding the LOC for Avian Chronic Risk
Appendix 13 - Avian Chronic Risk (Time to RQ=1)
Appendix 13a -% Pesticide Contribution to Sum of Days to Time to RQ=1
Appendix 14 - Acute Toxicity of Insecticides to Freshwater Fish
Appendix 15 - Acute Toxicity of Insecticides to Marine/Estuarine Fish
Appendix 16 - Acute Toxicity of Insecticides to Freshwater Invertebrates
Appendix 17 - Acute Toxicity of Insecticides to Marine/Estuarine Crustaceans
Appendix 18 - Acute Toxicity of Insecticides to Marine/Estuarine Molluscs
Appendix 19 - Chronic Toxicity of Insecticides to Freshwater Fish
Appendix 20 - Chronic Toxicity of Insecticides to Freshwater Invertebrates
Appendix 21 - GENEEC Model Run Results
Appendix 22 - Freshwater Fish Acute Risk
Appendix 22a -% Pesticide Contribution to RQ Sum for Freshwater Fish Acute Risk
Appendix 22b -Frequency of Exceeding the LOC for Freshwater Fish Acute Risk
Appendix 23 - Marine/Estuarine Fish Acute Risk
Appendix 23a - % Pesticide Contribution to RQ Sum for Marine/Estuarine Fish Acute
Risk
Appendix 23b - Frequency of Exceeding the LOC for Marine/Estuarine Fish Acute Risk
Appendix 24 - Freshwater Invertebrate Acute Risk
Appendix 24a - % Pesticide Contribution to RQ Sum for Freshwater Invertebrate Acute
Risk
Appendix 24b - Frequency of Exceeding the LOC for Freshwater Invertebrate Acute
Risk
Appendix 25 - Marine/Estuarine Crustacean Acute Risk
Appendix 25a - % Pesticide Contribution to RQ Sum for Marine/Estuarine Crustacean
Acute Risk
Appendix 25b - Frequency of Exceeding the LOC for Marine/Estuarine Crustacean
Acute Risk
Appendix 26 - Marine/Estuarine Mollusc Acute Risk
Appendix 26a - % Pesticide Contribution to RQ Sum for Marine/Estuarine Mollusc
Acute Risk
Appendix 26b - Frequency of Exceeding the LOC for Marine/Estuarine Mollusc Acute
Risk
Appendix 27 - Freshwater Fish Chronic Risk
Appendix 27a - % Pesticide Contribution to RQ Sum for Freshwater Fish Chronic Risk
Appendix 27b - Frequency of Exceeding the LOC for Freshwater Fish Chronic Risk
Appendix 28 - Freshwater Invertebrate Chronic Risk
Appendix 28a - % Pesticide Contribution to RQ Sum for Freshwater Invertebrate
Chronic Risk
Appendix 28b - Frequency of Exceeding the LOC for Freshwater Invertebrate Chronic
Risk
Appendix 29 - Total Number of Bird and Fish Incidents Reported for 17 Insecticides on
-------
All Sites
Appendix 30 - Bird and Fish Incidents Reported for 17 Insecticides on Four Crops
Appendix 31 - EFED FATE Program
-------
1
I. INTRODUCTION
A. Purpose
This document describes a proposed approach and methodology, which is under
development, for comparing the ecological risk of pesticides and their uses in the
Environmental Protection Agency's (EPA or the Agency) Office of Pesticide Programs
(OPP). Risk assessors are often asked to compare the ecological risks posed by
different pesticides registered (or being considered for registration) for use on a
specific crop. Comparative analyses can help to ensure consistency in risk
management decisions and to focus more significant ecologically-based risk
management decisions on those pesticides that pose the greatest risk to fish and
wildlife. Therefore, OPP seeks to define standard methods for comparative ecological
risk assessment that are scientifically sound and capable of being implemented using
currently available data and resources. We would expect to update or replace these
methods as additional data and probabilistic assessment tools become available.
Methods will be presented for comparing the potential ecological risk of
pesticides used on similar crop sites. Risk indices or risk quotients (RQs)1 are
calculated and the results are compared to established levels of concern (LOCs)2. In
addition, since numerous RQ calculations are made using a range of use rates and
ecotoxicity values, pesticides and their use sites are compared based on frequency (%)
of LOC exceedances. The resultant exceedances and frequencies are used to rank
pesticides used on the similar use sites according to risk. The comparisons include
acute and chronic endpoints for terrestrial and aquatic organisms, as well as incident
reports and information on extent of use. Pesticide specific ecotoxicology data and
environmental fate and transport data are used in the analysis. Screening models such
1A risk quotient is the ratio of the estimated environmental concentration of a chemical to a toxicity
test effect level for a given species. It is calculated by dividing an appropriate exposure estimate (e.g. EEC)
by an appropriate toxicity test effect level (e.g. LC50).
2Levels of Concern (LOC's) are criteria used to indicate potential risk to non-target organisms and
the need to consider regulatory action. The criteria indicate that a pesticide, when used as directed, has the
potential to cause adverse effects on non-target organisms. Since the issuance of a 1992 policy by OPPTS
[1 and 2], OPP has generally pursued ecological risk mitigation whenever these levels of concern are
exceeded.
-------
2
as GENEEC3 and FATE4 are used here to estimate pesticide exposure, but results from
more sophisticated models such as PRZM/EXAMS5 also could be used.
B. Background
Risk managers in EPA's Office of Pesticide Programs have had a longstanding
desire to better understand the relative ecological risk posed by pesticides so that this
information can be factored into decisions regarding priorities for risk management and
decisions regarding degrees of needed risk mitigation. With the passage of FQPA and
its mandate for EPA to conduct cumulative human health risk assessments for
pesticides with common mechanisms of toxicity, risk managers are even more
interested in understanding which pesticides pose the greater or lesser ecological risk.
Simply stated, risk managers understand that the purposeful release of biological
poisons into the environment will result in impacts to exposed non-target aquatic and
terrestrial species; what risk managers most desire is to focus the more significant
ecological risk mitigation actions on those pesticides and pesticide use patterns which
result in the greatest threat to non-target species. Although EFED is undertaking a
major multi-year effort to improve its risk assessment methods (an effort which will likely
lead to the use of probabilistic risk assessment methods and the collection of some
different data than has been historically required for registration and reregistration), a
major challenge for scientists in EFED today is to effectively use the data which are
currently and typically provided to support registration and reregistration and current
risk assessment methods in order to provide risk managers with an accurate sense of
relative risk.
For a typical food-use pesticide, EPA requires and generally has available the
following ecotoxicological data:
1. Mammalian Acute Toxicity (Rat LD50)
2. Avian Acute Toxicity (one species - Oral LD50; two species -
Dietary LC50's)
3. Avian Chronic Toxicity (two species, avian reproduction - NOAEC)
3GENEEC is a PC - based computer program which is designed to allow the userto quickly calculate
a set of generic (non-site specific) estimated environmental concentrations (EEC's) given limited
environmental fate data and pesticide label information [21],
4FATE model is PC - based computer program designed to allow the user to quickly calculate
conservative, non-site specific, exposure values for avian and mammalian risk assessments [Appendix 31],
5PRZM/EXAMS is a combination of a runoff model (PRZM2, Pesticide Root Zone Model) [26 and
27] and a surface water receiving model (EXAMS, Exposure Analysis Modeling System) [28] designed to
provide a distribution of EEC values, in time and space, for the crop area in which the pesticide has been
applied.
-------
3
4. Honey Bee Contact or Residue Toxicity
5. Terrestrial Plant Toxicity (vegetative vigor & seedling emergence -
10 species - EC50's )
6. Fish Acute Toxicity (two species, coldwater & warmwater - LC50's)
7. Fish Chronic Toxicity (one species - NOAEC)
8. Aquatic Invertebrate Acute Toxicity (one species - EC50)
9. Aquatic Invertebrate Chronic Toxicity (one species - NOAEC)
10. Estuarine/Marine Acute Toxicity (three species, fish, shrimp,
crustacean, mollusc - EC50's)
11. Aquatic Plant Toxicity ( 5 species - EC50's)
For a typical food-use pesticide, EPA requires and generally has available the
following exposure/fate and transport data:
1. Solubility in water
2. Volatility (vapor pressure)
3. Octanol/Water Partition Coefficient (as Log Kow)
4. Hydrolysis at pH 5 , 7, and 9; (when applicable, dissociation
constant)
5. Photolysis in water (half-life)
6. Photolysis on soil (half-life)
7. Aerobic/Anaerobic soil metabolism (half-life) (includes information
on soil type)
8. Aerobic/Anaerobic aquatic metabolism (half-life)
9. Mobility in soil data for parent & major degradates (includes
information on soil type)
10. Field dissipation studies, according to use pattern (includes brief
description of study sites)
11. Soil Adsorption/desorption with Kd and sorption coefficients,
values (range & median) for parent and environmental degradates
12. Bioaccumulation in fish
Field studies for investigating terrestrial and/or aquatic effects are not required
for all pesticides, but only those whose labeled use raised concern for high risk. Such
pesticides usually have high risk quotients for acute or chronic effects to birds, fish or
aquatic invertebrates. In addition, some other information characterizing the risk is
often available such as kill incident reports, data on environmental persistence, multiple
applications, or a pattern of widespread use.
This is not the first attempt by EPA's Office of Pesticide Programs (OPP) to
develop an approach for comparing the ecological risks posed by different pesticides in
a regulatory context. In March, 1992, EPA's Office of Pesticide Programs published a
-------
4
document titled "Comparative Analysis of Acute Avian Risk from Granular
Pesticides"[3], It described OPP's approach for screening granular formulation
pesticides to identify those that may pose acute lethal risk to birds. That document
focused on granular pesticides because of the particular acute risk they pose to birds.
The analysis was based on the calculation of an acute risk quotient and a weight-of-
evidence approach to characterize the ecological risk, considering confirmatory field
studies and bird kill incident reports. The analysis found that 14 granular pesticides
pose potentially high risk to birds due to their high acute toxicity and availability in the
environment. The report was released to the public and the regulated community.
Based on written technical comments provided by the registrants, the Agency
developed a series of generic risk assessment issues that could benefit from further
research [4], Such issues as the effect of granular substrate, avian preference, the
efficiency of various incorporation methods, the effect of watering in granules, and
insecticide/fertilizer mixes on risk of granular pesticides to birds were identified. The
registrants have already submitted data that have been useful in refining Agency risk
assessment for granular pesticides.
Given the limitations of the data generally available to EPA and the limitations of
available EPA resources, EFED recognizes it is necessarily limited in its ability to use
relative risk methods to make "fine" distinctions between the ecological risk posed by
particular pesticides. That is, given the nature of the available data and the amount of
time and effort that can be routinely dedicated to completing ecological risk
assessments6, OPP cannot expect to rank all active ingredients with precision.
However, EFED does believe that it is possible to use available data and current risk
assessment methods to identify those pesticides and pesticide uses which pose
significantly less ecological risk than others, so that EFED can answer the very real
questions that risk managers ask every day. EFED believes that the proposed risk-
based methodology for comparing pesticides may provide us with an appropriate tool to
distinguish between pesticides that are greatly different from one another.
The document provides two basic approaches to aid decision-making: the first is
a graphical presentation consisting of a series of bar charts comparing pesticides
based on risk functions; the second is a tabular presentation with a summarized
ranking of the same pesticides based on decision analysis of the risk functions. Ideally,
both would be used concurrently and would complement one another.
A series of questions are being posed to the SAP based on the two approaches.
Where the panel may conclude that an approach or method is inappropriate, the
6EFED technical staff of 78 scientists has responsibility for conducting ecological risk assessments
and water resource assessments for over 400 registered pesticide active ingredients and 10 to 20 new active
ingredients (per year).
-------
5
Agency is very interested in suggestions for alternative approaches and/or methods
that can be adopted within the constraints of available data and available resources. .
First, the fundamental question:
Based on the typical sets of studies, data, and information provided for pesticide
risk assessment purposes in the regulatory context and the use of OPP's current
risk assessment methods, is this approach useful/meaningful for evaluating the
relative potential risk of pesticides and pesticide uses, especially for
ascertaining large distinctions between the risk posed by pesticides?
Second, concerning the graphical presentation:
Is this approach useful/meaningful for comparing the relative potential risk of
pesticides and pesticide uses?
In the previous comparative analysis of granular insecticides the Agency
compared the pesticides based on calculated risk quotients. In this analysis, the
Agency sought to incorporate more of the available information into the
comparison. Thus, there are a number of new calculations for expressing
potential risk, such as the % contribution to the RQ sum, the frequency of RQ
exceedance, the % contribution to the Time to RQ=1 sum, and % risk. Are these
useful parameters for comparing the potential risk of pesticides?
The graphs are presented in order of decreasing percent of acres treated.
Otherwise, the extent of use is not factored into the risk calculations. Is this
appropriate?
The use of granular formulations is likely to present chronic risk to birds;
however, OPP currently has no method to calculate this risk. Therefore the
proposed approach addresses chronic risk to birds only for sprayable
formulations. Does the Panel agree that the avian chronic risk should be
included for sprayable formulations despite the Agencies inability to include this
risk element for granular formulations? Should the Agency explore ways to use
the avian chronic risk quotient for sprayable formulations as a surrogate to
address this risk factor when comparing granular formulations?
Third, concerning the tabular presentation based on decision analysis:
Is this approach useful/meaningful for comparing the relative potential risk of
pesticides and pesticide uses?
One of the simplifications of the methodology used in the decision analysis
-------
6
software is that there are no uncertainties. This is certainly not the case in
comparative ecological risk analysis. However, one of the ways the software
permits the user to deal with some uncertainty is by running multi-scenarios.
This was done for the case study. These runs show how sensitive the
differences between pesticides are to change in the importance of the criteria.
The results can add confidence to overall conclusions. Does the panel agree
that his approach is useful and can increase the confidence of conclusions
derived from the results?
Incidents were treated as important when they exist; however, when there were
no incident reports this element was given zero weight in the analysis. Is this an
appropriate use of incident data for this comparative analysis?
C. Scope and Methods
This comparative analysis builds upon the earlier comparative risk document.
The Comparative Analysis of Acute Avian Risk from Granular Pesticides was focused
solely on one kind of formulation (granular) and one endpoint (acute risk to birds). In
order to present the approach and methods for this more expansive and complex
analysis, we chose to present a case study of 17 insecticide chemicals and four use
sites, alfalfa, corn, cotton, and peanuts. In order to maintain the focus of this analysis
on the methodology and not on individual pesticides, the 17 insecticides were
designated as Chemical A through Q.
In this new analysis eleven endpoints were selected for comparing both at-plant
granular formulations and post-emergent sprays. They include both acute and chronic
risk for birds, fish and aquatic invertebrates. The aquatic endpoints covered exposure
in both the freshwater and marine/estuarine environments. In addition, the analysis
incorporates the standards of ecological risk assessment and management, the Levels
Of Concern (LOCs), provided in the 1992 Agency policy document [2], Consequently,
risk comparisons can be made in a more equitable fashion for all pesticides included in
this analysis.
EPA calculated risk quotients for acute and chronic risk to birds, fish and aquatic
invertebrates. These risk indices are based upon estimates of pesticide exposure and
ecotoxicity. In turn, these estimates are based on available pesticide label information,
ecotoxicity and environmental fate data, as well as widely accepted models.
The risk quotient methodology described in this analysis has previously been
available for both public and scientific review. It has become common in ecological risk
assessment to present potential risk in terms of a ratio of the estimated environmental
exposure or EEC, divided by the hazard of toxicity such as the LC50 EC50 or No
-------
7
Observed Adverse Effects Concentration (NOAEC). EPA first presented this risk index
method in the Standard Evaluation Procedure for Ecological Risk Assessment in 1986
[4], These ratios are used to express potential acute and chronic risk to birds, fish and
aquatic invertebrates.
Ecological risk assessment is an evolving field. EPA sponsors research and
works with industry and other agencies in a continuing effort to refine the Agency's
ecological risk assessment methodologies. Of particular note is the Ecological
Committee on FIFRA Risk Assessment Methods (ECOFRAM) which was formed in
June 1997. Its purpose is to develop tools and processes within the FIFRA framework
for predicting the magnitude and probabilities of adverse effects to non-target aquatic
and terrestrial species resulting from the introduction of pesticides into their
environment. ECOFRAM was convened in response to a review of OPP's ecological
risk assessments and guidelines in May of 1996 by the FIFRA Scientific Advisory Panel
(SAP). While recognizing and generally affirming the utility of the current assessment
process and methods for screening risk assessment purposes, the SAP noted that OPP
has relied on deterministic methods of assessing the ecological effects of pesticides
and strongly encouraged OPP to develop and validate tools and methodologies to
conduct probabilistic assessments of ecological risk. This resulted in the formation of
ECOFRAM. As tools for probabilistic ecological risk assessments become available
and are implemented in OPP/EFED, this comparative analysis of ecological risk
assessments will need to be revised and updated.
As noted previously, the risk quotients in this screening analysis are used to
indicate potential ecological risk. They find their greatest utility when used as the basis
for comparing the potential acute and chronic risk to birds, fish and aquatic
invertebrates posed by different pesticides used on the same sites, under similar
exposure scenarios. When used with additional information such as reports of bird and
fish kill incidents, refined estimates of exposure, common practice mitigatory measures,
unique site characteristics, etc. this analyses is useful for identifying those pesticides
which pose comparatively higher or lower risk. However, it is important to clarify the
limitations of this approach.
This analysis is similar to a predictive model. It is based on data inputs such as
laboratory eco-toxicity data, fate data from laboratory and/or field studies, computer
generated model exposure estimates, and use data from the pesticide labels. The
quality of the results from any model reflect the quality of the input data and the
adequacy of the models used to accurately represent the most significant processes
affecting a pesticide's fate and biological effects in the environment, and the
dependence of those behaviors on the selected input parameters. The limitations of the
approaches used here are discussed more fully in Section VIII of this paper.
Nevertheless, the Agency believes that the choices of input data and risk calculations
are useful for comparing potential and relative risks among pesticides used as
-------
8
alternatives on the same site.
The current analysis is intended only to compare the potential acute and chronic
risk to birds, fish and aquatic invertebrates posed by these insecticides used on these
four crop sites. A complete ecological risk assessment of any of these pesticides would
include an evaluation of other information including acute and chronic risk to other non-
target organisms such as wild mammals and non-target plants, consideration of
exposure refinements based on site-specific use data, an account of the extent,
location and ecological sensitivity of the areas treated and a comprehensive
assessment of available field effects data (terrestrial field studies and mesocosms),
including detailed incident reports. These factors were not included as part of the
current analysis.
EPA recognizes the potential risk to wild mammals and non-target plants from
these and other highly toxic pesticides and will address those risks in future
comparative assessments. Additionally, the results of this analysis could change if the
input data changes. New eco-toxicity or environmental fate data, or updated use
information will result in changes in the quotients, their LOC exceedance, and the
relative comparisons. As such, the results of the analysis and the conclusions drawn
are dynamic. The Agency recognizes that as additional information is made available,
both the results and conclusions are subject to change.
This analysis is valuable in that it identifies and compares pesticides and uses
presenting the relative acute and chronic potential risk to birds, fish and aquatic
invertebrates. The scope of this analysis is considerably expanded over the previous
analysis for granular pesticides. However, it is still incomplete. Despite this, the
analysis presents an interesting approach for comparing ecological risk based on
sound data and a well documented methodology.
D. Approach - Case Study Using Selected Insecticides & Use Sites
Four major crop sites were selected for analysis - alfalfa, corn, cotton and
peanuts. Current pesticide labels for insecticides commonly used on these sites were
reviewed . Based on this review, 17 pesticides were selected and designated as
Chemical A through Q. For this case study, we will assume that these 17 pesticides
were all of the same chemical class, with similar modes of action. This resulted in a
total of 38 pesticide-use site combinations (See Table 1). Hypothetical usage data was
generated and is found in Appendix 1. Appendix 2 lists the information collected from
selected pesticide labels. This included various formulations, timing, types, and
methods for application for each pesticide-use combination.
This case study analysis is not intended to characterize all the risks of the
-------
9
chemicals in the analysis or serve as the sole basis for decision-making. Rather, it is
provided as an illustrative example demonstrating how pesticides used as alternatives
on the same site can be compared based on ecological risk and the results used to aid
decision-making.
-------
10
Table 1. Chemicals & Uses Chosen for Comparative Ecological Risk Analysis (17 Chemicals & 4 Uses)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Chemical Names ^Selected Use
mites
1 Alfalfa (10)* Corn (8) Cotton (14) Peanuts (6)
# of Usd
S/'fesI
per Chemical\
2
B
2
C
3
D
3
E
G
F
'
G
3
H
f
t.
I
1
J
1
K
2
L
3
M
2
N
2
0
V
P
1
* The Number of Chemicals per use Site
-------
11
II. AVIAN EFFECTS AND EXPOSURE ASSESSMENT
A. Effects
EPA typically receives the following required laboratory studies to use in
performing avian risk assessments: acute bird LD507 (mg/kg) and LC508 (ppm) studies,
and chronic bird reproduction studies, providing a NOAEC9 (ppm). The Agency
evaluates the studies and classifies them as either core10, supplemental11 or invalid12,
as well as indicating whether the supplemental and invalid studies are upgradable13.
The toxicity values from the core and supplemental studies are used in risk
assessment.
This analysis used core and supplemental eco-toxicity data in the Eco Tox data
base [5] to characterize the effects of the selected pesticides on birds. All the data used
in this analysis has been updated and verified. However, some errors in this analysis
could result from entry errors. In addition, the data used in this analysis reflects the
status of the data base as of July, 1998. It does not include data available after that
date. More recent data could change the results of this analysis.
Avian LD50 (mg/kg) and LC50 (ppm) values and chronic NOAEC (ppm) values for
the most sensitive species tested (the lowest values) were selected from the data base.
In addition, the median LD50 (mg/kg) and LC50 (ppm) values were calculated. These
median values provided a less conservative estimate of the toxicity values (compared
to the lowest) and provided additional values for determining a range of toxicity values
for a particular endpoint.
Since relatively few species are used in standard toxicity testing, it is likely that
the species most sensitive to each pesticide has not been tested. The few species that
are tested often provide a range of toxicity values, reflecting the combined effects of
measurement error, variability in sensitivity among individuals within a species, and
7 Median lethal dose necessary to affect (kill) 50% of the test population.
8 Median lethal concentration in the diet necessary to affect (kill) 50% of the test population.
9 The highest concentration tested in the study where no adverse effects were observed.
10 Scientifically sound study which also meets EPA published guideline requirements.
11 Scientifically sound study with some deviations from published EPA guideline requirements.
12 Study has flaws that make it's results unreliable to use in risk assessment.
13 Additional data (e.g., sample storage stability data) could make a study useable for risk
assessment.
-------
12
species-to-species variation in sensitivity to the pesticide being tested. Because of this
variation in sensitivity, it is unlikely that this analysis will show the worst case risk for
each pesticide considered. Rather, based on the calculated toxicity values, the analysis
will provide a range of risk values for purposes of comparison and identification of
those pesticides that are more likely to cause adverse effects in actual use.
1. Acute Toxicity to Birds
Based on years of experience in preparing risk assessments, EPA/OPP has
found that the LD50 value, compared to the LC50 value, is often a better indicator of
acute toxicity to birds [5], This seems to be true especially for pesticides with LD50
values less than or equal to 50 mg/kg. Alternately, the LC50 value may be a better
indicator of acute toxicity to birds if their LD50 values are greater than 50 mg/kg and
they persist in the environment with a half-life greater than one day. For this analysis,
however, both the avian acute oral LD50 value and the avian subacute dietary LC50
value were included and used in the risk calculations.
The avian LD50 value is usually expressed in mg/kg of body weight. However, it
is well established that the body weight of a bird is a very important consideration when
determining how sensitive any individual bird will be to acute pesticidal effects,
Therefore, the LD50 value was adjusted by multiplying it by the weight of a bird to arrive
at an LD50 per bird value. For this analysis, EPA chose 20 gm to represent small birds
(e.g., songbirds); 100 gm to represent medium size birds such as small upland game
birds (e.g., quail); and 1000 gm to represent large upland game birds and waterfowl
(e.g., pheasants and geese).
EPA ranked the 17 pesticides in order of their lowest acute oral LD50 toxicity
values (Appendix 3) and their lowest acute dietary LC50 toxicity values (Appendix 4).
The median of the data including data on all bird species tested for each pesticide was
included.
2. Chronic Toxicity to Birds
The NOAEC are typical values resulting from the avian reproduction test.
Typically, two species, bobwhite quail and mallard ducks, are tested. Common
reproductive effects found in these tests are egg thinning, cracked eggs, reduced
hatchability, decreased survival rate, reduced growth of Fgeneration and reduced egg
production.
EPA ranked the 17 pesticides in order of their lowest chronic toxicity, that is
NOAECs (Appendix 5). For chemicals lacking data on chronic effects to birds, a value
-------
13
was estimated by applying an acute to chronic ratio value. This was calculated based
on a regression analysis (Table 2) of acute toxicity values over long-term exposure
effect values for the all the pesticides in this class. This regression equation was used
to predict missing values. Although revalues for the regression suggested that they
were not predictive over the entire range of the regression, the regression coefficient
was significant (P<0.05) and residuals were minimal for acute toxicity values requiring
the regression analysis.
-------
14
Table 2. Linear regression parameter estimates following the format: Dependent
Variable = Slope (Independent Variable) + y-intercept. Regression analyses were
conducted using toxicity estimates from studies involving > 50% active ingredient.
Dependent
Variable1
slope
y-intercept
Independent
Variable
r2
avian chronic
NOAEC
0.03081
6227.22
median acute
avian LC502
0.39
freshwater fish
chronic NOAEC
0.157
-24.5931
median
freshwater fish
acute LC503
0.27
freshwater fish
chronic NOAEC
0.05933
-6.809434
median
freshwater fish
acute LC503
0.79
marine/estuarine
fish acute LC50
0.076924
458.1463
median
freshwater fish
acute LC503
0.99
freshwater
crustacean NOAEC
0.0127
2.372717
median
freshwater
crustacean
acute EC50
0.57
marine/estuarine
crustacean acute
ec50
0.599898
7.282905
median
freshwater
crustacean
acute EC504
0.97
marine/estuarine
mollusc acute EC50
2.489109
47.852861
median
marine/estuarine
crustacean
acute EC50
0.79
1estimate expressed as parts per billion (ppb).
degression equation developed using Guideline 71-2 median toxicity (LC50) estimates,
degression equation developed using Guideline 72-4 toxicity estimates of 4,000 ppb.
degression equation developed using Guideline 72-3 median toxicity estimates
excluding estimate of 43 ppb.
-------
15
B. Exposure
Both granular and non-granular formulations are being considered in this
analysis. The amount of toxicant a bird is likely to consume in the diet or by preening,
ingest as a single dose, inhale, or absorb via the eye or through the skin, is currently
not quantified as it is for human exposure. Research has begun, but is limited at this
time [6], Among bird species, there are tremendous differences in feeding, mating,
migration, and other behaviors. These and other factors explain why a definitive avian
exposure model is not currently available.
Environmental exposure has two components: the frequency and duration of
contact with the pesticide; and, the amount or concentration of a pesticide in the
environment and available to non-target organisms. The Comparative Analysis of
Acute Avian Risk from Granular Pesticides provided an in-depth discussion showing
that birds are present in fields treated with pesticides; that the pesticide is available to
birds in the fields; and, birds can and do ingest pesticide granules, contaminated plant
material, insects, and soil.
Only limited data are currently available to determine to what extent ingestion of
pesticide granules or food items with pesticide residues is incidental, accidental,
selected for, avoided or some combination of these possibilities. Birds may
inadvertently ingest granules along with other material, may mistake the granules for
seeds, grit, or other food items, or may actively select or avoid contaminated insects,
plant material or pesticide granules. With accidental or incidental exposure, both
dietary consumption and oral ingestion are assumed to be proportional to availability.
Since the amount of pesticide actually consumed or ingested by birds is difficult
to quantify, the Agency used two simple exposure models to estimate exposure in
terms of availability of the pesticide active ingredient: one to estimate avian acute oral
dose exposure, and one to estimate avian dietary exposure, both acute and chronic.
1. Acute Exposure to Granular Pesticides via Oral Dose (mg a.i./ ft2
available)
For granular pesticides, a simple exposure model for avian oral dose exposure
assumes that the amount of toxicant available to birds per unit area of the treated field
provides an indication of the actual amount of pesticide available that birds could
ingest. It is important to note that the Agency is not attempting to estimate the actual
number of birds that would receive a lethal dose, nor the probability of a given bird
consuming a lethal dose. Estimates of that sort would depend on the number of acres
treated, the species and numbers of birds present in a given area and many factors of
bird behavior, that have not yet been adequately documented.
-------
16
Methods and timing of applications vary with the specific product, the crop, and
reason for treatment. Further, preferred application methods also vary with crop and
location. Though some application-incorporation regimes are more effective than others
at reducing exposure, wildlife exposure to pesticides can result from all pesticide
application methods including ground spray, aerial spray, band, in-furrow, drill,
shanked-in, broadcast, side-dress and aerial broadcast.
For the purposes of this analysis, the Agency assumed that applications
requiring soil incorporation of the pesticide would result in only 15% of the pesticide
being available to birds. For in-furrow applications the Agency assumed that only 1% of
the pesticide would be available to birds. If labels did not specify any incorporation, no
reduction in exposure was calculated. Further, the Agency calculated the exposure
using the maximum application rate and one-half that rate. The latter was included to
provide a range of exposure estimates. An exposure at one-half the maximum is not
intended to represent any particular label rate. Since this is a screening analysis, it is
assumed that information on typical rates would not be readily available for all the
pesticides included in the comparison.
The Agency is using 15% and 1% as a representative values, recognizing that
specific application methods provide more or less efficient incorporation. In previous
product-specific assessments, the Agency has used a range of incorporation efficiency
values, reflecting the range of application methods. Erbach and Tollefson [7] and other
published data document the efficiency of various incorporation methods.
Field study and incident data confirm that birds can and do consume sufficient
amounts of the pesticide formulations examined in this analysis to cause mortality.
Furthermore, multiple lethal doses are readily available to birds in the relatively small
area of one square foot.
Birds may ingest pesticide granules or food items contaminated by pesticides
remaining on or just below the soil surface after a pesticide application. These
granules or contaminated food items may be consumed while a bird is foraging for
seed, grit or insects on the surface or probing below the surface of the soil.
Furthermore, subsurface granules and contaminated food items may also have
exposure potential via routes other than direct ingestion (e.g., dermal exposure via
contaminated water after irrigation or rainfall). Data are not available to estimate the
amount of pesticide ingested by birds probing below the soil surface. Therefore, for this
analysis, the Agency has considered only the amount of pesticide on the surface of the
soil after a pesticide application. See equation (1).
(1) #Pounds Active Ingredient/Acre x 453,600 mq/lbs = #mq Active Ingredient
43,560 ft2/Acre ft2
-------
17
which equals
#Pounds Active Ingredient/Acre x 10.4132 mq/lbs = #mq Active Ingredient
ft2/Acre ft2
This analysis, like the Comparative Analysis of Acute Avian Risk from Granular
Pesticides, uses one square foot as the unit for calculating toxicant availability,
although any constant unit area could be used. DeWitt [8] suggested this unit for
calculating environmental exposure when he related quantities of toxic pesticides
ingested by birds to quantities of toxic pesticide deposited per square foot using
several laboratory and field studies. Felthousen [9] proposed Agency risk criteria for
granular pesticides related to the amount of toxic pesticide per square foot available to
an animal. Current EPA ecological risk assessment procedures for pesticides use a
similar approach for determining the amount of toxicant available [10 and 11 ].
Appendix 6 gives the results of the calculations of amount of toxicant available
on the major use sites in terms of milligrams per square foot.
2. Acute Exposure to Sprayed Pesticides via Diet (ppm available in
diet)
In the Standard Evaluation Procedures for Ecological Risk Assessment [10,
Table 5], EPA presented a generalized table for estimating pesticide residues on avian
food items based on the data compiled by Hoerger and Kenaga [12], The pesticide
residues in the table (all 0-day residues for 1 lb a.i./acre application) have been used to
estimate maximum residues likely to be found in avian diets such as 240 ppm for small
grasses, estimate ranges from maximum to typical such as 240 to 125 for small
grasses or 58 to 33 ppm for forage crops, and to estimate residues in diets of specific
species. These estimates have been recently updated based on Fletcher et al [13],
These estimates, ranging from maximum to average, are 240 to 85 ppm for small
grasses, 110 to 36 ppm for long grasses, 135 to 45 ppm for broadleaf plants, and 15 to
7 ppm for fruits.
Since this is a screening analysis, the Agency chose to keep the acute exposure
simple and use the maximum residue value for small grasses adjusted by the maximum
single application rate on the label and one-half this value for comparison purposes.
This residue value was input to the FATE Model as well as the aerobic soil metabolism
half-life value (see Table 3). The aerobic soil metabolism half-life value is used here as
an estimate of foliar dissipation14. Run for 30-days, the model out-put provided the
14The aerobic soil metabolism (ASM) half-life value is almost always a conservative estimate of
foliar dissipation. For many pesticides, a more refined estimate may be found in Willis and McDowell [14],
ASM values greater than 30 days are very rare. Where the ASM value is greater than 30-days, the data in
-------
18
maximum and average estimated residues on avian food items in ppm. Appendix 7
shows the results for spray formulations.
this reference may provide a better estimate of foliar dissipation. Since this is a screening analysis, only the
ASM value was used to estimate the foliar dissipation.
-------
19
Table 3. Environmental Fate Parameters for Inputs into Terrestrial FATE Model & GENEEC Surface Water Model
Chemical
Name
Water
Solubility
Hydrolysis
Half-life
(days @ pH 7)
Photolysis
Half-life
(days)
Aerobic Soil
Metabolism
(days)
Anaerobic Soil
Metabolism
(days)
Aerobic Aquatic
Metabolism
(days)
GENEEC
Koc
A
80.1 g/l
163
stable
2.3
NA
NA
2.73
B
25.1 mg/l
37
3.2
95.6
NA
NA
725
C
2 mg/l
72
29.6
180
NA
NA
3680
D
32000
mg/l
38
175
27
NA
54
10
E
15 mg/l
323
3.87
19.39
NA
NA
386
F
843 mg/l
stable
stable
300
300
NA
108
G
24 mg/l
stable
30
174
NA
5.2
232
H
400 mg/l
706
0.218
13.29
NA
NA
106
I
145 mg/l
3.2
94
3
NA
3.3
151
J
200 g/l
27
90
1.75
NA
NA
D.88
K
250 mg/l
48
11
9
NA
NA
113
L
2000 mg/l
D.64
stable
3
NA
NA
39
M
miscible
40
137
9.6
10.5
NA
2
N
60 mg/l
40
2.04
11.25
NA
NA
230
0
50 mg/l
3
1
9
96
NA
150
P
25 mg/l
D.39
Stable
9
45
NA
1260
Q
15 mg/l
15
1.2
81
216
NA
333
-------
20
3. Chronic Exposure via Diet (ppm available in diet)
The Agency has noted that the chronic exposure is the weakest point in the
avian risk assessment [10], It noted that Hoerger and Kenaga [12] even adjusted by
Fletcher et al [13] data are of minimal value since the values presented are generally
those found immediately after application. Further, in the past, the residues likely to be
found over time have been estimated on a case-by-case basis and used for the chronic
avian EEC's.
Fletcher et al. [13] looked at pesticide persistence by examining residue-decay
curves of pesticides administered at rates between 0.5 and 1.5 lb/acre. All the data fit
exponential decay curves except systemic pesticides applied as either granules or
dust. For such pesticides, "no apparent exponential decay curve occurred over the first
30 to 40 days." The residues that remained were generally "below the 0-day levels
predicted by the Kenaga nomogram." More research is needed to expand this
prediction.
Further, the findings of Rattner et al. [15], Bennett and Bennett, [16], and
Bennett et al. [17] have shown that pesticide effects on avian reproduction for some
pesticides are not simply a function of chronic exposure. They found that exposure of
breeding bobwhite quail and mallard ducks to organophosphate compounds can
negatively impact reproduction with exposure periods as short as 8-10 days. Again,
this research needs to be expanded to more accurately predict when short exposure
periods can lead to reproductive impairment.
Considering all of the above, the Agency chose to use the acute exposure EEC's
as estimates in Appendix 7 to be compared with chronic test endpoints for this analysis
of 17 insecticides.
III. AVIAN RISK QUOTIENTS AND ECOLOGICAL LEVELS OF CONCERN
The Agency currently uses the quotient method to express ecological risk. The
quotient method compares the estimated environmental concentration of a pesticide to
the toxicity test effect level for a given species. The result is a risk quotient (RQ). An
RQ is calculated by dividing an appropriate exposure estimate (e.g. EEC) by an
appropriate toxicity test effect level (e.g. LC50, LD50, NOAEC). We assume that the
higher the specific risk quotient for an endpoint, generally, the greater the relative risk.
Equation (2) is a general statement of this relationship:
(2) Estimated Environmental Concentration (e.g.EEC) = Risk Quotient
Toxicity Test Effect Level (e.g., LC50)
-------
21
The risk quotients are intended to be used as rough indicators of comparative
risk, and cannot be used to predict how many birds will actually die or experience
impaired reproduction. Further, they are not intended to predict the probability of a bird
receiving a lethal or chronic dose. Site-specific considerations such as the
attractiveness of the treated fields, the species distribution, the species density, as well
as the number of acres treated would affect the number of these organisms actually
exposed.
Furthermore, the quotient does not provide a definitive value for the amount of
pesticide that will be available to birds. The actual amount of pesticide available will
vary depending on the application method, configuration and calibration of equipment,
wind speed and other field conditions.
In order to provide industry and the public with clear standards for ecological risk
assessment and management that can be applied in an equitable fashion and to
facilitate ecological risk comparisons, the Agency established levels of concern (LOC's)
for ecological effects of pesticides on non-target organisms [2], These LOC's are
criteria used by the Agency to indicate potential risk to non-target organisms and the
need for a regulatory action. If the criteria are exceeded, it indicates that a pesticide,
when used as directed, has the potential to cause adverse effects on non-target
organisms.
There are two general categories of LOC's for avian species, acute and chronic.
In order to determine if an LOC has been exceeded, first a risk quotient must be
calculated and then compared to the appropriate LOC. When the risk quotient exceeds
the LOC for a particular endpoint, risk for that endpoint is presumed to exist. The
LOC's for birds used in this analysis plus the corresponding risk presumptions are as
follows:
ENDPOINT LOC PRESUMPTION
Acute Dietary RQ> 0.5 High Acute Risk
Acute Oral Dose RQ> 0.513 High Acute Risk
Chronic RQ> 1.0 High Chronic Risk
The specific equations used to calculate the risk quotients for acute and chronic
13 In 1992, the Agency announced the selection of 1 LD50/ft2 as the cutoff level of concern based upon
field study data submitted to the Agency at that time which indicated that pesticide applications resulting in
environmental concentrations of at least 1 LD50/ft2 have resulted in avian mortality. Since that time, the
Agency proposed changing the level of concern (LOC) to 0.5 LD50/ft2 primarily to add a level of safety to the
risk estimate [2], This proposal has generated considerable discussion both within and outside the Agency.
-------
22
risk to birds used in this analysis follow. Calculations using these equations were
conducted for the 17 pesticides used on the seven use sites. Maximum and average
exposure values as well as one-half these residues were used as numerators; the
lowest toxicity value and the median toxicity value (where calculated) were used as
denominators. Avian acute (LD50/ft2) RQs calculations were limited to granular
formulations typically applied pre- or at-plant, while avian dietary (EEC/LC50) RQs,
avian acute bird per day (EEC x %Food Ingestion per Day/ LD50) RQs, and avian
chronic (EEC/NOAEC) RQs were limited to spray formulations, primarily applied post-
emergent. Twelve acute avian dose risk quotients were calculated for each granular
pesticide/use combination; four avian dietary risk quotients were calculated for each
spray pesticide/use combination; twenty-four acute bird per day risk quotients were
calculated for each spray pesticide/use combination; and, four avian chronic risk
quotients were calculated for each spray pesticide/use combination.
A. Calculation of the Acute Avian Risk Quotients
In U.S. EPA 1986 [10], the avian dietary LC50 was presented as the primary
acute toxicological endpoint to be compared to the acute exposure. However, Hill [18]
points out that "ingestion is believed to be the most common route of pesticidal
exposure in birds and therefore th[e] oral tests of lethality [LD50 ] provide a sound basis
for preliminary screening." Further, he states that "when used in combination and
judiciously, the two tests of lethality are invaluable tools for preliminary evaluation of
potential hazard of pesticides to wild birds."
The Agency chose to estimate acute risk to birds for these pesticides by using
both the avian acute oral LD50 test and the avian dietary LC50 test in the risk
assessment.
1. Avian Acute Risk via Dose Ingestion - Granular Formulations
The avian acute risk via dose ingestion is calculated for granular formulations
using equation (3).
(3) EEC (ma ai/ft2) = # of LD^s = RQ (Risk Quotient)
LD50 (mg/kg) x Bird Weight (kg) ft2
This equation describes acute avian risk as a quotient of the amount of toxicant readily
available to birds within a square foot (a rough indicator of exposure) of treated area, to
the avian acute oral toxicity (expressed as an LD50 per bird). General bird body
weights used were 1.000 kg (1000 gm) to represent large birds such as mallard ducks
and pheasants, 0.100 kg (100 gm) to represent medium size birds such as doves and
quail, 0.020 kg (20 gm) to represent small birds such as songbirds. The result is an
-------
23
expression of acute risk to birds in terms of the number of LD50s per square foot.
Appendix 9 lists all acute avian risk quotient calculations for the granular
formulation pesticide use site combinations. The results are presented in order of
descending risk quotients for the pesticides on each use site. The greater the number,
the greater the potential acute dose risk to birds.
2. Avian Dietary Risk - Spray Formulations
The avian acute risk via the diet is calculated for spray formulations using
equation (4).
(4) EEC (ppm in the diet) = Risk Quotient
LC50 (ppm)
This equation describes acute avian dietary risk as a quotient of the concentration of
toxicant likely to be available in bird diets, to the subacute avian dietary toxicity
(expressed as an LC50). The result is an expression of acute risk to birds in terms of
concentration exposed to concentration tested.
Appendix 10 lists all avian dietary risk quotient calculations for the spray
formulation pesticide use site combinations. The results are presented in order of
descending risk quotients for the pesticides used on each use site. The greater the
number, the greater the potential acute dietary risk to birds.
3. Avian Acute Bird per Day Risk - Spray Formulations
The avian acute bird per day ingestion risk is also calculated for spray
formulations using equations (5) and (6).
(5 ) EEC x (Food lnqestion14/bird weight) = Risk Quotient
ld50
This equation describes acute avian bird per day ingestion risk as a quotient of the
14 Bird food ingestion rates (in grams dry matter per day) were calculated using an equation
developed by Nagy [19] and referenced by EPA [20],
Food Ingestion (g/day) = 0.648 x bird weight 0651 (g)
The Agency selected the general equation for all birds over other more specific equations for passerines,
non-passerines, and seabirds. We assumed that the lowest and median toxicity values used in the risk
quotients represented all birds. Thus, we chose the generalized food ingestion rate for all birds.
-------
24
quantity of toxicant likely to be ingested daily [20] by a bird, to the acute oral avian
dose toxicity (expressed as an LD50 expressed as mg/kg). The result is an expression
of acute risk to birds in terms of daily acute dose from ingestion of contaminated food
items.
As previously noted, EPA/OPP has found that the LD50 value is often a better
indicator of acute toxicity to birds especially for pesticides with acute LD50 values less
than or equal to 50 mg/kg. Appendix 3 shows that 14 out of the 17 Chemicals
considered here have LD50 values less than 50 mg/kg. Also, comparing the toxicity
rankings in Appendix 3 and 4, a number of the most toxic insecticides via the oral dose
are less toxic via the diet, e.g., Chemical 0, Chemical G, Chemical E. Thus, the Agency
decided that a risk quotient using the LD50 toxicity values should be included for spray
formulations.
Appendix 11 lists all avian acute bird per day ingestion risk quotient calculations
for the spray formulation pesticide use site combinations. The results are presented in
order of descending risk quotients for the pesticides used on each use site. The
greater the number, the greater the potential acute bird per day risk to birds.
B. Calculation of the Chronic Avian Risk Quotients
1. Avian Chronic Risk - Spray Formulations
The avian chronic quotient is calculated for spray formulations using equation
(6).
(6) EEC (ppm in the diet) = Risk Quotient
NOAEC (ppm)
This equation describes chronic risk to birds as a quotient of the concentration of
toxicant likely to be available in bird diets, to the no observed adverse effect
concentration (NOAEC) in the avian reproduction test. The result is an expression of
chronic risk to birds in terms of concentration exposed to concentration tested.
Appendix 12 contains a list of avian chronic risk quotient calculations for all
spray formulation pesticide use site combinations. The results are presented in order of
descending risk quotients for the pesticides used on each use site. The greater the
number, the greater the potential avian chronic risk to birds.
Granular formulations may also present a chronic risk to birds. However, the
Agency dose not presently have a method to evaluate this risk.
-------
25
2. Number of Days to Reach the Avian Chronic Level of Concern
(RQ=1) - Spray Formulations
In addition to using an RQ approach for estimating potential avian chronic risk,
the Agency elected to use a risk index that better reflected pesticide persistence. This
was accomplished by calculating the total number of days post-application required for
estimated pesticide concentrations in avian food items to be degraded/dispersed to a
point of equivalence with the avian long-term exposure toxicity endpoint (reproduction
NOAEC)15.
The Avian Chronic Time to RQ=1 calculation was performed for spray
formulations using the equation (7):
(7) Ln ((NOAEC (ppmVEEC (ppm in the diet)) = Time (days)
-K
where, the EEC (starting food item concentration) is the Agency standard short-grass
estimated concentration based on 1 lb a.i./A (240 ppm) [13] adjusted by the application
rate on the label. K is the foliar degradation rate constant as estimated by aerobic soil
metabolism half-life and an assumption of first-order degradation kinetics (See
Appendix 8).
Appendix 13 contains a list of avian chronic Time to RQ=1 calculations for all
spray formulation pesticide use site combinations. The results are presented in order of
descending number of days for the pesticides used on each use site. The greater the
number of days, the greater the potential avian chronic risk to birds.
C. The Avian Risk Column - % Pesticide Contribution to RQ Sum for Each
Endpoint on Each Site (Crop)
When comparing pesticides using RQs, the RQ scales for each endpoint are an
important consideration. As shown below, the scales for the four avian endpoints vary
widely (pesticide and use site for the highest values are presented parenthetically):
Acute Dose (LD50/ft2) 0 to 2519 [Chemical O on Peanuts]
Acute Dietary (EEC/LC50) 0 to 131 [Chemical L on Cotton]
Acute Bird per Day (EECx%FI/LD50) 0 to 2940 [Chemical G on Cotton]
Chronic Bird (EEC/NOAEC) 0 to 206 [Chemical B on Cotton]
15 Note that the Agency is not suggesting that avian exposures occurring after this time are
inconsequential; only that new exposures starting after that point are not expected to present significant risk
to birds.
-------
26
These differences raise many questions. For example: Should we be more concerned
with acute bird per day risk and avian acute dose risk than acute dietary risk and
chronic bird risk? Is the magnitude of the avian acute dose risk approximately 24 times
greater than the dietary risk for pesticides used on potatoes? Do these differences
reflect inherently different ranges of toxicities for the different endpoints? Additional
analyses and perhaps research is needed to answer these questions. Not having
adequate answers at present, the Agency decided that it would be easier to compare
avian risk between pesticides without having to deal with these scale differences.
With the above information in mind, the Agency focused on each crop site. It
assumed that all the potential risk for a particular endpoint on a particular crop site
could be represented by the sum of the RQ values that exceeded the LOC for all the
pesticides used on that crop site. If the RQ values for that endpoint for each pesticide
used on that crop site were summed, then the quotient of the sum of the individual
pesticide RQs over the sum of all the RQs for all the pesticides, would represent the
percent (%) contribution of each pesticide to the total risk for that endpoint on that site.
See equation (8).
(8) Vendpoint RQ values/pesticide/crop site x 100 = % Pesticide Contribution to RQ
£ endpoint RQ values for all pesticides/crop site Sum on Each Crop Site
This calculation provides a relative estimate of potential avian risk per pesticide
per avian endpoint by which the pesticides can be compared on a crop site basis. The
comparison is relative to the total risk for each endpoint and on each crop, represented
by the sum of the RQ values which exceed the LOCs. It eliminates the problem of
widely varying scales. Appendices 9a, 10a, 11a and 12a show these calculations for
avian acute dose risk, avian dietary risk, acute bird per day risk and avian chronic risk,
respectively.
If all the avian risk per endpoint can be represented by the sum of all the RQ
values exceeding the LOC for that endpoint, the total risk can be viewed as column.
Each pesticide used on that site contributes a certain percentage toward filling the
column. The major contributors to the total risk can be determined by showing the
individual percentages (See example in Figure 1).
-------
27
Figure 1. Avian Risk
Acute Bird per Day Risk on Alfalfa
Others l0A0ofc
Chemcial D (0.60^
Chemical K <0.90%^
Chemical L (0.90%^
Chemical C (3.90
M
Chemical G (93.30%i
-------
28
In this example, Chemical G contributes the greatest and a majority of the
potential acute bird per day risk on alfalfa. Chemical C is a distant second, while the
other pesticides used on alfalfa contribute less than 1%.
D. Frequency of LOC Exceedance (%)
Where the risk quotients can provide an estimate of the magnitude of potential
avian risk, considering how often a risk quotient exceeds an LOC can provide an
estimate of the frequency of the potential risk. Twelve avian acute dose risk quotients
were calculated for each granular pesticide/use combination (See Appendix 9); four
avian dietary risk quotients were calculated for each spray pesticide/use combination
(See Appendix 10); twenty-four acute bird per day risk quotients were calculated for
each spray pesticide/use combination (See Appendix 11); and, four avian chronic risk
quotients were calculated for each spray pesticide/use combination (See Appendix 12).
The lowest and median toxicity values were included in all calculations except the
avian chronic, where the lowest value was the only value available. The maximum
residue values and one-half these values were included in all calculations. In the bird
per day RQ calculations and the avian chronic calculations, the average residue
values, as determined using the FATE model were also included. Both the avian acute
dose and the acute bird per day RQ calculations included LD50 values adjusted for 20,
100, and 1000 gram birds.
Appendices 9b, 10b, and 11b show how often (%) the calculated acute risk
quotients exceeded the LOC's for pesticide/use combination and each endpoint
analyzed. The pesticides in each appendix were ordered by crop site and by
decreasing summed RQ values exceeding the LOC. The higher the frequency (%), the
more times the RQ's exceeded the LOC's. This shows the relative frequency with which
a pesticide/use combination is likely to exceed an LOC. The frequency of exceedance
was not calculated for avian chronic risk because (1) greater than 93% [9/124; see
Appendix 12] of the chronic RQ calculations exceeded the LOC for chronic avian risk,
and (2) the Time to RQ=1 (# of days) calculation along with the RQ calculation were
thought to be risk indices that together better reflected pesticide persistence.
-------
29
IV. EFFECTS AND EXPOSURE CHARACTERIZATION FOR AQUATIC
ORGANISMS
A. Effects
EPA typically reviews the following laboratory studies in performing aquatic risk
assessments: acute freshwater LC50 (ppb) studies, acute marine/estuarine fish LC50
(ppb) studies, acute freshwater invertebrate EC50 (ppb) studies, marine/estuarine
crustacean EC50 (ppb) studies, marine/estuarine mollusc EC50 (ppb) freshwater fish
chronic (partial fish life-cycle) study providing a NOAEC (ppm), and freshwater
invertebrate life-cycle providing a NOAEC. The Agency evaluates the studies and
classifies them as either core, supplemental or invalid, as well as indicating whether the
supplemental and invalid studies are up gradable. The results of the core and
supplemental studies, the toxicity values, are used in risk assessment.
Core and supplemental eco-toxicity data in the EcoTox Data Base [9] were used
to characterize the effects on fish and aquatic invertebrates. EPA updated and verified
all the data used in this analysis. However, some errors in this analysis could result
from entry errors. In addition, the data used in this analysis reflects the status of the
data base as of July, 1998. It does not include data entered after that date. More
recent data could change the results of this analysis.
Freshwater fish LC50 (ppb), marine/estuarine fish LC50 (ppb), freshwater
invertebrate EC50 (ppb), marine/estuarine crustacean EC50 (ppb), marine/estuarine
mollusc EC50 (ppb) acute values as well as freshwater fish and aquatic invertebrate life-
cycle NOAEC (ppb) values for the most sensitive species tested (the lowest values)
were selected from the data base. In addition, the median fish and aquatic invertebrate
LC50 (ppb) and EC50 (ppb) values were calculated. These median values provided a
less conservative estimate of the acute toxicity (compared to the lowest) and provided
additional values for determining a range of toxicity values for a particular endpoint.
Since relatively few species are used in standard toxicity testing, it is likely that
the species most sensitive to each pesticide has not been tested. Because of this
variation in sensitivity, it is unlikely that this analysis will show the worst case risk for
each pesticide, but rather will provide a range of risk values for purposes of comparison
and identification of those pesticides that are more likely to cause adverse effects in
actual use.
1. Acute Toxicity to Freshwater Fish
EPA typically requires 96-hour acute LC50 toxicity studies on two fish species,
one cold water fish such as a rainbow trout, and one warm water fish such as a bluegill
sunfish. These toxicity data are used to assess the pesticide's potential to cause acute
-------
30
lethality in freshwater fish.
EPA ranked the 17 pesticides in order of their lowest acute freshwater fish LC50
toxicity values (Appendix 14). The median values were also calculated.
2. Acute Toxicity to Marine/Estuarine Fish
EPA typically requires 96-hour acute LC50 toxicity studies on one
marine/estuarine fish species such as the sheepshead minnow when the use of the
pesticide is likely to contaminate marine/estuarine environments. These toxicity data
are used to assess acute effects on marine/estuarine fish. For pesticides lacking data
on this endpoint, a value was estimated based on a regression analysis (Table 2) of
median freshwater fish acute values over median marine/estuarine fish for all the
pesticides in the class. Median values are not presented for this endpoint because EPA
typically receives data on only one marine/estuarine fish species.
EPA ranked the 17 pesticides in order of their lowest acute marine/estuarine fish
LC50 toxicity values (Appendix 15).
3. Acute Toxicity to Freshwater Invertebrates
EPA typically requires one 48-hour EC50 study on daphnia spp. These data are
used by OPP to assess acute effects on freshwater invertebrates.
EPA ranked the 17 pesticides in order of their lowest acute EC50 toxicity values
(Appendix 16). The median value is also included since there were sufficient data for
each pesticide in the analysis to calculate this value.
4. Acute Toxicity to Marine/Estuarine Crustaceans
EPA typically requires 96-hour acute EC50 toxicity studies on one
marine/estuarine crustacan species such as the mysid when the use of the pesticide is
likely to contaminate marine/estuarine environments. These toxicity data are used to
assess acute effects on marine/estuarine crustaceans. For pesticides lacking data on
this endpoint, a value was estimated based on a regression analysis (Table 2) of
median freshwater crustacean acute values over median marine/estuarine crustacean
values for all the pesticides in this class. Median values are not presented for this
endpoint because EPA typically receives data on only one marine/estuarine crustacean
species.
EPA ranked the 17 pesticides in order of their lowest acute marine/estuarine
crustacean EC50 toxicity values (Appendix 17).
5. Acute Toxicity to Marine/Estuarine Molluscs
-------
31
EPA typically requires 96-hour acute EC50 toxicity studies on one
marine/estuarine mollusc species such as the eastern oyster when the use of the
pesticide is likely to contaminate marine/estuarine environments. These toxicity data
are used to assess acute effects on marine/estuarine molluscs. For pesticides lacking
data on this endpoint, a value was estimated based on a regression analysis (Table 2)
of median marine/estuarine crustacean acute values over median marine/estuarine
mollusc values for all the pesticides in this class. Median values are not presented for
this endpoint because EPA typically receives data on only one marine/estuarine
crustacean species.
EPA ranked the 17 pesticides in order of their lowest acute marine/estuarine
mollusc EC50 toxicity values (Appendix 18).
6. Chronic Toxicity to Freshwater Fish
The NOAEC is the typical value resulting from the partial or full life-cycle fish
test. Typically, one species is tested, often the fathead minnow. Common life-cycle
effects found in these tests are reduced hatchability, reduced juvenile survival, reduced
growth of Fgeneration, etc. For pesticides lacking data on this endpoint, a value was
estimated based on a regression analysis (Table 2) of median freshwater acute fish
values over freshwater fish chronic values for all the pesticides in this class.
EPA ranked the 17 pesticides in order of their lowest chronic toxicity NOAECs
(Appendix 19).
7. Chronic Toxicity to Freshwater Invertebrates (crustaceans)
The NOAEC is the typical value resulting from the aquatic invertebrate life-
cycle test requirement in the regulations. Typically one species such as Daphnia, spp.
is tested. Common life-cycle effects found in these tests are reduced number of young
per female, reduced juvenile survival, reduced growth of Fgeneration, etc. For
pesticides lacking data on this endpoint, a value was estimated based on a regression
analysis (Table 2) of median freshwater crustacean acute values over freshwater
invertebrate chronic values for all the pesticides in this class.
EPA ranked the 17 pesticides in order of their lowest chronic toxicity NOAECs
(Appendix 20).
-------
32
B. Exposure
1. Acute and Chronic Exposure Modeling
To provide a basis for comparison, EPA used the GENEEC [21] to estimate the
concentration of the 17 pesticides in ponds adjacent to pesticide applications on the
four use sites using both the maximum use rates and one-half the maximum use rates.
The GENEEC is a screening model that mimics the PRZM-EXAMS model behavior. In
the model, the number of days between treatment and rain-induced runoff is set at 2. It
assumes runoff from a 10-hectare field to a standard 1 hectare pond two meters deep.
Further, it assumes 10% runoff of a total annual pesticide application. A Generic
Estimated Environmental Concentration (GEEC) is produced and this value may be
increased by adding spray drift. Spray drift for an aerial application is added at 5%
application rate with 95% application efficiency. Spray drift for a ground application is
added at 1% application rate with 99% application efficiency. The GEEC may be
reduced by factoring in adsorption to soil using the Koc value and by considering
incorporation. The model calculates chronic GEEC's using aerobic aquatic, hydrolysis
and/or aquatic photolysis half-life values (days). The model produces a report
consisting of the peak GEEC as well as the average GEEC at 4-days, 21-days and 56-
days. Following the standard procedures used in EFED Science Chapters for
Registration Eligibility Documents (REDs), EPA chose to use the peak GEEC for the
acute exposure, i.e., the Estimated Environmental Concentration (EEC), to fish and
aquatic invertebrates; the 21-day GEEC for the chronic EEC to aquatic invertebrates;
and, the 56-day GEEC for the chronic EEC to fish. Table 3 in Section II provides a
listing of the environmental fate parameters for each of the 17 pesticides used to load
the GENEEC model. Table 4 below provides the report format for the model.
Appendix 21 gives the results of the GENEEC model runs for the 38
pesticide/use combinations for the maximum label rates and one-half the maximum
label rates.
-------
33
Table 4. GENEEC Model Report Format
RUN No. 1 FOR [name of pesticide] INPUT VALUES
RATE (#/AC) APPLICATIONS SOIL SOLUBILITY % SPRAY INCORP
ONE(MULT) NO.-INTERVAL KOC (PPM) DRIFT DEPTH(IN)
0.0( 0.000) 0 0 0.0 000.0 0 .0 0.0
FIELD AND STANDARD POND HALFLIFE VALUES (DAYS)
METABOLIC DAYS UNTIL HYDROLYSIS PHOTOLYSIS METABOLIC COMBINED
(FIELD) RAIN/RUNOFF (POND) (POND-EFF) (POND) (POND)
0.00 0 000.00 .00- .00 .00 000.00
GENERIC EECs (IN PPB)
PEAK AVERAGE 4 AVERAGE 21 AVERAGE 56
GEEC DAY GEEC DAY GEEC DAY GEEC
00.00
00.00
00.00
00.00
-------
34
V. AQUATIC RISK QUOTIENTS AND LEVELS OF CONCERN
As noted previously, the Agency currently uses the quotient method to express
ecological risk. The quotient method compares the estimated environmental
concentration of a chemical to the toxicity test effect level for a given species. The
result is a risk quotient. A risk quotient is calculated by dividing an appropriate
exposure estimate (e.g. EEC) by an appropriate toxicity test effect level (e.g. LC50). We
assume that the higher the specific risk quotient, the greater the relative risk. Once
again equation (2) is presented as a general statement of this relationship:
Estimated Environmental Concentration (e.g.EEC) = Risk Quotient
Toxicity Test Effect Level (e.g., LC50)
The acute aquatic effect levels typically are: LC50 for fish, and the EC50 for
aquatic invertebrates. The aquatic chronic effect level for both fish and aquatic
invertebrates is the NOAEC.
The risk quotients are intended to be used as rough indicators of comparative
risk, and cannot be used to predict how many fish or aquatic invertebrates will actually
die or experience adverse effects on their life-cycle or reproduction. Further, the risk
quotients are not intended to predict the probability of a fish or aquatic invertebrate
receiving a lethal dose. Site-specific considerations such the water temperature,
quality and pH, vagaries of weather especially precipitation, the species distribution,
the species density, as well as the number of acres treated, would affect the number of
organisms actually exposed, the concentrations and durations of the exposures, and
their consequences. .
Furthermore, the quotient does not provide a definitive value for the amount of
pesticide that will be available to fish or aquatic invertebrates. The actual amount of
pesticide available will vary depending on application method, configuration and
calibration of equipment, and specific field conditions.
In order to provide industry and the public with clear standards for ecological risk
assessment and management that can be applied in an equitable fashion and to
facilitate ecological risk comparisons, the Agency established levels of concern (LOC's)
for ecological effects of pesticides on fish and aquatic invertebrates [2], These LOC's
are criteria used by the Agency to indicate potential risk to non-target organisms and
the need for a regulatory action. Exceeding the criteria indicates that a pesticide, when
used as directed, has the potential to cause undesirable effects on non-target fish and
aquatic invertebrates.
There are two general categories of LOC's for fish and aquatic invertebrates,
-------
35
acute and chronic. In order to determine if an LOC has been exceeded, first a risk
quotient must be calculated and then compared to the appropriate LOC. When the risk
quotient exceeds the LOC for a particular category, risk to that particular category is
presumed to exist. The LOC's for fish and aquatic invertebrates used in this analysis
plus the corresponding risk presumptions are as follows:
ENDPOINT LOC PRESUMPTION
Acute RQ > 0.5 High Acute Risk
Chronic RQ > 1.0 High Chronic Risk
The specific equations used to calculate the risk quotients for acute and chronic
risk to fish and aquatic invertebrates used in this analysis follow. Calculations using
these equations were conducted for all of the pesticides use site combinations.
Exposure values were modeled using the maximum and one-half the maximum use
rates. Five acute RQs (acute freshwater fish RQs (EEC/LC50), acute marine/estuarine
fish RQs (EEC/LC50), acute freshwater invertebrate RQs (EEC/EC50), acute
marine/estuarine crustacean RQs (EEC/EC50), and acute marine/estuarine mollusc
RQs (EEC/EC50)), and two chronic RQs (chronic freshwater fish and chronic freshwater
invertebrate EEC/NOAEC)) were calculated for both granular and spray formulations.
Four quotients were calculated for each pesticide use site combination for acute
freshwater fish and aquatic invertebrate endpoints, reflecting the combinations of
maximum and one-half the maximum use rates with the lowest and median LC50 or EC50
values. Two quotients were calculated for the other endpoints, reflecting the two use
rate assumptions and the lowest LC50 or EC50 values.
A. Calculation of the Acute Fish & Aquatic Invertebrate Risk Quotients
1. Acute Freshwater Fish Risk
The freshwater fish acute risk quotient is calculated using equation (9).
(9) EEC (ppb in pond water; peak) = Risk Quotient
LC50 (ppb)
This equation describes acute freshwater fish risk as a quotient of the concentration of
toxicant likely to occur in ponds adjacent to pesticide applications as estimated by the
GEN EEC model, to the acute freshwater fish toxicity value (expressed as an LC50). The
-------
36
result is an expression of acute risk to freshwater fish in terms of concentration
exposed to concentration tested.
Appendix 22 lists the freshwater fish risk quotient calculations for all pesticide
use site combinations. The results are presented in order of descending risk quotients
by crop.
2. Acute Marine/Estuarine Fish Risk
The marine/estuarine fish acute risk quotient is calculated using equation (9).
This equation describes acute marine/estuarine fish risk as a quotient of the
concentration of toxicant likely to occur in estuarine areas adjacent to pesticide
applications as estimated by the GENEEC model, to the acute marine/estuarine fish
toxicity value (expressed as an LC50). The pond values are used as a rough
approximation of pesticide concentrations in the estuarine environment. The result is
an expression of acute risk to marine/estuarine fish in terms of concentration exposed
to concentration tested.
Appendix 23 lists the marine/estuarine fish risk quotient calculations for all
pesticide use site combinations. The results are presented in order of descending risk
quotients by crop.
3. Acute Freshwater Invertebrate Risk
The freshwater invertebrate acute risk quotient is calculated using equation (10).
(10) EEC (ppb in pond water; peak) = Risk Quotient
EC50 (ppb)
This equation describes acute freshwater invertebrate risk as a quotient of the
concentration of toxicant likely to occur in ponds adjacent to pesticide applications as
estimated by the GENEEC model, to the acute freshwater invertebrate toxicity value
(expressed as an EC50). The result is an expression of acute risk to freshwater
invertebrate in terms of concentration exposed to concentration tested.
Appendix 24 lists the freshwater invertebrate risk quotient calculations for all
pesticide use site combinations. The results are presented in order of descending risk
quotients by crop.
4. Acute Marine/Estuarine Crustacean Risk
The marine/estuarine crustacean acute risk quotient is calculated using equation
-------
37
(10). This equation describes acute marine/estuarine crustacean risk as a quotient of
the concentration of toxicant likely to occur in estuarine areas adjacent to pesticide
applications as estimated by the GENEEC model, to the acute marine/estuarine
crustacean toxicity value (expressed as an EC50). The pond values are used as a rough
approximation of pesticide concentrations in the estuarine environment. The result is
an expression of acute risk to marine/estuarine crustaceans in terms of concentration
exposed to concentration tested.
Appendix 25 lists the marine/estuarine crustacean risk quotient calculations for
all pesticide use site combinations. The results are presented in order of descending
risk quotients by crop.
5. Acute Marine/Estuarine Mollusc Risk
The marine/estuarine mollusc acute risk quotient is calculated using equation
(11). This equation describes acute marine/estuarine mollusc risk as a quotient of the
concentration of toxicant likely to occur in estuarine areas adjacent to pesticide
applications as estimated by the GENEEC model, to the acute marine/estuarine
mollusc toxicity value (expressed as an EC50). The pond values are used as a rough
approximation of pesticide concentrations in the estuarine environment. The result is
an expression of acute risk to marine/estuarine molluscs in terms of concentration
exposed to concentration tested.
Appendix 26 lists the marine/estuarine mollusc risk quotient calculations for all
pesticide use site combinations. The results are presented in order of descending risk
quotients by crop.
B. Calculation of the Chronic Fish & Aquatic Invertebrate Risk Quotients
1. Chronic Freshwater Fish Risk
The freshwater fish chronic quotient is calculated using equation (11).
(11) EEC (ppb in pond water; 56-dav average) = Risk Quotient
NOAEC (ppb)
This equation describes chronic freshwater fish risk as a quotient of the concentration
of toxicant likely to occur in ponds adjacent to pesticide applications as estimated by
the GENEEC model, to the chronic freshwater fish toxicity value (expressed as an
NOAEC). The result is an expression of chronic risk to freshwater fish in terms of
concentration exposed to concentration tested.
-------
38
Appendix 27 lists all chronic freshwater fish risk quotient calculations for all
pesticide use site combinations.
2. Chronic Freshwater Invertebrate Risk
The freshwater invertebrate chronic quotient is calculated using equation (12).
(12) EEC (ppb in pond water; 21-day average) = Risk Quotient
NOAEC (ppb)
This equation describes chronic freshwater invertebrate risk as a quotient of the
concentration of toxicant likely to occur in ponds adjacent to pesticide applications as
estimated by the GENEEC model, to the chronic freshwater invertebrate toxicity value
(expressed as an NOAEC). The result is an expression of chronic risk to freshwater
invertebrates in terms of concentration exposed to concentration tested.
Appendix 28 lists all chronic freshwater invertebrate risk quotient calculations for
all pesticide use site combinations.
A. The Aquatic Risk Column - % Pesticide Contribution to RQ Sum for Each
Endpoint on Each Site (Crop)
When comparing pesticides using RQs, the RQ scales for each endpoint are an
important consideration. As shown below, the scales for the four aquatic endpoints vary
widely (pesticide and use site for the highest values are presented parenthetically):
Acute Freshwater Fish (EEC/LC50) 0 to 443 [Chemical
Acute Marine/Estuarine Fish (EEC/LC50) 0 to 50 [Chemical
Acute Freshwater Invertebrate (EEC/EC50) 0 to 4979 [Chemical
Acute Marine/Estuarine Crustacea (EEC/EC50) 0 to 830 [Chemical
Acute Marine/Estuarine Mollusc (EEC/EC50) 0 to 249 [Chemical
Chronic Freshwater Fish (EEC/NOAEC) 0 to 334 [Chemical
Chronic Freshwater Invertebrate (EEC/NOAEC) 0 to 22494 [Chemical
B on Cotton]
B on Cotton]
G on Cotton]
G on Cotton]
G on Cotton]
B on Cotton]
G on Cotton]
These differences raise many questions. For example: Should we be more concerned
with chronic and acute freshwater invertebrate risk than the others, especially acute
marine/estuarine fish risk? Is the magnitude of the chronic freshwater invertebrate risk
approximately 450 times greater than the acute marine/estuarine fish risk for pesticides
used on cotton? It would be easier to compare aquatic risk between pesticides without
having to deal with these scale differences.
With the above information in mind, the Agency focused on each crop site. It
-------
39
assumed that all the potential risk for a particular endpoint on a particular crop site
could be represented by the sum of the RQ values that exceeded the LOC for all the
pesticides used on that crop site. If the RQ values for that endpoint for each pesticide
used on that crop site were summed, then the quotient of the sum of the individual
pesticide RQs over the sum of all the RQs for all the pesticides, would represent the
percent (%) contribution of each pesticide to the total risk for that endpoint on that site.
See equation (8) again.
Vendpoint RQ values/pesticide/crop site x 100 = % Pesticide Contribution to RQ
£ endpoint RQ values for all pesticides/crop site Sum on Each Crop Site
This calculation provides a relative estimate of potential fish or aquatic
invertebrate risk per pesticide per aquatic endpoint by which the pesticides can be
compared on a crop site basis. The comparison is relative to the total risk for each
endpoint and on each crop, represented by the sum of the RQ values which exceed the
LOCs. It eliminates the problem of widely varying scales. Appendices 22a, 23a, 24a,
25a, 26a, 27a, and 28a, show these calculations for acute freshwater fish risk, acute
marine/estuarine fish risk, acute freshwater invertebrate risk, acute marine/estuarine
crustacean risk, acute marine/estuarine mollusc risk, chronic freshwater fish risk, and
chronic freshwater invertebrate risk, respectively.
If all the aquatic risk per endpoint can be represented by the sum of all the RQ
values exceeding the LOC for that endpoint, the total risk can be viewed as a column.
Each pesticide used on that site contributes a certain percentage toward filling the
column. The major contributors to the total risk can be determined by showing the
individual percentages (See example in Figure 2).
-------
40
Figure 2. Aquatic Risk
Acute Freshwater Fish Risk on Alfalfa
Others (0.50%W
Chemical G (4.30%-f
Chemical I (8.10%*-
Chemical K <17.40%*
Chemical C (22.20%*
Chemical B (47.50%*
-------
41
In this example, Chemical B contributes the greatest potential acute freshwater
fish risk on alfalfa, followed by Chemical C and Chemical K. Chemical I and Chemical
G contribute a combined total of 12.4%. Other Chemicals used on alfalfa contribute
less than 1%.
D. Frequency (%) of LOC Exceedance
Where the risk quotients can provide an indicator of the magnitude of potential
aquatic risk, considering how often a risk quotient exceeds an LOC can provide an
indicator of the frequency of the potential risk. Four acute freshwater fish risk quotients
were calculated for each pesticide/use combination (See Appendix 22); two acute
marine/estuarine fish risk quotients were calculated for each pesticide/use combination
(See Appendix 23); four acute freshwater invertebrate risk quotients were calculated for
each pesticide/use combination (See Appendix 24); and, two acute marine/estuarine
crustacean risk quotients were calculated for each pesticide/use combination (See
Appendix 25); two acute marine/estuarine mollusc risk quotients were calculated for
each pesticide/use combination (See Appendix 26); two chronic freshwater fish risk
quotients were calculated for each pesticide/use combination (See Appendix 27); and,
two chronic freshwater invertebrate risk quotients were calculated for each
pesticide/use combination (See Appendix 28). The lowest and median toxicity values
were included in the acute freshwater fish and invertebrate risk quotient calculations.
Only the lowest toxicity value was available for the other risk quotient calculations. The
maximum use rate EEC values and one-half these values were included in all
calculations.
Appendices 22b, 23b, 24b, 25b, 26b, 27b, and 28b show how often (%) the
calculated acute and chronic risk quotients exceeded the LOC's for each pesticide/use
combination and each endpoint analyzed. The pesticides in each appendix were
ordered by crop site and by decreasing summed RQ values exceeding the LOC. The
higher the frequency (%), the more times the RQ's exceeded the LOC's. This shows the
relative frequency with which a pesticide/use combination is likely to exceed an LOC.
VI. COMPARATIVE ECOLOGICAL RISK ANALYSIS
This comparative analysis relies upon readily available data for use in estimating
potential risk. Ecotoxicity [5] and environmental fate data were readily available in
Agency files and used to calculate risk quotients, % contributions to RQ sums, and
frequency (%) of exceeding LOCs. Use information [Appendix 1], such as the number
of acres treated by pesticide and crop site, as well as bird and fish incident report data
[22] can also be considered in characterizing the potential risk. Together, they can be
used to compare the ecological risk of pesticides used on the same crop site.
-------
42
A. Calculation of Potential Risk (% of Risk)
Up to this point in the analysis, the risk quotients for all the pesticides have been
summed for each endpoint on each crop site to arrive at the total calculated risk for
each endpoint and each crop site. Next, the percentage contribution of each pesticide
use/combination to the total risk has been determined for each endpoint. This will be
called the percentage (%) contribution to the RQ sum, e.g., % contribution to the RQ
sum for avian acute risk, % contribution to the RQ sum for avian chronic risk, %
contribution to the RQ sum for acute freshwater fish risk.
Further, the frequency that the calculated risk quotients for each pesticide
use/combination exceeded the LOC for each endpoint has been calculated. This will be
called the frequency (%) of exceedance.
The potential risk (which will be called the percentage (%) of risk) for a pesticide
use/combination and any endpoint was assumed to be a function of both the %
contribution of the RQ sum and the frequency % of exceedance. This is expressed in
equation (13).
(13) % contribution to the RQ sum x frequency (%) of exceedance = % of risk
100% 100%
Thus, the % avian acute risk for Chemical 0 applied at-plant to corn is equal to the
product of the % contribution of the RQ sum for avian acute risk sum and the frequency
(%) that the calculated risk quotients exceeded the LOC=0.5.
The avian chronic endpoint presents a slightly different situation. As for the other
endpoints, the % contribution to the RQ sum for avian chronic risk was calculated.
However, as described above, the Time to RQ=1 was calculated in place of the
frequency of exceedance. This calculation better captured the persistence of the
pesticide.
Similar to the % contribution to the RQ sum, the % contribution to the Time to
RQ=1 was calculated for each pesticide use/combination. The potential avian chronic
risk, % avian chronic risk, for a pesticide use/combination was assumed to be the
average of the % contribution to the RQ sum and the % contribution to the Time to
RQ=1. This is expressed in equation (14).
(14) % contribution to the RQ sum/100% + % contribution to the Time toRQ=1/100% = % of avian
2 chronic risk
Most of the pesticides considered in this case study have low NOAEC values
and most calculated RQs exceed the LOC=1. Equation (14) assigns greater % avian
-------
43
chronic risk to those pesticides with greater persistence.
B. Percentage (%) Acres Treated
When considering a practical estimate for extent of risk, the pesticide usage
information in Appendix 1 provided data on number of acres treated. The acre
treatments data incorporated information on multiple applications of the same pesticide
on the same site during the year, and best represented the full extent of the use of the
pesticide. The % acres treated per pesticide provided an estimate of the extent of the
use of a pesticide on a crop site. When comparing pesticides on each crop site, those
used at-plant which are primarily granular formulations, and those used as post-
emergent sprays were compared separately. It was assumed that pesticides applied at-
plant could not substitute for those applied post-emergent, and vice versa.
C. Incident Reports - Bird and Fish Kills
An ecological incident has been defined as an adverse effect on non-target
organisms in the environment (Brassard et al., in press [23]); incidents may range from
incapacitation to mortality among non-target species. Incident data have been used by
the Agency in identifying and confirming ecological risk to non-target organisms. For
almost two decades the OPP has collected incident data. Prior to 1991 these data
were sporadically provided by state and federal agencies on a limited number of
pesticides; however, in 1991 the OPP began to actively solicit data from a variety of
sources that include state agencies, registrants (companies responsible for registering
pesticides), U.S. Fish and Wildlife Service, the National Biological Survey, and the
National Oceanic and Atmospheric Administration. As of early 1998, a total of 4,341
incidents had been reported to OPP; these data were recorded in the Ecological
Incident Information System (EMS) [22], Incident reports are categorized (certainty
index) relative to the likelihood of their being associated with a particular pesticide.
Thus, an incident is classified with one of the following certainty indices: highly
probable, probable, possible, unlikely, and unrelated. A classification of highly
probable indicates the presence of particular chemical residues and/or evidence of a
pesticide-specific effect, e.g., cholinesterase inhibition among organophosphorus
pesticides. A classification of probable implies direct information linking the pesticide
and incident; however, there was no residue data. A certainty index of possible implies
that the pesticide was present but several other compounds may also have been
implicated. The remaining two classifications, i.e., unlikely and unrelated, imply that
there is little to no evidence directly linking a particular pesticide with an incident.
Appendix 29 summarizes the total number of bird and fish kill incidents reported
for the 17 pesticides on all sites recorded in the Ecological Incident Information System
(EMS) as of June of 1998. The incident data base was subjected to two selection
criteria: (1) that the certainty index was either probable or highly probable, and (2) that
-------
44
the cause of the incident was other than misuse. Incidents that have not been
screened or entered into EMS have not been included in this list. The systems reports a
total of 184 bird incidents and 607 fish incidents for these 17 pesticides.
Appendix 30 shows the bird and fish incidents reported for the 17 pesticides on
each of the four crop sites. No incidents were reported for peanuts. The reported
number of bird incidents for alfalfa, corn and cotton were similar, but primarily due to
different pesticides: Chemicals D and G for alfalfa, Chemicals 0 and Q for corn, and
Chemicals B and G for cotton. The site with the greatest number of reported fish
incidents was cotton, primarily due to one pesticide, Chemical B. Chemical Q had the
greatest number for corn.
The existence of highly probable and probable incident reports tends to add a
weight of certainty to the acute risk concerns indicated by the LOC exceedances for
acute risk to birds and fish. Reported incidents for pesticide use\combinations with
LOC exceedances tend to confirm the prediction of mortality based on the acute risk
quotients calculated using laboratory eco-toxicity data and exposure estimates using
models. Due to the fact that the incident data base is still in development, the lack of
reported incidents does not reduce the certainty of risk for pesticide/uses with risk
quotients that exceed the LOC.
D. Comparison of Potential Risk by Crop Site
The % of risk for each endpoint was calculated and compared for each pesticide
used on a crop site . At-plant applications of pesticides , primarily granular
formulations, were compared separately from post-emergent spray applications of
pesticides. Further, it seemed best to compare avian endpoints (four) and aquatic
endpoints (7) separately. It also seemed helpful to compare the pesticides in order of
decreasing % acres treated to add the element of extent of use to the comparison.
Finally, if any bird and/or fish incidents have been reported, the number of bird and fish
incidents was indicated for each pesticide, again in order of decreasing % acres
treated. In addition, the percentage (%) of the total number of incidents reported in EMS
for each site was calculated to provide some perspective on the importance of kills for a
pesticide relative to crop site.
1. Alfalfa - Post-Emergent Spray Formulations
Figure 3 provides overview graphs of the comparative avian risk, aquatic risk
and the incident reports for the pesticides used as post-emergent sprays on alfalfa.
Table 5 shows the data included in Figure 3. There was only one granular at-plant use
reported for alfalfa, and that was for Chemical C. The Chemical C label also included
post-emergent sprays. Consequently, for this analysis, we assumed that most of the
Chemical C was used as post-emergent sprays. Figures 3a, b, and c show individual
graphs for comparative avian risk, aquatic risk and the incident reports for pesticides
-------
45
used as post-emergent sprays on alfalfa.
The total combined use for these pesticides is 4.6 million treated acres.
Chemical C and Chemical D lead with greater than 20% each followed by Chemical G,
Chemical I, and Chemical L at greater than 10% each.
Comparing avian risk, Chemical G leads with the greatest % avian acute
bird/day risk and % avian chronic risk. Chemical L leads with the greatest % avian
dietary risk. Similarly, Chemical G leads the comparative aquatic risk with the greatest
% freshwater fish chronic risk, % freshwater invertebrate acute risk, and the %
freshwater invertebrate chronic risk, % marine/estuarine crustacean acute risk, and %
marine/estuarine mollusc acute risk. However, Chemical B has the greatest %
freshwater fish acute risk, and Chemical C has the greatest % marine/estuarine fish
acute risk.
Incidents have been reported for birds only, and are limited to Chemical D and
Chemical G. The one incident report for Chemical G is not surprising in view of the high
% avian acute bird/day risk. However, the three incidents reported for Chemical D,
representing 50% of bird incidents reported for Chemical D on all sites, is surprising.
The % avian acute bird/day risk is comparatively very low. Additional information is
needed to attempt to characterize these incidents.
In summary for postemergent sprays on alfalfa, Chemical G stands out as
presenting the greatest potential acute and chronic risk to birds, and aquatic
invertebrates, and chronic risk to fish. The acute risk to birds appears to be supported
by and incident report. Chemical B presents the greatest acute risk to freshwater fish,
Chemical C presents the greatest acute risk to marine/estuarine fish, and Chemical L
presents the greatest dietary risk to birds. The bird incident reports raises a question
concerning the comparatively low risk of Chemical D to birds. Overall, Chemical N,
Chemical M, and Chemical P appear to be comparatively less risky than the others.
-------
46
Comparative Avian Risk for Pesticides
Used on Alfalfa as PostEmergent Sprays
JS
o
on
100%
80%
60%
40%
20%
0% Jill n
1
¦
~
I
I
(83)
fa
In Order of
Decreasing %
Acres Treated
( ) = Maximum
RQ Values
Combined # Acres
Treated = 4.6 Million
JL
M
Pesticides
% Avian Acute Bird/Day Risk
% Avian Dietary Risk
~
% Avian Chronic Risk
% Acres Treated
100%
80%
60%
40%
20%
0%
Comparative Aquatic Risk forPesticides
Used on Alfalfa as PostEmergent Sprays
l*-*359)
| -^(9745)
In Order of
Decreasing %
Acres Treated
( ) = Maximum
RQ Values
(8.9)
ill
(63)
Combined # Acres
Treated = 4.6 Million
jj
U
JL.
£L_
_
% Freshwater Fish Acute Risk
_
% Freshwater Fish Chronic Risk
z
% Marine/Estuarine Fish Acute Risk
Z
% Freshwater Invertebrate Acute Risk
~
% Freshwater Invertebrate Chronic Risk
~
% Marine/Estuarine Crustacean Acute
% Marine/Estuarine Mollusc Acute Risk
% Acres Treated
Incident Reports for Pesticides
Used on Alfalfa as PostEmergent Sprays
20
c 15
10
U. 5
=tfc
L K
P
Pesticides
¦ # Reports - Birds
¦ # Reports - Fish
~ % Total Reports - Birds
~ % Total Reports - Fish
J1
m
100% ra
75%
si
<
£
O
50% o
25% ^
0%
-------
47
Comparative Avian Risk for Pesticides
Used on Alfalfa as PostEmergent Sprays
a>
>
¦
+->
w
a>
0£
100%
80%
60%
40%
(1259)
(113)
(83)
II
In Order of
Decreasing %
Acres Treated
( ) - Maximum
RQ Values
_
_
Combined # Acres
Treated = 4.6 Million
£L
K
B
N
Pesticides
¦ % Avian Acute Bird/Day Risk ~
~ % Avian Dietary Risk
% Avian Chronic Risk
% Acres Treated
-------
48
100%
80%
60%
40%
20%
0%
Comparative Aquatic Risk forPesticides
Used on Alfalfa as PostEmergent Sprays
_
_
(107.7)
(2154)
(8.9)
(63)
;
-(359)
-(9745)
IjhJ
In Order of
Decreasing %
Acres Treated
I
I
( ) = Maximum
RQ Values
Combined # Acres
Treated = 4.6 Million
jJ
JUL
XL
(49)
tL-
Pesticides
¦
~
~
~
% Freshwater Fish Acute Risk
% Marine/Estuarine Fish Acute Risk
% Freshwater Invertebrate Chronic Risk
% Marine/Estuarine Mollusc Acute Risk
~
~
~
% Freshwater Fish Chronic Risk
% Freshwater Invertebrate Acute Risk
% Marine/Estuarine Crustacean Acute
% Acres Treated
-------
49
20
W
+¦>
c
0)
T3
15
3 10
o
Q.
Q> -
Q£ 5
Incident Reports for Pesticides
Used on Alfalfa as PostEmergent Sprays
-
m
K
B
l P
Pesticides
M
¦ # Reports - Birds
# Reports - Fish
~ % Total Reports - Birds
~ % Total Reports - Fish
100% £
75%
50%
re
I—
s=
<
c
o
W
+¦>
c
0)
T3
25% ^
o
(0
o
0%
-------
50
Table 5. Data for Figure 3
Pesticides Used on Alfalfa as Post-Emergent Sprays
Birds
Chemical
name
% Avian Acute
Bird/Day Risk
% Avian Chronic Risk
% Avian Dietary
Risk
% Acres Treated
C
3.9%
22.1%
8.2%
29.7%
D
0.4%
2.3%
0.1%
26.3%
G
93.3%
38.8%
10.7%
16.0%
1
0.0%
0.4%
0.0%
12.2%
L
0.9%
19.5%
78.8%
11.6%
P
0.0%
1.5%
0.1%
2.8%
K
0.8%
2.1%
0.5%
0.6%
M
0.0%
0.5%
0.0%
0.4%
B
0.0%
8.4%
0.0%
0.3%
N
0.1%
4.5%
0.1%
0.0%
Aquatic
Organisms
Chemical
name
% Freshwater
Fish
Acute Risk
% Freshwater Fish
Chronic Risk
% Marine/Estuarine
Fish Acute Risk
% Freshwater
Invertebrate
Acute Risk
% Freshwater
Invertebrate
Chronic Risk
%
Marine/Estuarine
Crustacean
Acute Risk
%
Marine/Estuarine
Mollusc Acute
Risk
% Acres
Treated
C
22.2%
19.6%
55.8%
10.3%
26.0%
23.5%
0.2%
29.7%
D
0.0%
0.0%
0.0%
0.1%
0.0%
0.0%
0.0%
26.3%
G
2.1%
48.3%
0.0%
67.7%
70.9%
59.4%
99.2%
16.0%
1
4.0%
0.4%
0.0%
2.7%
1.1%
2.7%
0.0%
12.2%
L
0.0%
0.0%
0.0%
6.5%
0.8%
3.3%
0.2%
11.6%
P
0.1%
0.0%
1.2%
0.1%
0.0%
0.3%
0.0%
2.8%
K
17.4%
2.9%
6.9%
0.3%
0.4%
8.3%
0.0%
0.6%
M
0.0%
0.0%
0.0%
5.7%
0.3%
1.3%
0.0%
0.4%
B
47.5%
28.4%
34.8%
5.9%
0.4%
1.2%
0.0%
0.3%
N
0.0%
0.0%
0.0%
0.3%
0.0%
0.0%
0.0%
0.0%
-------
51
Incidents
JChemical
name
ff Reports - Birds
% Total Reports -
Birds
# Reports - Fish
% Total Reports ¦
Fish
C
0
0%
0
0%
D
3
50%
0
0%
G
1
3%
0
0%
I
0
0%
0
0%
L
0
0%
0
0%
P
0
0%
0
0%
K
0
0%
0
0%
M
0
0%
0
0%
B
0
0%
0
0%
N
0
0%
0
0%
-------
52
2. Corn - At-Plant Granular & Post-Emergent Spray Formulations
Granular At-Plant - Figure 4 provides overview graphs of the comparative avian
risk, aquatic risk and the incident reports for pesticides used as at-plant granular
formulations on corn. Table 6 shows the data included in Figure 4. Chemical Q,
Chemical C, and Chemical 0 were used at-plant. We assumed and the labels seemed
to support the assumption that there was no overlap between those pesticides used at-
plant and those used as post-emergent sprays, i.e., Chemical L, Chemical D, Chemical
G, Chemical E, and Chemical I. Figures 4a, b, and c show individual graphs for
comparative avian risk, aquatic risk and the incident reports for pesticides used as at-
plant granular formulations on corn.
The total combined use for these pesticides is 15.8 million treated acres.
Chemical Q and Chemical C lead with greater than 40% each, and followed by
Chemical 0 at approximately 10%.
Comparing avian risk, Chemical 0 leads with the greatest % avian acute risk.
This is the only avian risk endpoint compared for granular formulations. Chemical 0
also leads the comparative aquatic risk with the greatest % risk for five out of the
seven endpoints: % freshwater fish acute risk, % freshwater fish chronic risk, %
marine/estuarine fish acute risk, % marine/estuarine crustacean acute risk, %
marine/estuarine mollusc acute risk. Chemical C leads the comparative aquatic risk
with the greatest % risk for two out of the seven endpoints: % freshwater invertebrate
acute risk, and the % freshwater invertebrate chronic risk.
Incidents have been reported for birds and fish for Chemical Q and Chemical 0.
only. No incidents have been reported for Chemical C. The bird and fish incident
reports for Chemical 0 are not surprising in view of the high acute % risk for birds and
fish. The bird and fish incident reports for Chemical Q are probably attributed to the
extensive use and may indicate a greater ability to move to water. It is somewhat
surprising that there are no reported incidents for Chemical C.
In summary for at-plant granular pesticides on corn, Chemical 0 stands out as
presenting the greatest potential acute risk to birds, as well as to acute and chronic risk
to freshwater fish, and acute risk to marine/estuarine crustaceans and molluscs. The
acute risk to birds and fish appears to be supported by incident reports. Chemical C
presents the greatest acute and chronic risk to aquatic invertebrates. Chemical Q may
not present comparatively less acute risk to birds and fish considering the reported
incidents.
-------
53
Figure 4
Comparative Avian Risk for Pesticides
Used on Corn as Granular At-Plant
100%
80%
60%
| 40%
20%
0%
Combined # Acres
Treated = 15.8 Million
In Order of
Decreasing %
Acres Treated
( ) = Maximum
RQ Values
I
(1092)
~
% Avian Acute Risk
~
% Acres Treated
Comparative Aquatic Risk - Pesticides
Used on Corn as Granular At-Plant
100%
80%
60%
40%
20%
0%
In Order of
Decreasing %
Acres Treated
( ) = Maximum RQ
Values
Combined #Acres
Treated = 15.8 Million
(2682.6)
(212.7) r n
ttol jrf.J
(32.^ (23.8)
jA
132.7^ i
£
(13.6)
¦ % Freshwater Fish Acute Risk
~ % Marine/Estuarine Fish Acute Risk
~ % Freshwater Invertebrate Chronic Risk
~ % Estuarine/Marine Mollusc Risk
~ % Freshwater Fish Chronic Risk
~ % Freshwater Invertebrate Acute Risk
~ % Estuarine/Marine Crustacean Acute
I % Acres Treated
Incident Reports for Pesticides
Used on Corn
25
¦£ 20
a)
¦g
I 15
T3
a)
t 10
o
a.
a)
* 5
—^
_r-T
¦
Q
c
O
¦ # Reports - Birds
I # Reports - Fish
~ % Total Reports - Birds
~ % Total Reports - Fish
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
-------
54
Comparative Avian Risk for Pesticides
Used on Corn as Granular At-Plant
100%
80%
55 60%
a)
>
+5
40%
20%
0%
~
% Avian Acute Risk
% Acres Treated
Combined # Acres
Treated = 15.8 Million
In Order of
Decreasing %
Acres Treated
(1092)
( ) = Maximum
RQ Values
Q O
C
-------
55
Comparative Aquatic Risk - Pesticides
Used on Corn as Granular At-Plant
100%
80% -H
so
°
| 40%
£
20%
0%
In Order of
Decreasing %
Acres Treated
_
( ) = Maximum RQ
Values
_
Combined # Acres
Treated = 15.8 Million
.(2682.6)
(212.7)
(32.7) (23.8)
(92.6)
(32.fl
Jfr
(13.6)
-(298.8)
% Freshwater Fish Acute Risk
% Marine/Estuarine Fish Acute Risk
% Freshwater Invertebrate Chronic Risk
% Estuarine/Marine Mollusc Risk
% Freshwater Fish Chronic Risk
% Freshwater Invertebrate Acute Risk
% Estuarine/Marine Crustacean Acute
% Acres Treated
-------
56
Incident Reports for Pesticides
Used on Corn
J2
c
0)
"O
"O
0)
L.
o
Q.
0)
a:
*
25
20
15
10
5
0
¦ ¦
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
o
o
c
o
(0
+->
c
d)
;o
o
c
<4—
o
=tfc
+->
o
# Reports - Birds
# Reports - Fish
% Total Reports - Birds
% Total Reports - Fish
-------
57
Table 6. Data for Figure 4
Pesticides Used on Corn as Granular At-Plant Formulations
Birds
Chemical
1% Avian Acute
name
flisk
% Acres Treated
Q
J 1.0%
45.6%
C
| 12.8%
43.0%
O
| 84.4%
11.4%
Aquatic
Organisms
Chemical
I
name
% Freshwater
Fish
Acute Risk
% Freshwater
Fish
Chronic Risk
%
Vlarine/Estuarine
Fish Acute Risk
% Freshwater
Invertebrate
Acute Risk
% Freshwater
Invertebrate
Chronic Risk
%
Vlarine/Estuarine
Crustacean Acute
Risk
%
Marine/Estuarine
Mollusc Acute
Risk
% Acres
Treated
Q
21.1%
16.3%
31.3%
28.0%
7.9%
19.9%
0.0%
45.6%
C
6.3%
31.0%
15.0%
42.1%
49.5%
21.0%
1.2%
43.0%
O
66.3%
52.7%
53.7%
29.9%
42.6%
59.1%
97.5%
11.4%
Incidents
Chemical
name
# Reports - Birds
% Total Reports -
Birds
ff Reports - Fish
% Total Reports ¦
cish
Q
1
20%
21
46%
C
0
0%
0
0%
O
2
3%
3
75%
-------
58
Post-Emergent Sprays - Figure 5 provides overview graphs of the comparative
avian risk, aquatic risk and the incident reports for pesticides used as post-emergent
sprays on corn. Table 7 shows the data included in Figure 5. Chemical L, Chemical D,
Chemical G, Chemical E, and Chemical I were used post-emergent sprays. Figures 5a
and b show individual graphs for comparative avian risk and aquatic risk for pesticides
used as post-emergent sprays on corn. No incidents have been reported for any of the
pesticides used as post-emergent sprays on corn.
The total combined use for these pesticides is 2.6 million treated acres.
Chemical L leads with greater than 50%, and followed by Chemical D at 20% and
Chemical G at slightly less than 20%.
Comparing avian risk, Chemical G leads with the greatest % avian acute and
chronic risk. Chemical L leads with the greatest % avian dietary risk. Chemical G also
leads the comparative aquatic risk with the greatest % risk for five out of the seven
endpoints: % freshwater fish chronic risk, % freshwater invertebrate acute risk, %
freshwater invertebrate chronic risk, % marine/estuarine crustacean acute risk, %
marine/estuarine mollusc acute risk. Chemical I leads the comparative aquatic risk with
the greatest % risk for two out of the seven endpoints: % freshwater fish acute risk, %
marine/estuarine fish acute risk. It is important to note that the maximum RQ value for
the latter is only 0.6, just exceeding the LOC of 0.5.
In summary for post-emergent sprays on corn, Chemical G stands out as
presenting the greatest potential acute and chronic risk to birds, as well as to acute and
chronic risk to freshwater invertebrates, and acute risk to marine/estuarine crustaceans
and molluscs. Chemical L presents the greatest potential avian dietary risk, and
Chemical I the same for freshwater fish. Overall, Chemical D and Chemical E appear to
be comparatively less risky than the others.
-------
59
Figure 5
Comparative Avian Risk for Pesticides
Used on Corn as PostEmergent Sprays
100%
80%
In Order of
Decreasing %
Acres Treated
(1678.5)
(115.1)
( ) = Maximum
RQ Values
Combined # Acres
T reated = 2.6
¦ % Avian Acute Bird/Day Risk ~ % Avian Dietary Risk
~ % Avian Chronic Risk % Acres Treated
100%
Comparative Aquatic Risk - Pesticide
Used on Corn as PostEmergent Sprays
Combined # Acres
T reated = 2.6
Million
0)
>
at
OL
¦
~
~
~
% Freshwater Fish Risk
% Marine/Estuarine Fish Acute Risk
% Freshwater Invertebrate Chronic Risk
% Marine/Estuarine Mollusc Acute Risk
~
~
~
%Freshwater Fish Chronic Risk
% Freshwater Invertebrate Acute Risk
% Marine/Estuarine Crustacean Acute
% Acres Treated
No Incident Reports
-------
60
Comparative Avian Risk for Pesticides
Used on Corn as PostEmergent Sprays
100%
_
_
In Order of
Decreasing %
Acres Treated
J
(1678.5)
(115.1)
( ) = Maximum
RQ Values
Combined # Acres
Treated = 2.6
¦ % Avian Acute Bird/Day Risk ~ % Avian Dietary Risk
~ % Avian Chronic Risk % Acres Treated
-------
61
100%
Comparative Aquatic Risk - Pesticide
Used on Corn as PostEmergent Sprays
80% -
o 60%
>
-------
62
Table 7. Data for Figure 5
Pesticides Used on Corn as Post-Emergent Sprays
Birds
Chemical
name
% Avian Acute
Bird/Day Risk
% Avian Dietary
Risk
% Avian Chronic
Risk
% Acres Treated
L
0%
75%
19%
53%
D
1%
1%
5%
20%
G
96%
21%
69%
18%
E
2%
2%
5%
8%
1
0%
0%
1%
0%
Aquatic
Organisms
Chemical
name
% Freshwater
Fish
Acute Risk
% Freshwater Fish
Chronic Risk
%
Vlarine/Estuarine
Fish Acute Risk
% Freshwater
Invertebrate Acute
Risk
% Freshwater
Invertebrate Chronic
Risk
% Marine/Estuarine
Crustacean Acute
Risk
% Marine/Estuarine
Mollusc Acute Risk
% Acres
Treated
L
0%
0%
0%
4%
1%
3%
0%
53%
D
0%
0%
0%
0%
0%
0%
0%
20%
G
14%
95%
0%
90%
92%
92%
100%
18%
E
2%
2%
0%
0%
6%
0%
0%
8%
1
51%
3%
50%
5%
2%
5%
0%
0%
No Incident
Reports
-------
63
3. Cotton - At-Plant Granular & Post-Emergent Spray Formulations
Granular At-Plant - Figure 6 provides overview graphs of the comparative avian
risk, aquatic risk and the incident reports for pesticides used as at-plant granular
formulations on cotton. Table 8 shows the data included in Figure 6. Chemical A,
Chemical E, Chemical 0, and Chemical H were used at-plant. We assumed and the
labels seemed to support the assumption that there was no overlap with the use of
Chemical 0 and Chemical H at-plant and those Chemicals used as post-emergent
sprays. However, the remaining Chemicals used on cotton had labels for both at-plant
and post-emergent sprays. There was no information separating use by formulation
and timing of application. Total use figures were used for both at-plant and post-
emergent sprays. Figures 6a and b show individual graphs for comparative avian risk
and aquatic risk for pesticides used as at-plant granular formulations on cotton. There
were no incident reports for pesticides used at-plant on cotton.
The total combined use for these pesticides is 2.8 million treated acres.
Chemical A leads with 60%. Chemical E follows with 22%, and Chemical 0 is third with
15%.
Comparing avian risk, Chemical 0 leads with the greatest % avian acute risk.
This is the only avian risk endpoint compared for granular formulations. Chemical 0
also leads the comparative aquatic risk with the greatest % risk for five out of the
seven endpoints: % freshwater fish acute risk, % freshwater fish chronic risk, %
marine/estuarine fish acute risk, % marine/estuarine crustacean acute risk, %
marine/estuarine mollusc acute risk. Chemical H leads the comparative aquatic risk
with the greatest % risk for the % freshwater invertebrate acute risk, while Chemical E
leads with the greatest % risk for freshwater invertebrate chronic risk.
In summary for at-plant granular pesticides on cotton, Chemical 0 stands out as
presenting the greatest potential acute risk to birds, as well as to acute and chronic risk
to freshwater fish, and acute risk to marine/estuarine crustaceans and molluscs.
Chemical H and Chemical E present the greatest risk freshwater to invertebrates, acute
and chronic risk respectively. Overall, Chemical A appears to present comparatively
less acute risk to birds and fish.
-------
64
Figure 6
Comparative Avian Risk for Pesticides
Used on Cotton as Granular At-Plant
100%
80%
60%
40%
20%
0%
Combined # Acres
Treated = 2.8 Million
(1847.5)
J
_
In Order of
Decreasing %
Acres Treated
( ) = Maximum
RQ Values
~
% Avian Acute Risk
% Acres Treated
Comparative Aquatic Risk - Pesticides
Used on Cotton as Granular At-Plant
100%
80%
g 60%
| 40%
a:
20%
0%
Combined # Acres
Treated = 2.8 Million
( ) = Maximum
RQ Values
(49.7)
I
(12.7)
J1737.3)
JUj
I
(7.3)
In Order of
Decreasing %
Acres Treated
i
¦
~
~
~
% Acute Freshwater Fish Risk
% Marine/Estuarine Acute Fish Risk
% Chronic Freshwater Invertebrate R|
% Marine/Estuarine Acute Mollusc Ri:
~
~
% Chronic Freshwater Fish Risk
% Acute Freshwater Invertebrate Risk
% Marine/Estuarine Acute Crustacean
% Acres Treated
No Incident Reports
-------
65
Comparative Avian Risk for Pesticides
Used on Cotton as Granular At-Plant
100%
80%
60%
40%
20%
0%
_
Combined # Acres
Treated = 2.8 Million
(1847.5)
In Order of
Decreasing %
Acres Treated
( ) = Maximum
RQ Values
H
% Avian Acute Risk
% Acres Treated
-------
66
Comparative Aquatic Risk - Pesticides
Used on Cotton as Granular At-Plant
100%
80%
g 60%
J2
0
£
40%
20%
0%
Combined # Acres
Treated = 2.8 Million
( ) = Maximum
RQ Values
(49.7)
I
(12.7)
(1737.3)
(17.4)
Li
(7.3)
In Order of
Decreasing %
Acres Treated
(160.3)
(82.8)
H
% Acute Freshwater Fish Risk
% Marine/Estuarine Acute Fish Risk
% Chronic Freshwater Invertebrate Ri
% Marine/Estuarine Acute Mollusc Ri
% Chronic Freshwater Fish Risk
% Acute Freshwater Invertebrate Risk
% Marine/Estuarine Acute Crustacean
% Acres Treated
-------
67
Table 8. Data for Figure 6
Pesticides Used on Cotton as Granular At-Plant Formulations
Birds
Chemical
name
% Avian Acute
Risk
% Acres Treated
A
0%
60%
E
7%
22%
O
90%
15%
H
1%
3%
Aquatic
Organisms
Chemical
name
% Freshwater
Fish
Acute Risk
% Freshwater Fish
Chronic Risk
% Marine/Estuarine
Fish Acute Risk
% Freshwater
Invertebrate Acute
Risk
% Freshwater
Invertebrate
Chronic Risk
% Marine/Estuarine
Crustacean Acute
Risk
% Marine/Estuarine
Mollusc Acute Risk
% Acres Treated
A
0%
0%
0%
0%
0%
0%
0%
60%
E
1%
10%
0%
9%
47%
3%
0%
22%
O
78%
49%
66%
37%
34%
88%
100%
15%
H
19%
41%
34%
54%
19%
9%
0%
3%
No Incident
Reports
-------
68
Post-Emergent Sprays - Figure 7 provides overview graphs of the comparative
avian risk, aquatic risk and the incident reports for pesticides used as post-emergent
sprays on cotton. Table 9 shows the data included in Figure 7. Chemical L, Chemical
B, Chemical A, Chemical D, Chemical C, Chemical I, Chemical E, Chemical J,
Chemical G, Chemical N, Chemical M, and Chemical Kwere used post-emergent
sprays. Figures 7a, b, and c show individual graphs for comparative avian risk, aquatic
risk and the incident reports for pesticides used as post-emergent sprays on cotton.
The total combined use for these pesticides is 11.3 million treated acres.
Chemical L leads with greater than 20%. Chemical B, Chemical A, Chemical D, and
Chemical C, followed with greater than 10%.
Comparing avian risk, Chemical G leads with the greatest % avian acute and
chronic risk. Chemical L leads with the greatest % avian dietary risk. Chemical G also
leads the comparative aquatic risk with the greatest % risk for four out of the seven
endpoints: % freshwater invertebrate acute risk, % freshwater invertebrate chronic risk,
% marine/estuarine crustacean acute risk, % marine/estuarine mollusc acute risk.
Chemical B leads the comparative aquatic risk with the greatest % risk for two out of
the seven endpoints: % freshwater fish acute risk, % freshwater fish chronic risk, and %
marine/estuarine fish acute risk.
Incidents have been reported for birds and fish. Bird incident reports are
reported for Chemical B (3) and Chemical G (1). Fish incidents are reported for
Chemical B (126) and Chemical L (2). The bird report for Chemical G and the fish
reports for Chemical B are not surprising considering the high acute % risk for birds
and fish, respectively. The greater number of bird incidents reported for Chemical B is
probably due to its greater use.
In summary for post-emergent sprays on cotton, Chemical G stands out as
presenting the greatest potential acute and chronic risk to birds, as well as to acute and
chronic risk to freshwater invertebrates, and acute risk to marine/estuarine
crustaceans and molluscs. Chemical L presents the greatest potential avian dietary
risk. Chemical B presents the greatest acute risk to freshwater and marine/estuarine
fish, as well as chronic risk to freshwater fish. Overall, Chemical A, Chemical D,
Chemical I, Chemical J and Chemical N appear to be comparatively less risky than the
others.
-------
69
Figure 7
Comparative Avian Risk for Pesticides
Used on Cotton as PostEmergent Sprays
100%
80%
3?
(5
60%
I 40%
20%
0%
jj
(49.8)
I
(334.4)
J
Combined # Acres
Treated = 11.2 Million
In Order of
Decreasing %
Acres Treated
j
1
Bdllll
(22493.6)
I
= (4978.8)
( ) = Maximum
RQ Values
(829.8)
JL
ML
% Acute Freshwater Fish Risk ~ % Chronic Freshwater Fish Risk
~ % Marine/Estuarine Acute Fish Risk ~ % Acute Freshwater Invertebrate Risk
~ % Chronic Freshwater Invertebrate Ris o % Marine/Estuarine Acute Crustacean
~ % Marine/Estuarine Acute Mollusc Risk^^J % Acres T reated
200
U)
c
160
<1)
¦o
()
c
120
¦O
<1)
t
O
80
Q.
-------
70
Comparative Avian Risk for Pesticides
Used on Cotton as PostEmergent Sprays
100%
80%
_
_
60% -
40% -
20% \-
0%
Combined # Acres
Treated = 11.3
(131.0)
:li J
(2940.2)
B
( ) =
Maximum RQ
tforfr
In Order of
Decreasing %
Acres Treated
(201.7)
% Avian Acute Bird per Day Risk
% Avian Chronic Risk
% Avian Dietary Risk
% Acres Treated
-------
71
Comparative Aquatic Risk - Pesticides
Used on Cotton as PostEmergent Sprays
A I G K
% Acute Freshwater Fish Risk
% Chronic Freshwater Fish Risk
% Acute Freshwater Invertebrate Risk
% Marine/Estuarine Acute Crustacean
% Acres Treated
% Marine/Estuarine Acute Fish Risk
% Chronic Freshwater Invertebrate Ri
% Marine/Estuarine Acute Mollusc Ris
-------
72
Incident Reports for Pesticides
Used on Cotton
C/>
+-<
C
0
"O
¦
o
-a
0
t
o
Q.
0
a:
*
200
160
120
80
40
-
.
L
D
B C
A I
E
N
J M
G K
¦
# Reports - Birds
# Reports - Fish
% Total Reports - Birds
% Total Reports - Fish
100% g
90% Q
80% c
70%
60%
50%
40%
30%
20%
10%
"%
o
JQ
c
0
"O
¦
o
o
*
"re
¦*->
o
-------
73
Table 9. Data for Figure 7
Pesticides Used on Cotton as Post-Emergent Sprays
Birds
Chemical
name
% Avian Acute
Bird/Day Risk
% Avian Dietary
Risk
% Avian Chronic
Risk
% Acres Treated
L
1%
70%
13%
23%
B
1%
4%
26%
16%
A
0%
0%
2%
15%
D
0%
0%
1%
14%
C
3%
7%
17%
10%
1
0%
0%
0%
7%
E
1%
0%
2%
5%
J
0%
1%
1%
4%
G
93%
14%
32%
3%
N
0%
0%
4%
2%
M
0%
0%
0%
0%
K
1%
1%
2%
0%
Aquatic
Organisms
Chemical
name
% Freshwater
Fish
Acute Risk
% Freshwater Fish
Chronic Risk
% Marine/Estuarine
Fish Acute Risk
% Freshwater
Invertebrate
Acute Risk
% Freshwater
Invertebrate Chronic
Risk
% Marine/Estuarine
Crustacean Acute
Risk
%
Marine/Estuarine
Mollusc Acute
Risk
% Acres Treated
L
0%
0%
0%
8%
1%
4%
1%
23%
B
81%
63%
75%
21%
2%
4%
0%
16%
A
0%
0%
0%
0%
0%
0%
0%
15%
D
0%
0%
0%
0%
0%
0%
0%
14%
C
6%
7%
20%
6%
17%
14%
0%
10%
1
1%
1%
0%
2%
1%
2%
0%
7%
E
0%
1%
0%
0%
6%
0%
0%
5%
J
0%
0%
0%
0%
0%
0%
0%
4%
G
1%
27%
0%
61%
73%
56%
99%
3%
N
0%
0%
0%
0%
0%
0%
0%
2%
M
0%
0%
0%
2%
0%
11%
0%
0%
K
8%
2%
4%
0%
0%
9%
0%
0%
-------
74
Incidents
JChemical
name
# Reports - Birds
% Total Reports -
Birds
# Reports - Fish
% Total Reports ¦
Fish
L
0
0%
2
50%
B
3
60%
126
34%
A
0
0%
0
0%
D
0
0%
0
0%
C
0
0%
0
0%
1
0
0%
0
0%
E
0
0%
0
0%
J
0
0%
0
0%
G
1
3%
0
0%
N
0
0%
0
0%
M
0
0%
0
0%
K
0
0%
0
0%
-------
75
4. Peanuts - At-Plant Granular & Post-Emergent Spray Formulations
Granular At-Plant - Figure 8 provides overview graphs of the comparative avian
risk, aquatic risk and the incident reports for pesticides used as at-plant granular
formulations on peanuts. Table 10 shows the data included in Figure 8. Chemical A,
Chemical 0, Chemical E, Chemical F, and Chemical H were used at-plant. Chemical A
and Chemical F had labels for post-emergent sprays as well as for at-plant granular
formulations. Chemical I labels support only post-emergent sprays. There was no
information separating use by formulation and timing of application. Total use figures
were used for both at-plant and post-emergent sprays. Figures 8a and b show
individual graphs for comparative avian and aquatic risk for pesticides used as at-plant
granular formulations on peanuts. There were no incident reports for pesticides used
at-plant on peanuts.
The total combined use for these pesticides is 0.5 million treated acres.
Chemical A leads with 35%, with Chemical 0 and Chemical E closely following with
30% and 25%, respectively.
Comparing avian risk, Chemical 0 leads with the greatest % avian acute risk.
This is the only avian risk endpoint compared for granular formulations. Chemical 0
also leads the comparative aquatic risk with the greatest % risk for six out of the seven
endpoints: % freshwater fish acute risk, % freshwater fish chronic risk, %
marine/estuarine fish acute risk, % freshwater fish acute risk, % freshwater fish chronic
risk, and % marine/estuarine crustacean acute risk. Chemical F leads with the greatest
% marine/estuarine mollusc acute risk.
In summary for at-plant granular pesticides on peanuts, Chemical 0 stands out
as presenting the greatest potential acute risk to birds, as well as to acute and chronic
risk to freshwater fish, freshwater invertebrates and marine/estuarine invertebrates.
Chemical F presents the greatest acute risk to marine/estuarine mollusc. Overall,
Chemical A appears to present comparatively less acute risk to birds and fish.
-------
76
Figure 8
Comparative Avian Risk for Pesticides
Used on Peanuts as Granular At-Plant
100%
80%
a> 60%
>
£ 40%
Combined # Acres
Treated = 0.5 Million
20%
0% BJ —
(2519.3)
I J
J D
In Order of
Decreasing %
Acres Treated
( ) = Maximum
RQ Values
~
% Avian Acute Risk
~
% Acres Treated
100%
80%
60%
40%
20%
0%
Comparative Aquatic Risk - Pesticides
Used on Peanuts as Granular At-Plant
In Order of
Decreasing
% Acres
Treated
i
i FT
(93.2) fl (300.6) !,
iz —l L
fflr ~(23-9)
. —(155.3)
•5> —I' 1
—! i —(2298.9)
( ) = Maximum
RQ Values
i jii
Combined #
Acres Treated =
0.5 Million
m % Freshwater Fish Acute Risk ~ % Freshwater Fish Chronic Risk
~ % Marine/Estuarine Fish Acute Risk ~ % Freshwater Invertebrate Acute Risl^
~ % Freshwater Invertebrate Chronic a
~ % Estuarine/Marine Mollusc Risk | |
% Estuarine/Marine Crustacean Acute
% Acres Treated
No Incident Reports
-------
77
Comparative Avian Risk for Pesticides
Used on Peanuts as Granular At-Plant
100%
80%
a) 60%
>
¦
JH
® 40%
20%
0%
_
Combined # Acres
Treated = 0.5 Million
(2519.3)
In Order of
Decreasing %
Acres Treated
_
( ) = Maximum
RQ Values
H
| | % Avian Acute Risk % Acres Treated
-------
78
Comparative Aquatic Risk - Pesticides
Used on Peanuts as Granular At-Plant
100%
—
(93.2)
(300.6)
80%
60%
40%
20%
0%
In Order of
Decreasing
% Acres
Treated
--T
r*~(23-9)
J. f (-<155.3)
(32.5) —to* I
™ ¦ ¦—(2298.9)
( ) - Maximum
RQ Values
nj
i
(92.7)
Combined #
Acres Treated
0.5 Million
H
% Freshwater Fish Acute Risk
% Freshwater Fish Chronic Risk
% Freshwater Invertebrate Acute Risk
% Estuarine/Marine Crustacean Acute
% Acres Treated
% Marine/Estuarine Fish Acute Risk
% Freshwater Invertebrate Chronic Ri
% Estuarine/Marine Mollusc Risk
-------
79
Table 10. Data for Figure 8
Pesticides Used on Peanuts as Granular At-Plant Formulations
Birds
Chemical
name
% Avian Acute
Risk
% Acres TreatecJ
A
0%
35%
O
59%
30%
E
8%
25%
F
22%
5%
H
11%
4%
Aquatic
Organisms
Chemical
name
% Freshwater
Fish
Acute Risk
% Freshwater
Fish
Chronic Risk
%
Vlarine/Estuarine
Fish Acute Risk
% Freshwater
Invertebrate
Acute Risk
% Freshwater
Invertebrate Chronic
Risk
%
Marine/Estuarine
Crustacean
Acute Risk
%
Marine/Estuarine
Mollusc Acute
Risk
% Acres Treated
A
0%
0%
0%
0%
0%
0%
0%
35%
O
92%
52%
70%
64%
44%
92%
13%
30%
E
1%
7%
0%
11%
44%
2%
0%
25%
F
0%
31%
22%
4%
6%
4%
87%
5%
H
4%
10%
8%
21%
6%
2%
0%
4%
No Incident
Reports
-------
80
Post-Emergent Sprays - Figure 9 provides overview graphs of the comparative
avian risk, aquatic risk and the incident reports for pesticides used as post-emergent
sprays on peanuts. Table 11 shows the data included in Figure 9. Chemical A,
Chemical F and Chemical I were used post-emergent sprays. Figures 9a and b show
individual graphs for comparative avian risk and aquatic risk for pesticides used as
post-emergent sprays on peanuts. No incidents were reported for these pesticides used
as post-emergent sprays on peanuts.
The total combined use for these pesticides is 0.2 million treated acres.
Chemical A leads with over 80%. Chemical F follows with 13%.
Comparing avian risk, Chemical F leads with the greatest % avian acute, %
avian dietary risk, and % avian chronic risk. Chemical F also leads the comparative
aquatic risk with the greatest % risk for five out of the seven endpoints: % freshwater
fish chronic risk, % marine/estuarine fish acute risk, % freshwater invertebrate chronic
risk, % marine/estuarine crustacean acute risk, and % marine/estuarine mollusc acute
risk. Chemical I leads the comparative aquatic risk with the greatest % risk for two out
of the seven endpoints: % freshwater fish acute risk, and % freshwater invertebrate
acute risk,.
In summary for post-emergent sprays on peanuts, Chemical F stands out as
presenting the greatest potential acute, dietary and chronic risk to birds, as well as
potential acute risk to marine/estuarine fish, chronic risk to freshwater fish and
invertebrates, and acute risk to marine/estuarine crustaceans and molluscs. Chemical I
presents the greatest potential acute risk to freshwater fish and invertebrates. Overall,
Chemical A appears to be comparatively less risky than the others.
-------
81
Figure 9
Comparative Avian Risk for Pesticides
Used on Peanuts as Post-Emergent Spray
100%
80%
(u
60%
>
to
+-»
TO
-------
82
100%
80%
60%
>
¦
+-<
I 40%
£
20%
0%
Comparative Avian Risk for Pesticides
Used on Peanuts as Post-Emergent Spray
(78.1)
(43.6)
(154.7)
In Order of
Decreasing %
Acres Treated
( ) = Maximum
RQ Values
Combined #
Acres Treated
0.2 Million
% Avian Acute Bird/Day Risk
% Avian Chronic Risk
% Avian Dietary Risk
% Acres Treated
-------
83
Comparative Aquatic Risk - Pesticides
Used on Peanuts as PostEmergent Spray
100%
£
80% -
® 60%
¦2 40% -|
0
20% 4J
0%
In Order of
Decreasing
% Acres
Treated
Combined #
Acres Treated
= 0.2 Million
(19.2)
(7.3)
(92.7)
( ) - Maximum
RQ Values
<295.8)
-(12.8)
(5.4)
(42.9)
% Freshwater Fish Acute Risk
% Marine/Estuarine Fish Acute Risk
% Freshwater Invertebrate Chronic Ri
% Estuarine/Marine Mollusc Risk
% Freshwater Fish Chronic Risk
% Freshwater Invertebrate Acute Risk
% Estuarine/Marine Crustacean Acute
% Acres Treated
-------
84
Table 11. Data for Figure 9
Pesticides Used on Peanuts as Post-Emergent Sprays
Birds
Chemical
name
% Avian Acute
Bird/Day Risk
% Avian Dietary
Risk
% Avian Chronic
Risk
% Acres Treatecj
A
0%
0.0%
8%
86%
F
100%
100.0%
92%
13%
I
0%
0.0%
0%
1%
Aquatic
Organisms
Chemical
I
name
% Freshwater
Fish
Acute Risk
% Freshwater
Fish
Chronic Risk
%
Marine/Estuarine
Fish Acute Risk
% Freshwater
Invertebrate
Acute Risk
% Freshwater
Invertebrate
Chronic Risk
%
Marine/Estuarine
Crustacean Acute
Risk
%
Marine/Estuarine
Mollusc Acute
Risk
% Acres Treated
A
0%
0%
0%
0%
0%
0%
0%
86%
F
2%
97%
100%
17%
77%
57%
100%
13%
1
45%
2%
0%
83%
23%
43%
0%
1%
No Incident
Reports
-------
85
5. Overall Summary
Figures 4, 6 and 8 show that Chemical 0 stands out as consistently presenting
the greatest overall potential risk for granular at-plant pesticides applied to these four
crops.
Figures 3, 5, 7 and 9 present a more complex picture of potential risk for
pesticides applied as post-emergent sprays to these four crops. Chemical G appears to
present the greatest risk to birds and aquatic invertebrates on alfalfa, corn and cotton.
Chemical B presents the greatest risk to fish on alfalfa and cotton; while Chemical I
presents the greatest risk to fish on corn. Peanuts is a separate case and is described
above.
-------
86
VII. Decision Support Analysis to Aid Decision-Making
A. Decision Support Software16
Comparative risk assessment can become a daunting process when the risk
assessor(s) are faced with ecological risks for a number of alternative pesticides
covering multiple endpoints. When attempting to decide which pesticides present the
greatest overall risk or which are comparatively less risky than others, the matrix-like
result of such comparisons can lead the assessor(s) to rely on individual or group
intuition more than they prefer. Further, when additional information characterizing the
risk, whether quantitative or not, is added for consideration in the decision making
process, paralysis (indecision) can set in. To address this situation, the Agency chose
to explore the use of commercially available software called DecideRight (Version 1.2)u
to aid in decision-making for this case study. This software was primarily designed for
use in businesses. It is a user friendly tool to help choose among alternatives when
many factors must be considered. The underlying methodology is SMART, the Simple
Multi-attribute Rating Technique. It was developed about 27-years ago and has
become a standard in decision modeling. When faced with a number of choices and a
number of criteria, SMART prescribes that (1) each choice be rated on each criterion,
(2) each criterion be assigned a measure of importance to the decision-maker, (3) a
summary score for each choice be calculated as a weighted average of the ratings,
where the weights represent the relative importance of the criteria. Thus, the higher the
summary score, the better the choice. The result of this process has proved to be
superior to the alternative of reliance on intuition.
SMART is not rooted in probability. The assigned ratings are assumed to be
based on full knowledge of how an option works. There are no uncertainties. However,
some uncertainty can be dealt with in the ratings by sensitivity testing of the results.
Similarly, this analysis is not rooted in probability. However, when the tools for
probabilistic ecological risk assessments become available and are implemented in
OPP/EFED, a decision tool that is capable of handling probabilities should be
considered for comparative analysis. ERGO18 by Arlington Software Corporation
currently is capable of handling probability in decision analysis.
16Much of the software description here was based on a software review by Len Tashman and Sara
Munro [24],
17 DecideRight was developed by Avantos Performance Systems of Emertville, California. The
company has since closed; however, the software is still available from Parsons Technology, P.O. Box 100,
Hiawatha, IA 52233 [http://www.parsonstech.coml Mention of this commercial product does not constitute
a recommendation or endorsement by EPA.
18Arlington Software Corporation, 740 Saint-Maurice, Suite 410, Montreal, Quebec, Canada, H3C
1L5 [http://www.arlinqsoft.coml Mention of this commercial product does not constitute a recommendation
or endorsement by EPA.
-------
87
In addition, there are no interactions between attributes in SMART. Thus,
SMART ignores any interplay between two criteria. This interaction effect is handled
better by more complex models such as ERGO.
DecideRight is easy to use and provides the results in intuitive and colorful
decision tables. In addition to a baseline scenario, the software allows the development
various additional scenarios through its scenario facility. This permits consideration of
additional risk characterization information and also preferences of different members
in a team environment. EPA selected DecideRight for this exploratory case study.
B. Baseline Scenario - Alfalfa
The Agency chose to enter the data in Table 5, for all pesticides used on alfalfa
as post-emergent sprays, into DecideRight. The twelve criteria i.e., the risk endpoints,
were considered to be equally important for this scenario, and set at a level of medium
importance. The criteria included # of bird incident reports, # offish incident reports, %
avian bird/day risk, % avian dietary risk, % avian chronic risk, % freshwater fish acute
risk, % marine/estuarine fish acute risk, % freshwater fish chronic risk, % freshwater
invertebrate acute risk, % freshwater invertebrate chronic risk, % marine/estuarine
crustacean acute risk, % marine/estuarine mollusc acute risk.
Each pesticide was rated for each criteria using the % risk values in Table 5.
The option was chosen to rate lower % risk values better. Thus, lower % risk ratings
result in higher summary ratings. The question posed was "Which of the Insecticides
Used on Alfalfa are Less Risky?" The corollary is "Which of the Insecticides are More
Risky?" The results are presented in a decision table format in Figure 10.
-------
88
Figure 10. Baseline: Decision Table for Which of the Insecticides Used on Alfalfa are Less Risky
# of bird incident reports
% avian acute bird/day risk
% avian chronic risk
% avian dietary risk
% freshwater fish acute risk
% freshwater fish chronic risk
% freshwater invertebrate acute risk
% freshwater invertebrate chronic risk
% Marine/Estuarine Fish Acute Risk
% Marine/Estuarine Mollusc Acute Risk
% Marine/Estuarine Crustacean Acute Ris
# offish incident reports
Summary
Chemical P
N/A
0.00
1.50
0.10
0.10
0.00
0.10
0.00
1.20
0.00
0.30
N/A
9.94
Chemical N
N/A
0.10
4.50
0.10
0.00
0.00
0.30
0.00
0.00
0.00
0.00
N/A
9.89
Chemical M
N/A
0.00
0.50
0.00
0.00
0.00
5.70
0.30
0.00
0.00
1.30
N/A
9.89
Chemical I
N/A
0.00
0.40
0.00
4.00
0.40
2.70
1.10
0.00
0.00
2.70
N/A
9.81
Chemical K
N/A
0.80
2.10
0.50
17.40
2.90
0.30
0.40
6.51
0.00
8.30
N/A
9.25
Chemical D
Poor
0.40
2.30
0.10
0.00
0.00
0.10
0.00
0.00
0.00
0.00
N/A
9.04
Chemical L
N/A
0.90
19.50
78.80
0.00
0.00
6.50
0.80
0.00
0.20
3.30
N/A
8.33
Chemical B
N/A
0.00
8.40
0.00
28.40
5.90
0.40
34.80
0.00
1.20
N/A
7.47
Chemical C
N/A
3.90
22.10
8.20
22.20
19.60
10.30
26.00
55.80
0.20
23.50
N/A
6.51
Chemical G
Poor
93.30
38.80
10.70
2.10
48.30
67.70
70.90
0.00
99.20
59.40
N/A
2.57
-------
89
Alternative choices considered are listed down the left side of the table. The
criteria used to evaluate the various options are listed along the top. Initially entered in
no particular order, both the choices and the criteria were then repositioned according
to importance of criteria, held constant in this case at medium, and effectiveness of
individual choices in meeting them.
As criteria are evaluated and weights assigned according to which factors are
considered to be most significant, the factors are sorted from left to right in order of
importance (i.e., the factor considered by the decision maker to be most significant in
meeting overall needs ends up in the leftmost position). In this baseline scenario, all
are equally important.
Similarly, as choices are evaluated according to effectiveness in meeting
criteria, the best choices migrate to the top of the list. When the process is complete,
the best choice should emerge at the top, the worst at the bottom.
As selection alternatives and the criteria to be used in evaluating them are
entered into the table, weights are assigned to each of the evaluation factors so that
they are ranked in order of their importance in fulfilling the overall task. For the
baseline decision "Which Insecticides Used on Alfalfa are Less Risky ?," the twelve
criteria used to evaluate the choices were all weighted as Medium. Where one or more
bird incidents were reported, that criterion was rated as "poor" for that chemical. When
no bird incidents were reported, that criterion was rated as N/A for that chemical and it
did not contribute to the summary rating. No fish incidents were reported and thus this
criterion did not contribute to the summary for any chemical.
The colors under each criteria and under the summary show four rating
categories which indicate difference between options:
Medium Green = Top Option(s)
Dark Green = Second Option(s)
Yellow = Third Option(s)
Red = Worst Option(s)
The summary column displays the results of the hidden calculations based on
the assigned ratings and weights for the criteria. For a more detailed explanation of the
underlying mathematical algorithms, Tashman and Munro [24] recommend a chapter
entitled Decisions with Multiple Objectives in the decision analysis textbook by
Goodwin and Wright [25], The higher the summary score, the better the option, and
vice versa. The scale for the summary is from 10 to 1, and is expressed as four rating
categories as above.
Among the 10 choices (chemicals) considered, 7 were considered to be "top
-------
90
options." (A top option is defined as follows: If the choice immediately following the
preferred choice is rated in the same rating category as the recommended selection,
then all choices in that category are considered top options. If the second ranking
choice is in a different category, the top options are considered to be the recommended
choice plus all choices in the same category as the second-place option. Thus, the
"top options" list will always have at least two choices in it and may include all of the
choices considered in the entire table.) DecideRight also provides pair-wise
comparisons between the choices where the critical rating factor is identified. This was
not included in this document.
For the decision of "Which Insecticides Used on Alfalfa are Less Risky ?," the
top options were:
Chemical P
Chemical N
Chemical M
Chemical I
Chemical K
Chemical D
Chemical L
After a careful evaluation of each option, Chemical P appears to be the best
choice. The worst choice was Chemical G, followed by Chemicals C and B.
Finally, the relative strengths of the various choices, i.e., chemicals, in each of
the factors is illustrated in Figure 11. The wider the band, the better the rating, i.e.,
less risk, for that criteria and for that pesticide. Missing bands indicate the worst rating
for that criteria and that pesticide.
-------
91
Figure 11. The Relative Strengths of the 10 Choices Considering the 11 Risk
Endpoints
Chemical P
Chemical N
Chemical M
Chemical I
Chemical K
Chemical D
Chemical L
:
:
~
¦
Chemical B
Chemical C
¦
¦
~
Chemical G
[] # of bird incident reports
I I % avian acute bird/day risk
n % avian chronic risk
HI % avian dietary risk
| % freshwater fish acute risk
HI % freshwater fish chronic risk
| % freshwater invertebrate acute risk
HI % freshwater invertebrate chronic risk
|~| % Marine/Estuarine Fish Acute Risk
I I % Marine/Estuarine Mollusc Acute Risk
n % Marine/Estuarine Crustacean Acute Risk
HI # of fish incident reports
C. Ecological Risk Characterization Information
-------
92
Following the initial screening analysis for the case study, which is based on
readily available data, we considered additional information that could further
characterize the potential risk of the chemicals to non-target organisms on alfalfa. This
information may be quantitative in nature such as refined estimates of aquatic exposure
based on PRZM/EXAMS modelling, updated ecotoxicity data, additional reports of bird
and fish incidents not recorded in the EMS, or measured residues in the environment. It
may also come in the form of technical or literature reports addressing the use site,
application methods, and/or vulnerable populations of non-target organisms.
Terrestrial and aquatic field studies are important information to include in a
comprehensive risk assessment for a pesticide. For this case study on alfalfa, however,
field studies had been required to support the registration for six of the ten pesticides,
but not for the other four. While most studies confirm the risk concerns raised by the
preliminary analysis, four of the pesticides lack any field testing. Since all ten cannot be
compared using this criterion, an evaluation of available field study data has not been
included in this comparative analysis.
Following is a chemical by chemical breakdown of typical information that can be
used to further characterize, and/or refine the potential risk of these 10 pesticides to
non-target organisms when used on alfalfa.
Chemical B
Two additional incidents were found:(1) In 1997 in California, 25 birds were
killed when Chemical B was sprayed on alfalfa. Residue analysis confirms Chemical B
as being the cause of the bird kill; (2) In 1991, again in California, 100+ fish were killed
after an application of Chemical B to an alfalfa field. No residue analysis was available,
however, the farmer confirmed the application of Chemical B the day prior to the
discovery of the fish kill by personnel from the California Department of Fish and
Game.
PRZM/EXAMS model results are available and show estimated residues
approximately 3X greater than the GEN EEC estimates.
Chemical C
PRZM/EXAMS model results are available and show estimated residues from 2
to 3 X lower than the GENEEC estimates.
Chemical D
PRZM/EXAMS model results are available and show estimated residues
approximately 2 X lower than the GENEEC peak estimates. However, by 21-days and
-------
93
60-days the factor increased to approximately 8X and 17X lower, respectively.
Chemical G
A 100-foot buffer from sensitive aquatic habitat is required for the use of this
chemical on alfalfa. PRZM/EXAMS model results are available. The spray drift
component of the EEC was modified. The results from the model show estimated
residues approximately 2 X lower than the GENEEC peak estimates. The 21-day
estimated were approximately equal, and the 60-day estimates were slightly greater
than the GENEEC estimates.
Chemical I
PRZM/EXAMS model results are available and show estimated residues slightly
less than the GENEEC peak estimates. However, by 21-days and 60-days the residue
estimates were approximately 7X and 12X lower, respectively.
Chemical K
New avian reproduction study results report a NOAEC of 1 ppm. This is
considerably lower than the predicted value of 23 ppm. In addition, PRZM/EXAMS
model results are available and show estimated residues approximately 3Xless than
the GENEEC peak estimates. However, by 21-days and 60-days the residue estimates
were approximately 4X lower.
Chemical L
PRZM/EXAMS model results are available and show estimated residues
approximately 7 X less than the GENEEC peak estimates. However, by 21-days and
60-days the residue estimates were approximately 16X and 20X lower, respectively.
Chemical M
PRZM/EXAMS model results are available and show estimated residues
approximately 3 X less than the GENEEC peak and 21-day estimates. However, by 60-
days the residue estimates were just slightly less than the GENEEC results.
Chemical N
PRZM/EXAMS model results are available and show estimated residues
approximately 11 X less than the GENEEC peak, 21-day, and 60-day estimates.
-------
94
Chemical P
PRZM/EXAMS model results are available and show estimated residues
approximately 6 X less than the GENEEC peak and 21-day estimates. By 60-days the
residue estimates were approximately 4 X less than the GENEEC results.
D. Scenario #1 - Baseline Plus Ecological Risk Characterization Information
The software has a scenario function where the weights and ratings of the
criteria in the baseline can be changed to answer "What if...?" questions. Thus, we
asked "What if the ratings of the criteria change based on the additional information in
Section C above? Then, "Which are the insecticides used on alfalfa are less risky?"
For this scenario, we modified the criteria ratings for % risk and # of incident reports
using the information in Table 12. The results for Scenario #1 are presented in Figure
12. below.
-------
95
Figure 12. Scenario #1: Decision Table for Which Insecticides Used on Alfalfa are Less Risky
# of bird incident reports
# of fish incident reports
% avian acute bird/day risk
% avian chronic risk
% avian dietary risk
% freshwater fish acute risk
% freshwater fish chronic risk
% freshwater invertebrate acute risk
% freshwater invertebrate chronic risk
% Marine/Estuarine Crustacean Acute Risk
% Marine/Estuarine Fish Acute Risk
% Marine/Estuarine Mollus
Summary
Chemical P
N/A
N/A
0.00
1.00
0.10
0.00
0.00
0.00
0.00
0.00
0.00
0.00
9.97
Chemical M
N/A
N/A
0.00
0.00
0.00
0.00
0.00
3.00
0.00
1.00
0.00
0.00
9.93
Chemical N
N/A
N/A
0.10
3.00
0.10
0.00
0.00
0.00
0.00
0.00
0.00
0.00
9.90
Chemical I
N/A
N/A
0.00
0.00
0.00
1.00
0.00
2.00
1.00
3.00
0.00
0.00
9.89
Chemical K
N/A
N/A
0.80
22.00
0.50
4.00
0.00
0.00
0.00
6.00
0.00
0.00
9.15
Chemical D
Poor
N/A
0.40
2.00
0.10
0.00
0.00
0.00
0.00
0.00
0.00
0.00
9.03
Chemical L
N/A
N/A
0.90
12.00
78.80
0.00
0.00
2.00
0.00
1.00
0.00
0.00
8.56
Chemical C
N/A
N/A
3.90
20.00
8.20
7.00
4.00
9.00
12.00
24.00
22.00
0.00
8.08
Chemical B
Poor
Poor
0.00
7.00
0.00
85.00
55.00
28.00
2.00
7.00
78.00
0.00
5 21
Chemical G
Poor
N/A
93.30
32.00
10.70
1.00
40.00
56.00
86.00
58.00
0.00
100.00
2.84
-------
96
Table 12. Data for Scenario #1
Pesticides Used on Alfalfa as Post-Emergent Sprays
(Includes Risk Characterization Information; Compare with Table 5)
Birds
Chemical
name
% Avian Acute
Bird/Day Risk
% Avian
Chronic Risk
% Avian Dietary
Risk
% Acres Treated
C
3.9%
20%
8.2%
29.7%
D
0.4%
2%
0.1%
26.3%
G
93.3%
32%
10.7%
16.0%
I
0.0%
0%
0.0%
12.2%
L
0.9%
12%
78.8%
11.6%
P
0.0%
1%
0.1%
2.8%
K
0.8%
22%
0.5%
0.6%
M
0.0%
0%
0.0%
0.4%
B
0.0%
7%
0.0%
0.3%
N
0.1%
3%
0.1%
0.0%
Aquatic
Organisms
Chemical
name
% Freshwater
Fish
Acute Risk
% Freshwater
Fish
Chronic Risk
%
Marine/Estuarine
Fish Acute Risk
% Freshwater
Invertebrate
Acute Risk
% Freshwater
Invertebrate
Chronic Risk
%
Marine/Estuarine
Crustacean
Acute Risk
%
Marine/Estuarine
Mollusc Acute
Risk
% Acres
Treated
C
7%
4%
22%
9%
12%
24%
0.0%
29.7%
D
0%
0%
0%
0%
0%
0%
0.0%
26.3%
G
1%
40%
0%
56%
86%
58%
100.0%
16.0%
1
1%
0%
0%
2%
1%
3%
0.0%
12.2%
L
0%
0%
0%
2%
0%
1%
0.0%
11.6%
P
0%
0%
0%
0%
0%
0%
0.0%
2.8%
K
4%
0%
0%
0%
0%
6%
0.0%
0.6%
M
0%
0%
0%
3%
0%
1%
0.0%
0.4%
B
85%
55%
78%
28%
2%
7%
0.0%
0.3%
N
0%
0%
0%
0%
0%
0%
0.0%
0.0%
-------
97
Incidents
JChemical
name
# Reports ¦
Birds
% Total
Reports
Birds
# Reports - Fish
% Total Reports ¦
Fish
C
0
0%
0
0%
D
3
50%
0
0%
G
1
3%
0
0%
1
0
0%
0
0%
L
0
0%
0
0%
P
0
0%
0
0%
K
0
0%
0
0%
M
0
0%
0
0%
B
1
17%
1
0%
N
0
0%
0
0%
-------
98
The addition of the risk characterization information provided results which are
very similar to the baseline. After a careful evaluation of each option in Scenario #1,
once again Chemical P appears to be the best choice. This time there were eight top
choices, with the addition of Chemical C. In the baseline, Chemical C followed the
worst choice, Chemical G. For Scenario #1, the worst choice again was Chemical G,
but his time it is followed by Chemical B.
E. Additional Scenarios - Changes in Criteria Importance
As noted previusly, one of the simplifications of the methodology used in the
decision analysis software is that there are no uncertainties. Certainly comparative
ecological risk analysis involves much uncertainty. Since the software permits the user
to deal with some uncertainty is by running multi-scenarios, we chose to run seven
additional scenarios for the case study. These runs show how sensitive the differences
between pesticides are to change in the importance of the criteria. The results can add
confidence to overall conclusions. The additional scenarios are listed below. The
summary columns are presented in Figure 13 below. The different scenarios reflect
changes in the importance given to various endpoints and groups of endpoints. In two
scenarios, certain criteria are considered to be unimportant and do not contribute to the
comparison. They reflect differences in importance that can typically arise during
ecological risk assessments, and during discussion with risk managers.
Scenario #2 -
Scenario #3 -
Scenario #4 -
Scenario #5 -
Scenario #6 -
Scenario #7 -
Additional Risk Characterization Data Added; Bird
Endpoints are Rated High; All Other Endpoints are Rated
Medium
Additional Risk Characterization Data Added; Aquatic
Emdpoints are Rated High; All Other Endpoints are Rated
Medium
Additional Risk Characterization Data Added; Bird & Fish
Incidents are Rated as N/A; All Other Endpoints are Rated
Medium
Additional Risk Characterization Data Added;
Marine/Estuarine Endpoints are Rated as N/A; All Other
Endpoints are Rated Medium
Additional Risk Characterization Data Added; Fish
Endpoints are Rated High; All Other Endpoints are Rated
Medium
Additional Risk Characterization Data Added; Aquatic
Invertebrate Endpoints are Rated High; All Other Endpoints
-------
99
are Rated Medium
Scenario #8 -
Scenario #9 -
Additional Risk Characterization Data Added; Bird
Endpoints are Rated High; Aquatic Endpoints are Rated
Low
Additional Risk Characterization Data Added; Aquatic
Endpoints are Rated High; Bird Endpoints are Rated Low
The results in Figure 13 show that Chemicals P, N, M, I, K, and D were
consistently rated as top options regardless of the changes in importance values for the
twelve criteria. Chemical P was the top option for all scenarios except #8 where bird
endpoints were given a high importance and aquatic endpoints a low importance. In
this scenario, Chemical M was the top choice, closely followed by Chemical P.
Chemical C was a top option except in the baseline option. Chemical L was a top
option in all scenarios exept scenario #8. Alternately, Chemical G is consistently rated
at the bottom as the worst option in all scenarios. Chemical B was the second worst
option except for the baseline scenario where Chemical B and C were rated equal as
the second worst option. In scenario #8, Chemicals B and L were equal as the second
worst option. Chemical D made the largest change within the top options in scenario
#4 where incidents were rated as unimportant.
-------
100
Figure 13.
Scenario
Summary
3aseline
Scenario
Scenario #1
Scenario #2
Scenario #3
Scenario #4
Scenario #5
Scenario #6
Scenario #7
Scenario #8
Scenario #9
Chemical P
3.94
Chemical P
3.97
Chemical P
3.95
Chemical P
3.98
Chemical P
3.97
Chemical P
3.95
Chemical P
3.97
Chemical P
3.98
Chemical M
3.96
Chemical P
3;9
Chemical N
3.89
Chemical N
3.93
Chemical M
3.95
Chemical N
3.94
Chemical D
3.93
Chemical M
3.92
Chemical N
3.96
Chemical N
3.93
Chemical P
3.94
Chemical N
Chemical M
3.89
Chemical M
3.90
Chemical I
3.91
Chemical M
3.92
Chemical M
3.93
Chemical I
3.92
Chemical M
3.93
Chemical M
3.90
Chemical I
3.93
Chemical M
Chemical I
3.81
Chemical I
3.89
Chemical N
3.85
Chemical I
3.87
Chemical N
3.90
Chemical N
3.86
Chemical 1
3.91
Chemical I
3.85
Chemical N
3.82
Chemical I
Chemical K
3.25
Chemical K
3.15
Chemical K
3.80
Chemical K
3.41
Chemical 1
3.89
Chemical K
3.92
Chemical K
3.31
Chemical K
3.32
Chemical K
3.59
Chemical K
Chemical D
3.04
Chemical D
3.03
Chemical D
3.58
Chemical D
3.41
Chemical K
3.15
Chemical D
3.66
Chemical D
3.24
Chemical D
3.29
Chemical D
3.31
Chemical D
Chemical L
3^33
Chemical L
3.56
Chemical C
7.93
Chemical L
3.13
Chemical L
3.56
Chemical C
3.24
Chemical L
3.89
Chemical L
3.93
Chemical C
M34
Chemical L
Chemical B
7.47
Chemical C
3.08
Chemical L
7.83
Chemical C
3.19
Chemical C
3^08
Chemical L
7.97
Chemical C
3.19
Chemical C
3.12
Chemical L
7.Z7
Chemical C
Chemical C
>.51
Chemical B
5.21
Chemical B
5.65
Chemical B
1.86
Chemical B
3.14
Chemical B
1.86
Chemical B
3.98
Chemical B
>.02
Chemical B
5.91
Chemical B
n
Chemical G
2.57
Chemical G
2.48
IHitJullATig 3EE1
pHW.'.UAnq area
grim |
Chemical G
yj
Baseline Scenario - All Criteria Medium Importance
Additional Scenario #1 - Additional Risk Characterization Information Added; All Criteria Medium Importance
Additional Scenario #2 - Additional Risk Characterization Information Added; Bird Endpoints High Importance; All Other Medium
Additional Scenario #3 - Additional Risk Characterization Information Added; Aquatic Endpoints High Importance; All Other Medium
Additional Scenario #4 - Additional Risk Characterization Information Added; Bird & Fish Incidents rated as N/A; All Other Medium
Additional Scenario #5 - Additional Risk Characterization Information Added; Marine/Estuarine Endpoints Rated as N/A; All Other Medium
Additional Scenario #6 - Additional Risk Characterization Information Added; Fish Endpoints High Importance; All Other Medium
Additional Scenario #7 - Additional Risk Characterization Information Added; Aquatic Invertebrate Endpoints High Importance; All Other
Medium
Additional Scenario #8 - Additional Risk Characterization Information Added; Bird Endpoints High; Aquatic Endpoints Low
Additional Scenario #9 - Additional Risk Characterization Information Added; Aquatic Endpoints High; Bird Endpoints Low
-------
101
In summary, the insecticides used on alfalfa as postemergent sprays which are potentially the
least risky based on the comparison using DecideRight software, including multi-scenarios, are these six:
Chemical P
Chemical N
Chemical M
Chemical I
Chemical K
Chemical D
Chemical P is the top option except where bird endpoints are rated high and aquatic endpoints rated low.
In this case, Chemical M is the top option. Alternately, Chemical G is the worst option across all
scenarios, and Chemical B is consistently rated as the second worst option.
This summary is similar to the summary based on the previous graphical analysis above:
"In summary for postemergent sprays on alalfa, Chemical G stands out as presenting the greatest
potential acute and chronic risk to birds, and aquatic invertebrates, and chronic risk to fish. The
acute risk to birds appears to be supported by and incident report. Chemical B presents the
greatest acute risk to freshwater fish, Chemical C presents the greatest acute risk to
marine/estuarine fish, and Chemical L presents the greatest dietary risk to birds. The bird incident
reports raises a question concerning the comparatively low risk of Chemical D to birds. Overall,
Chemical N, Chemical M, and Chemical P appear to be comparatively less risky than the others."
When the use of the pesticides on alfalfa is considered (See Figure 3), we find that Chemical C
and Chemical D lead based on current estimates of % acres treated with greater than 20% each
followed by Chemical G, Chemical I, and Chemical L at greater than 10% each. Considering the results
from the two previous analyses, the overall risk to non-target organisms from insecticides used on alfalfa
could be reduced by reducing the use of Chemical G.
-------
102
F. Limitations of this Analysis
This proposed methodology for comparing potential ecological risk from the use of pesticides, is
intended to identify those pesticides with large differences in risk. Twelve endpoints were considered;
however, other endpoints such as effects on wild mammals, non-target plants, non-target insects,
reptiles and amphibians were not included. These could be included in a more comprehensive analysis.
The available ecotoxicity and environmental fate data were limited, so that in some cases missing
values had to be estimated so that the comparison could proceed. The tools used to estimate missing
values is not perfect as shown by the later submission of an avian chronic NOAEC value of 1 ppm for
Chemical K, which differed significantly from the estimated value of 23 pm.
The exposure estimates were based on maximum label rates and one-half those rates. Ideally,
typical rates would be better information to use for comarison. It is interesting to note that while refined
aquatic EEC estimates based on PRZM/EXAMS were in most cases 2X to 20 X lower than the GENEEC
EECs, the overall top options and worst options did not change significantly (compare the summary for
the Baseline Scenario and Scenario #1 in Figure 13).
This analysis employs a number of new calculations for expressing potential risk, such as the %
contribution to the RQ sum, the frequency of RQ exceedance, the % contribution to the Time to RQ=1
sum, and % risk. These have not been used beyond this analysis for these 17 pesticides. Thus, we do
not know how useful they will prove to be in expressing risk for other groups and combinations of
pesticides.
We have already identified a number of limitations of the decision support software, DecideRight
and others are identified by Tashman and Munro [24], Comparions with other decision support products
have not yet been made.
-------
103
IX. REFERENCES
1. Memorandum from Linda J. Fisher, Assistant Administrator, Office of Pesticides
and Toxic Substances, to Douglas Campt, Director, Office of Pesticide
Programs, March 26.1992. Establishment of Ecological and
Environmental Fate Data Requirements Task Force
2. Memorandum from Linda J. Fisher, Assistant Administrator, Office of Pesticides
and Toxic Substances, to Douglas Campt, Director, Office of Pesticide
Programs, October 29, 1992. Decisions on the Ecological, Fate, and
Effects Task Force.
3. U.S. EPA. 1992 (March). Comparative Analysis of Acute Avian Risk from
Granular Pesticides. U.S. EPA, Office of Pesticide Programs.
4. U.S. EPA. 1994 (September). Avian Granular Risk Reduction Initiative -
Progress Report. U.S. EPA, Office of Pesticide Programs.
5. OPP Ecotoxicity Data Base, 1998. Environmental Fate & Effects Division, Office
of Pesticide Programs [7507C], Washington, D.C. 20460
6. Driver, C.J., M.W. Ligotke, P. Van Voris, B. D. Greenspan and D. B. Brown.
1991. Routes of uptake and their relative contribution to the toxicologic
response of northern Bobwhite (Colinus virginianus) to an
organophosphate pesticide. Environ. Toxicol. Chem. 10: 21-33.
7. Erbach, D. and J. Tollefson. 1983. Granular insecticide application for corn
rootworm control. Trans. ASAE-1983. Amer. Soc. Aqric. Eng. 0001-
2351/83/2603-0696.
8. DeWitt, J.B. 1966. Methodology for determining toxicity of pesticides to wild
vertebrates. J. Applied Ecol. 3 (suppl.):275-278.
9. U.S. EPA. 1977a. Memorandum, dated 9 September 1977, to Environmental
Safety Section, OPP, U.S. EPA: Classification of granulated formulations,
from R. Felthousen.
10. U.S. EPA. 1986. Hazard Evaluation Division Standard Evaluation Procedure
Ecological Risk Assessment. Office of Pesticide Programs, Washington,
D.C. EPA 540/9-86-167.
-------
104
11. U.S. EPA. 1989. Technical Support Document for the Special Review of
Carbofuran. January 1989.
12. Hoerger, F. and E. E. Kenaga. 1972. Pesticide Residues on Plants: Correlation
of Representative Data as a Basis for Estimation of Their Magnitude in
the Environment. 1N:F. Coulston and F. Corte, eds., Environmental
Quality and Safety: Chemistry. Toxicology, and Technology. Vol I. Georg
Thieme Publishers, Stuttgart, West Germany, pp. 9-28.
13. Fletcher, J.S., J.E. Nellesson and T.G. Pfleeger. 1994. Literature Review and
Evaluation of the EPA Food-Chain (Kenaga) Nomogram, an Instrument
for Estimating Pesticide Residues on Plants. Environ. Tox. and Chem.
13(9): 1383-1391.
14. Willis, G. H. and L. L. McDowell. 1987. Pesticide persistence on foliage. Rev.
Environ. Contam. & Tox. 100: 23-73.
15. Rattner, B. A., L. Sileo, and C. G. Scanes. 1982. Oviposition and the Plasma
Concentrations of LH, Progesterone and Corticosterone in Bobwhite Quail
(Colinus virginianus) fed Parathion. J. Reprod. Fert. 66: 147-155.
16. Bennett, J. K. and R. S. Bennett. 1990. Effects of Dietary Chemical L on
Northern Bobwhite Egg Production and Eggshell Quality. Environ.
Toxicol. Chem. 19: 907-912.
17. Bennett, R. S. and L. M. Ganio. 1991. Overview of Methods for Evaluating
Effects of Pesticides on Reproduction in Birds. U. S. EPA. Environmental
Research Laboratory, Corvallis, OR. EPA 600/3-91/048.
18. Hill, E. F. 1994. Acute and Subacute Toxicology in Evaluation of Pesticide
Hazard to Avian Wildlife. IN: R. J. Kendall and T. E. Lacher, Jr., eds.,
Wildlife Toxicology and Population Modeling: Integrated Studies of
Agroecosvstems. Special Publication of Society of Environmental
Toxicology and Chemistry. Lewis Publishers, Boca Raton, FL. pp. 207-
226 (Chapter 22).
19. Nagy, K. A. 1987, Field metabolic rate and food requirement scaling in mammals
and birds. Ecol. Monogr. 57:111-128.
20. U. S. EPA. 1993. Wildlife Exposure Factors Handbook. Volume I of II.
EPA/600/R-93/187a. Office of Research and Development, Washington,
-------
105
D C. 20460. p. 3-4.
21. R. D. Parker, H. P. Nelson, R. D. Jones*, 1995. GENEEC: A Screening Model for
Pesticide Environmental Exposure Assessment. The International
Symposium on Water Quality Modeling, American Society for Agricultural
Engineering. Orlando, Florida, 1995.
22. U.S. EPA. 1994 (July). Ecological Incident Information System. U.S. EPA, Office
of Pesticide Programs.
23. Brassard, C. A., L. P. Kendrick, S. M. Lees, and A. F. Maciorowski. In press. An
overview of ecological incident monitoring and reporting. Society of
Environmental Chemistry and Toxicology.
24. Tashman, L. and S. Munro. 1997 (January 15, 1997). Software Review -
DecideRight. In The Forum, the newsletter of the International
Associaction of Business Forecasting.
[http://www. loyola.edu/iabf/form. htm]
25. Goodwin, P. and G. Wright. 1991. Decision Analysis for Management
Judgement. John Wiley & Sons, publisher.
26. Carsel, R.F., C.N. Smith, L.A. Mulkey, J.D. Dean and P. Jowise. 1984. Users
manual for pesticide root zone model (PRZM): Release 1, Rep.EPA-
600/3-84-109, 219 pp. Environmental Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency,
Athens, GA.
27. Mullins, J.A., R.F.Carsel, J.E. Scarbrough and A.M. Ivery. 1993. PRZM2 users
manual version 1.0, Environmental Research Laboratory, Office of
Research and Development, U.S. Environmental Protection Agency,
Athens GA.
28.
Burns, L.A. 1991. Exposure analysis modeling system: users guide for EXAMS II
version 2.94, Environmental Research Laboratory, Office of Research
and Development, U.S. Environmental Protection Agency, Athens, GA.
------- |