Ecological Risk Assessment for Use of Fomesafen On Cotton (DP302766), Snap Beans ( DP314014) and Dry Beans (DP314112) January 2006

Transcription

1 Ecological Risk Assessment for Use of Fomesafen On Cotton (DP302766), Snap Beans ( DP314014) and Dry Beans (DP314112) January

2 Table of Contents 1. Executive Summary Nature of Stressor Potential Risk Endangered Species Assessment Problem Formulation Stressor Source and Distribution Environmental Fate Summary Mode of Action Use Characterization Assessment Endpoints Ecosystems Potentially at Risk Ecological Effects Conceptual Model Risk Hypotheses Key Uncertainties and Information Gaps Exposure Assessment Label Application Rates and Intervals Aquatic Exposure PRZM-EXAMS Modeling Inputs and Scenario Selection PRZM-EXAMS Modeling Output Registrant-submitted Aquatic Exposure Modeling SCIGROW Modeling for Ground Water Soil Accumulation Bird and Mammal Exposure (TREX) Terrestrial Plant Exposure (TerrPlant & AgDrift) TerrPlant AgDrift Effects Assessment Aquatic Guideline Data Aquatic Data from ECOTOX Terrestrial Plant Guideline Data Avian and Small Mammal Guideline Data Terrestrial Insect Data Terrestrial Data from ECOTOX Incident Database Review

3 5. Risk Estimation Aquatic RQ Summary Terrestrial RQ Summary Terrestrial Plants TerrPlant AgDrift Avian RQ Summary Small Mammal RQ Summary Terrestrial Insects Risk Description Aquatic Risk Terrestrial Risk Birds and Mammals Application Rate of 0.50 lb ai/a Application Rate of lb ai/a Application Rate of 0.2 lb ai/a (Alternative) Plants and Insects Endangered Species Aquatic Endangered Species Plants Birds, Reptiles & Amphibians Mammals Probability of Inidividual Effect

4 List of Tables Table 1 Endangered Species Occurring in the Same County as Proposed Crops... 8 Table 2 Label Application Rates and Intervals for Post-Emergent Fomesafen Use on Snap Beans and Dry Beans Table 3 Label Application Rates and Techniques for Fomesafen Use on Cotton Table 4 Input Parameters for PRZM-EXAMS Modeling of Fomesafen on Cotton and Soybeans Table 5 PRZM-EXAMS EECs for Fomesafen at lb a.i/a Table 6 PRZM-EXAMS EECs for Fomesafen at 0.50 lb ai/a Table 7 1-in-10 year Estimated Environmental Concentrations of Fomesafen from Cotton in LA Table 8 1-in-10 year Estimated Environmental Concentrations of Fomesafen from Soybeans in NE and MS Table 9 1 in 10 year Estimated Environmental Concentrations for Fomesafen from Snapbeans in WI Table 10 Input Parameters for SCIGROW Modeling for Fomesafen Table 11 Bird Dose Estimates Table 12 Mammal Dose Estimates Table 13 Terrestrial Plant Exposure Table 14 Estimated Point Deposition of Fomesafen from AgDrift... (Tier I, Ground Application, Low Boom, 90% Percentile Estimate) Table 15 Estimated Point Deposition of Fomesafen from AgDrift... (Tier I, Aerial Application, 90% Percentile Estimate) Table 16 Acute Aquatic Data from Registrant-submitted Studies Table 17 Chronic Aquatic Data from Registrant-submitted Studies Table 18 Terrestrial Plant Guideline Data Table 19 Avian and Small Mammal Guideline Data from Acute Studies Table 20 Avian and Small Mammal Guideline Data from Chronic Studies Table 21 Summary of Aquatic RQs Table 22 Terrestrial Plant Risk Quotients Based on TerrPlant Table 23 Required Distance from Application to Fall Below LOC Table 24 Avian RQ Summary 0.5 lb ai/a Table 25 Avian RQ Summary: lb ai/a Table 26 Avian RQ Summary 0.2 lb ai/a (alternative) Table 27 Small Mammal RQ Summary: 0.50 lb ai/a Table 28 Small Mammal RQ Summary: 0.375lb ai/a Table 29 Small Mammal RQ Summary: 0.2 lb ai/a (alternative) Table 30 Endangered Plants by Crop and Species Table 31 Endangered birds potentially at risk from fomesafen Table 32 Endangered reptiles and amphibians potentially at risk from fomesafen Table 33 Endangered mammals potentially at risk from fomesafen

5 List of Figures Figure 1 Figure 2 Figure 3 Figure 4 Conceptual Model for Fomesafen 16 Soil Accumulation of Fomesafen Aerial Application of Fomesafen Ground Application of Fomesafen...36 List of Maps Map 1 Map 2 Map 3 Acres of Crops by County, Cotton 12 Acres of Crops by County, Snap Beans 13 Acres of Crops by County, Dry Beans.. 14 Appendices Appendix A Appendix B Appendix C Appendix D Appendix E Appendix F Appendix G Appendix H Environmental Fate Studies PRZM EXAMS Output TREX Output TerrPlant Output Ecological Effects Data EIIS Output RQ Method and LOCs Endangered and Threatened Species 5

6 1. Executive Summary This ecological risk assessment addresses the proposed use of fomesafen (PC Code ) on cotton, dry beans, and snap beans. Fomesafen is recommended as a pre- or post-emergent herbicide on cotton and a post-emergent herbicide on dry beans and snap beans. Application rates and techniques are dependent on geographic regions, soil textural classification, and application timing. The methods of application assessed include aerial spray (0.375 lb ai/a) and ground spray (0.50 lb ai/a). An alternative ground spray application rate (0.2 lb ai/a) was also considered in the assessment. As expected for an herbicide, the greatest acute risk associated with fomesafen use was for non-target terrestrial plants. Due to the high toxicity of the compound to terrestrial plants and proposed high application rates, spray drift buffers will likely have a minimal impact on reducing risk to non-target plants unless they are extremely wide (>900 ft for aerial applications, >350 ft for ground applications). Another significant concern is the persistence and mobility of fomesafen in both soil and aquatic environments. The persistence of fomesafen in soil is expected to prolong phytotoxic effects to non-target plants. Environmental fate properties are also expected to favor fomesafen movement into ground and surface waters. Use of these waters for irrigation may pose a risk to nontarget plant species. 1.1 Nature of Stressor Fomesafen is an herbicide. It is applied as a foliar spray (both pre-emergent and postemergent) for control of broad-leaved weeds, grasses, and sedges. Mode of action is via cellular membrane disruption. It is highly persistent in soil ( days, dependent on soil type) resulting in a potential for accumulation in terrestrial environments. The label suggests not planting sensitive crops in a fomesafen-treated field for a 3-18 month period, due to the persistence of fomesafen in the soil. Additionally, it is highly mobile, and is expected to leach into groundwater and be transported from the site via runoff into surface waters. Based on physical properties, bioaccumulation and long-range transport are not expected to be of concern. It is extremely toxic to terrestrial plants, especially dicots, but of fairly low acute toxicity to fish and wildlife. Some chronic reproductive effects have been noted in mammals, and may also occur in birds. No major degradates of toxicological concern have been identified. 1.2 Potential Risk Terrestrial Plants Non-target organisms most at risk from fomesafen use are terrestrial plants near the use site(s). Disruption of plant communities could have effects on wildlife communities near the use site. The composition and health of plant communities determine quantity and type of wildlife present. Plant communities serve both as shelter and food. 6

7 Aerial applications are of particular concern. Based on modeling (AgDrift), plants at distances greater than 1000 ft away from the center of the flight line could be affected. This may include sensitive and high-value plant communities such as neighboring agricultural crops, wetlands, and riparian corridors. The majority of reported incidents are damage to crops, either as a result of spray drift or due to accidental misuse. Effects from ground-boom applied treatments do not extend as far, but at the proposed ground application rate (0.50 lb ai/a), point deposition from a low boom configuration does not drop below the acute risk level of concern until 350 ft away from the application source. Because of the range of these potential effects, two alternative ground application rates (0.375 lb ai/a and 0.2 lb ai/a) were also modeled. These two rates were selected based on the proposed aerial application rate, and the AgDrift estimate of herbicide deposition directly below the application source. Based on registrant-submitted data, these rates appear to be efficacious against some nuisance plants (Appendix E). Plant acute risk RQs based on these alternative application rates drop below the level of concern at 250 ft away from the application source (0.375 lb ai/a) and at 60 ft away from the application source (0.2 lb ai/a). Terrestrial Animals On an acute risk basis, fomesafen is practically non-toxic to birds and mammals. Potential risk to birds and mammals is evaluated based on the estimated pesticide residues on food items. No acute risk LOCs were exceeded at any application rate. The endangered species acute risk LOC was exceeded for small (15g) and medium (35g) mammals consuming short grass at the highest proposed application rate (0.50 lb ai/a), and for small (15 g) mammals consuming short grass at the lower proposed application rate (0.375 lb ai/a). Some medium-sized, but no small-sized endangered herbivorous mammals occur in counties where the proposed crops (cotton, dry beans, and snap beans) are grown. Reduction of the application rate to 0.2 lb ai/a eliminates all acute risk exceedences. At all application rates, there are exceedences of the chronic risk LOC for both birds and mammals. At the highest application rate, there are chronic exceedences for organisms consuming short grass, tall grass, broadleaf plants and small insects (i.e., all herbivores and insectivores). Chronic effects noted in the mammal study included a decrease in the number of young born live, and the number of young surviving, thus there is a potential for decreased fecundity in wildlife populations on or near the treated site as a result of fomesafen application. The guideline bird study currently available for fomesafen establishes a chronic no-effects level, but not a lowest effects level. Thus, the impact of exceeding the chronic level of concern for birds cannot be determined with any certainty. Reduction of the application rate to 0.2 lb ai/a decreases the exposure enough so only the birds and mammals consuming short grass exceed the chronic LOCs. Aquatic Plants and Animals Fomesafen is practically non-toxic to slightly toxic to aquatic animals. No-effects concentrations observed in laboratory toxicity tests for the most sensitive aquatic plants are higher than expected environmental concentrations based on modeled uses. No levels of concern were exceeded for aquatic animals at any application rate. 7

8 Chemical Properties Fomesafen is highly persistent and mobile in soil. These environmental fate properties are expected to promote year-to-year accumulation in soil as well as off-site movement by leaching and runoff. Because fomesafen is spray applied, there is a potential for drift onto non-target plants. Prolonged phytotoxic effects on non-target plants are expected based on the persistence of fomesafen. EFED has no data from which to determine how long it may be toxic, but the labels recommend not planting sensitive crops in the use site for 3 to 18 months following treatment. 1.3 Endangered Species Assessment As an initial screening to determine what endangered species may be affected by proposed uses of fomesafen, EFED used LOCATES (Ver ) to determine what species are known to occur in counties where the proposed crops are grown. Table 1 shows the number of unique species (i.e., species are counted only once, even if they are affected in multiple counties). Although it varies by use, some states contribute significantly to the total number of species, especially California, Florida, and Hawaii. Detailed information is contained in Appendix H. Table 1 Endangered Species Occurring in the Same County as Proposed Crops Taxa Cotton Snap Beans Dry Beans All Uses Plants Monocots Dicots Animals Amphibians Birds Reptiles Mammals Direct Effects -Plants To determine what plants may be affected would require a spatial analysis to identify proximity of the plants to potential use sites. EFED does not currently have the data and tools required for this type of analysis. Minimizing potential spray drift zones will offer some protection. Direct Effects Animals Based on categorization of animals into the diet and size classes used by EFED to develop risk quotients, the list of endangered species potentially at risk from fomesafen narrowed considerably. Based on this categorization, no species are at acute risk. At the highest application rate (0.50 lb ai/a), 28 species were identified as potentially being at risk for chronic effects. At the lower proposed application rate (0.375 lb ai/a), 23 species were identified as potentially being at risk for chronic effects. At the alternative rate (0.2 lb ai/a), only 5 species appear to be at risk for chronic effects. A spatial analysis would assist in evaluating the likelihood of exposure. 8

9 Indirect Effects Plants Pesticide-mediated indirect effects to plants are usually the loss of an important pollinator or dispersal species. Given the relatively low toxicity of fomesafen to animals in general, the likelihood of these types of effects appears to be extremely low. Some endangered plant species may require the presence of other non-endangered plant species to create a suitable habitat. Fomesafen effects on plant habitat could constitute an indirect effect, but without detailed information on the plant s life history and distribution, EFED is unable to even qualitatively evaluate this type of effect. Indirect Effects - Animals The most likely indirect effect on animals is modification of habitat as a result of damage to plants. The habitat modification could include reduced food supply, reduced locations for nesting or burrowing, and/or reduced cover for predator avoidance. EFED is currently unable to evaluate the likelihood or magnitude of such an effect. 9

10 2. Problem Formulation Problem formulation provides a strategic framework for the risk assessment. By identifying the important components of the problem, it focuses the assessment on the most relevant chemical properties, exposure routes, and endpoints. 2.1 Stressor Source and Distribution Proposed new uses for fomesafen are as a pre-plant, pre-emergence, and post-emergence, herbicide for use on broadleaf weeds, grasses, and sedges in snap beans, dry beans, and cotton. Distributions of these crops were determined based on AgCensus data. Source of the stressor includes application via ground and aerial sprays. Maps showing these distributions are shown in the Use Characterization section. Because fomesafen is persistent in soil (soil degradation t 1/2 = weeks at 0.50 lb ai/a application rate, (MRID )), residual in soil of the treated field is considered in this risk assessment as a secondary source. 2.2 Environmental Fate Summary Major routes of fomesafen dissipation are leaching, runoff, and microbial degradation. Because fomesafen is persistent and mobile in soil, it is expected to move from the application site into groundwater and surface water. Additionally, off-site movement of fomesafen is expected through spray drift from aerial and ground spray. The high persistence of fomesafen is expected to contribute to year-to-year accumulation in terrestrial and aquatic environments. Fomesafen is stable to abiotic hydrolysis. It undergos slow photodegradation in water (t 1/2 = 49 to 289 days). Fomesafen is persistent (t 1/2 =9 to 99 weeks) in aerobic soil and aquatic environments. However, it degrades rapidly (t 1/2 < 20 days) in anaerobic environments. The major degradation product of fomesafen is 5-(2-chloro-α,α,αtrifluoro-p-tolyloxy)-N-methylsulphonyl-panthranilamide (fomesafen amine). A minor degradation product is 5-(2-chloro-α,α,α-trifluoro-p-tolyloxy) anthranilic acid (fomesafen amino acid). Neither degradate has been identified as a toxicological concern. Fomesafen is expected to be very mobile in soil. Simple partitioning coefficients range from 0.51 in loamy coarse sand to 2.45 in sandy clay loam soil. Regression analysis indicates fomesafen sorption is not dependent on soil organic matter content. Aged soil column leaching studies indicate degradation products of fomesafen are not mobile in soils; less than 0.06% of applied radioactivity was detected in the leachate samples. Field dissipation studies in NC, IL, MS, AR, AL, TX, LA, SD, MN, KY, IA and MO indicate fomesafen is moderately persistent to persistent (t 1/2 = 50 to 150 days ) in surface soils under actual use conditions. Fomesafen was detected at depths up to 30 inches in the soil profile. Fomesafen amine was the only degradation product identified in field 10

11 dissipation studies. Prospective ground water monitoring in NC indicates fomesafen moved through the soil profile into medium and deep ground water. Fomesafen has a low potential for bioaccumulation in fish tissues. Bioaccumulation factors for fosmesafen were 0.7 for whole fish, 0.2 for edible tissues, and 5.2 for nonedible tissue. Bioaccumulated residues were depurated during a 14-day depuration period. 2.3 Mode of Action Fomesafen is a diphenylether. It disrupts the cell membrane of the plant ( by penetrating into the cytoplasm and causing formation of peroxides and free electrons ( The specific mode of action is inhibition of protoporphyrinogen oxidase ( Fomesafen generally acts quickly, and does not translocate. It has both foliar and soil activity. Other herbicides in this group include aciflourfen, lactofen, and oxyfluorfen. 2.4 Use Characterization Fomesafen is being proposed as a pre-plant, pre-emergence, and post-emergence herbicide for use on broadleaf weeds, grasses, and sedges, in snap beans, dry beans, and cotton. Methods of application are ground spray (0.5 lb ai/a, cotton) and aerial spray (0.375 lb ai/a, dry beans, snap beans, and cotton). Application is limited to once a year, or in alternate years, depending on location. Application rates are regionally specific. Maps 1, 2, and 3 show the locations of these crops according to USDA crop data. Cotton Cotton is grown in the southern U.S. with the greatest concentrations in four major areas: the eastern coastal plain (North Carolina to Alabama), the lower Mississippi River Valley (Arkansa, Tennessee, Mississippi and Louisiana), Texas (panhandle and hill country), and California (Central Valley). Dry Beans Dry beans are grown primarily in six areas: the upper Great Plains region (Minnesota, North Dakota), the western central Great Plains (western Kansas and Nebraska, eastern Colorado), western New York, central Michigan, Washington s Yakima Basin, and the California Central Valley. USDA classifies dry lima beans as a separate commodity group, and they are grown primarily in California s Central Valley. For the purpose of this risk assessment, dry lima beans have been included in the dry bean group. Snap Beans Snap beans are grown in a wide variety of locations, although certain areas do appear to have greater concentrations of snap bean culture than others. Areas of higher concentration include the eastern coastal plain in Maryland, Delaware, and New Jersey; Michigan and Wisconsin; Washington s Yakima Valley, Oregon s Willamette Valley, and the California Central Valley. 11

12 12

13 13

14 14

15 2.5 Assessment Endpoints Assessment endpoints are selected based on ecosystems typically at risk from agricultural pesticide applications. Specific ecological effects are evaluated based on toxicity information from guideline tests, and focus on the general categories of survival, growth, and reproduction. EFED currently does not assess behavioral or biochemical endpoints Ecosystems Potentially at Risk For typical crop applications, the ecosystem at risk is the field itself, in terms of organisms that might be sprayed during application, organisms affected by accumulation of fomesafen in the soil; and the adjacent aquatic and terrestrial environments affected due to runoff, spray drift, or groundwater contamination. In water bodies receiving runoff from agricultural fields, pelagic and benthic elements are considered. Terrestrial organisms assessed include non-target plants, insects, amphibians, reptiles, birds, and mammals. Because fomesafen is an herbicide, potential affects on non-target plants have been addressed at length Ecological Effects Evaluation of ecological effects focuses initially on direct effects to the groups of organisms residing in the ecosystems at risk, based on ratios of the estimated environmental concentration (EEC) to a designated toxicity endpoint for a surrogate test organism. If pre-established levels of concern (LOCs) are exceeded for direct effects, indirect effects to endangered species (e.g. food chain, decrease in community diversity) are evaluated based on the group of organisms exceeding the LOC. Direct Direct effects evaluated are the survival, growth, and reproduction of various taxa of organisms potentially exposed to fomesafen. Taxonomic groups evaluated include aquatic plants (algae and vascular), aquatic invertebrates, aquatic vertebrates, terrestrial plants, terrestrial invertebrates, birds, and mammals. Both acute and chronic effects are considered. Indirect When herbicides are applied, indirect effects may include a decline in primary productivity, or change in composition of plant communities proximate to the treated area or systems (wetlands and water bodies) receiving runoff from the site. If LOCs are exceeded for any taxa, potential indirect effects to endangered species are assessed. 2.6 Conceptual Model The conceptual model for fomesafen (Figure 1) is located on the following page. 15

16 Stressor Fomesafen applied to crop Source Runoff Spray drift Direct application Soil accumulation, leaching and runoff Receptors Aquatic plants invertebrates vertebrates Terrestrial insects Birds Mammals Terrestrial plants Soil invertebrates Attribute Change Individual organisms Reduced survival Reduced growth Reduced reproduction Food chain Reduction in primary productivity Reduction in prey Shift in community composition Habitat integrity Reduced cover Community change Figure 1 - Conceptual Model for Fomesafen 16

17 2.7 Risk Hypotheses Fomesafen deposited on plant surfaces may affect growth, survival, or fecundity of birds and/or small mammals ingesting the affected vegetation. Fomesafen accumulating in soil may be toxic to non-target plants. Fomesafen in runoff from treated areas may kill aquatic plants, aquatic invertebrates, or fish. Fomesafen in runoff from treated areas may reduce populations of aquatic plants, aquatic invertebrates, or fish, causing changes in the community. Fomesafen in runoff from treated areas may accumulate in sediments, resulting in chronic impacts to the benthic community. Fomesafen is expected to move from the application site by leaching into groundwater and runoff into surface water. Use of water resources with fomesafen occurrence as an irrigation source water may adversely impact non-target plants. 2.8 Key Uncertainties and Information Gaps Bioavailability of fomesafen to non-target organisms is influenced by site characteristics. An effort has been made to examine a variety of sites typical of areas where it might be used. Scenarios and input parameters have been selected to represent conditions where fomesafen is most bioavailable. Overall, conditions have been selected to develop a conservative, screening level estimate. In some locations and under some conditions, risk may be overestimated. Fomesafen is sufficiently persistent in soil that it may cause prolonged phytotoxic effects. EFED has no specific data to evaluate how long measurable toxic effects on plants may occur after initial treatment, thus it is difficult to assess the potential effects of chronic exposure. Natural genetic variability in wild organisms may affect their response to fomesafen, and the distribution of species sensitivity is not well understood. Risk may be under- or over-estimated by available toxicity data and use of surrogate taxa. At the time of the risk assessment, terrestrial and aquatic plant data were available, but had not yet been completely through EFED s data evaluation process. Provisional classification of these data is supplemental. Minor changes in the reported toxicological endpoints (i.e., LC 50, EC 25, NOAEC) may occur following statistical analysis of the data according the EFED s protocols, but these changes are not expected to affect the risk conclusions. Available avian reproduction studies did not establish LOAECs, and the NOAEC is set at the highest dose (approximately 50 ppm). Exceedences of the chronic LOCs for birds may or may not imply risk. The locations of sensitive and/or valuable plant communities are not available, thus effects on these communities can only be assessed in a general, qualitative sense. 17

18 3. Exposure Assessment EFED conducted the exposure assessment using standardized exposure modeling. The exposure was modeled separately for each proposed application rate, and in most cases, for each proposed application method. Two proposed application rates (0.50 lb ai/a for ground application and lb ai/a for aerial application) were modeled for all routes of exposure. An alternative application rate (0.2 lb ai/a) was also modeled for the terrestrial exposures. 3.1 Label Application Rates and Intervals Fomesafen is proposed as a pre-plant, pre-emergence, and post-emergence herbicide for use on broadleaf weeds, grasses, and sedges. The registrant and IR-4 are petitioning to use the sodium salt of fomesafen, formulated as REFLEX 2.5 Gallon (EPA Reg. No ) and REFLEX 2LC (EPA Reg. No ) on cotton, snap beans, and dry beans. Regional label application rates and application intervals for dry beans and snap beans are shown in Table 2. Fomesafen can be applied through aerial and ground spray at a maximum application rate of lbs ai/a. Table 2 Label Application Rates and Intervals for Post-Emergent Fomesafen Use on Snap Beans and Dry Beans 1 Region States in Region App Rate (lbs/a) Application Interval 1 AL, AK, GA, MS, MO, NC, OK, SC, TN, TX year 2 DE, KY, MD, VA, WV, IL, IN, OH, PA year 3 CN, IA, ME, MA, MI, NH, NJ, NY, PA, RI, VT, WI year 4 KS, MI, MN, NE, WI, ND, SD, year 5 ND, SD, MN year 1- REFLEX 2.5 Gallon (EPA Reg. No ) and REFLEX 2LC (EPA Reg. No ) Label restrictions for cotton are shown in Table 3. As post-emergent herbicide, fomesafen can be applied using ground spray at a maximum application rate of 0.5 lbs ai/a. Table 3 Label Application Rates and Techniques for Fomesafen Use on Cotton 1 Application Timing Application Technique Max App Rate (lbs/a) Pre-emergent Aerial or Ground Spray Post-directed Ground Spray Post-emergent Ground Spray REFLEX 2.5 Gallon (EPA Reg. No ) 18

19 3.2 Aquatic Exposure Tier II EFED aquatic exposure models use the linked Pesticide Root Zone Model and Exposure Analysis Model System (PRZM-EXAMS). PRZM uses the chemical s physical and environmental fate properties and the site characteristics to predict the concentration of pesticide in runoff and entrained sediment from the field. EXAMS estimates the concentration of pesticide in an edge-of-field small water-body receiving runoff from the field. The water-body has no outflow with a constant volume (20 million liters), and is intended to represent an upper-end occurrence concentration PRZM-EXAMS Modeling Inputs and Scenario Selection The aquatic exposure assessment for fomesafen was conducted to assess use on soybeans and cotton. Soybeans were used a surrogate for dry beans and snap beans, as EFED currently has no standard scenarios for these crops. Standard scenarios were selected to assess runoff potential from vulnerable use sites in MS (soybean and cotton), NC (cotton), and TX (cotton). Input parameters for fomesafen were selected according to EFED Input Parameter Guidance for PRZM/EXAMS 1. Input parameters are shown in Table 4. Table 4 Input Parameters for PRZM-EXAMS Modeling of Fomesafen on Cotton and Soybeans Parameter Value Comments Source Application Rate (kg a.i./ha)- Cotton 0.42 Aerial Spray Label 1 Application Rate (kg a.i./ha)- Cotton 0.56 Ground Spray Label 1 Application Rate (kg a.i/ha)- Soybean 0.42 Aerial Spray Label 1 Molecular Weight (grams/mole) 420 EPA Solubility (mg/l) 7; 20 0 c MRID Vapor Pressure (torr) 50 o C HSDB Henry s Constant (atm m 3 /mol) 7.5 x10-13 Estimated HSDB Kd (L/kg) 0.68 Lowest nonsand Acc No K d Aerobic Soil Metabolism Half-life (days) Upper 90 th percentile of mean 2 Acc No Acc. No Aerobic Aquatic Metabolism Half-life (days) Upper 90 th Acc. No percentile of mean 3 Anaerobic Aquatic Metabolism Half-life Stable Conservative No Data Available (days) Assumption Photodegradation in Water (days) MRID Hydrolysis Half-life (days) Acc No Reflect application rates on the REFLEX 2LC, REFLEX 2.5 and REFLEX labels 2-Calculated from half-lives of 187.6, 630, 57, 693, 349.3, 527.1, 207 days using a mean of days and standard deviation of days. 3- Calculated from half-lives of 139.9, 60.9, 92.4, and days using a mean of 102 days and standard deviation of days. 1 Guidance for Selecting Input Parameters in Modeling the Environmental Fate and Transport of Pesticides. Version II, 2/28/02. 19

20 3.2.2 PRZM-EXAMS Modeling Output For aerial applications (Table 5), peak 1 in 10 year estimated environmental concentrations (EECs) ranged from 7.5 ppb (soybeans, MS) to 12.2 ppb (cotton, TX). Chronic 1-in-10 year (21-day average and 60-day average) EECs ranged from 6.4 ppb (soybean, MS, 60-day average) to 11.4 ppb (cotton, MS &TX, 21-day average). Table 5 PRZM-EXAMS EECs for Fomesafen at lb a.i/a 1 Region Crop State Peak 4 days 21 days 60 days μg/l (ppb) 1 Soybean MS Cotton MS Cotton NC Cotton TX Concentrations were derived for lb ai/a using aerial applications Peak 1-in-10 year EECs for ground spray applications (Table 6) ranged from 10.6 ppb (cotton, NC) to 15.1 ppb (cotton, MS). Chronic 1 in 10 year (21-day average and 60-day average) concentrations ranged from 8.6 ppb (cotton, MS, 60-day average) to 14.2 ppb (cotton, MS, 21-day average). Table 6 PRZM-EXAMS EECs for Fomesafen at 0.50 lb ai/a Region Crop State Peak 4 days 21 days 60 days μg/l 1 Cotton MS Cotton NC Cotton TX Concentrations were derived for 0.50 lb ai/a using ground spray 20

21 3.2.3 Registrant-submitted Aquatic Exposure Modeling The registrant submitted Tier II PRZM/EXAMS modeling assessments for fomesafen use on cotton and soybeans (MRID , , ). The modeling scenarios used were developed to assess fomesafen runoff from vulnerable use sites (Hydrologic D soils) with high rainfall. Selected scenarios represent sites in Tensas County, LA (MRLA 131) for cotton; Leflore, MS (MRLA 131) for soybeans; and Dodge, NE (MRLA 102B) for soybeans, and Sheboygan, WI (MRLA 95B). Maximum application rates were 0.31 lbs for snap beans, 0.5 lbs a.i./a for cotton, and lbs a.i./a for soybeans. Only ground spray was considered, although the labels allow aerial spray. The sources (study identification numbers) of environmental fate data are not reported. Additionally, there are no explanations for selection of the environmental fate data in the exposure assessment. Registrant estimated environmental concentrations of fomesafen in the small water bodies are shown in Tables 7, 8, and 9. Table 7 1-in-10 year Estimated Environmental Concentrations of Fomesafen from Cotton in LA. 90- Annual Scenari Peak 4-day 21-day 60-day Application Rate day Average o ID (ppb) (ppb) (ppb) (ppb) (ppb) (ppb) Cotton lbs ai/a post emerg 0.25 lbs ai/a post emerg Cotton lbs ai/a pre-emerg 0.25 lbs ai/a post-emerg Cotton lbs ai/a pre-emerg 0.25 lbs ai/a pre-emerg Table 8 1-in-10 year Estimated Environmental Concentrations of Fomesafen from Soybeans in NE and MS Annual Peak 21-day 60-day Application Rate MLRA hour day Average (ppb) (ppb) (ppb) (ppb) (ppb) (ppb) 0.25 lbs ai/a in NE alternate years lbs ai/a every year (102B) MS (131) Table 9 1 in 10 year Estimated Environmental Concentrations for Fomesafen from Snapbeans in WI Application MLRA Peak 4 days 21 days 60 days Rate μg/l 0.31 lbs a.i./a alternate years 95B

22 3.2.4 SCIGROW Modeling for Ground Water Because fomesafen is mobile and persistent in soil, a screening level groundwater assessment using SCIGROW (ver. 2.3) was conducted to estimate the concentration of fomesafen in shallow groundwater, which could potentially be used for crop irrigation. Input parameters for SCIGROW are listed in Table 10. A groundwater monitoring study was submitted (MRID ), but the shallow groundwater wells were dry during the study. Fomesafen was detected in soil porewater at concentrations of 1 μg/l (at 4 months), up to 17 μg/l (at 1 month). It was detected at a concentration of 1 μg/l in the medium- to deep-depth wells. Table 10 Input Parameters for SCIGROW Modeling for Fomesafen Parameter Value Comments Source Application Rate (kg a.i./ha)- Cotton 0.56 Label 1 K oc (L/kg) 68 Estimated 2 Acc No Aerobic Soil Metabolism Half-life (days) Mean 3 Acc No Acc. No Reflect maximum application rates on the REFLEX 2LC, REFLEX 2.5 and REFLEX labels 2-Koc estimated using Kd/SOC=Koc; where Kd=0.68 and SOC=1% SOC percentage 3-Calculated from half-lives of 187.6, 630, 57, 693, 349.3, 527.1, 207 days using a mean of days and standard deviation of days. Based on the SCIGROW estimate, the concentration of fomesafen in shallow ground water in sand soils is not expected to exceed 6.68 μg/l. A groundwater monitoring study was submitted (MRID ), but the shallow groundwater wells were dry during the study. Fomesafen was detected in soil porewater at concentrations of 1 μg/l (at 4 months), up to 17 μg/l (at 1 month). It was detected at a concentration of 1 μg/l in the medium- to deep-depth wells. Because fomesafen is expected to leach to groundwater, EFED has calculated the maximum application rate of fomesafen from two inches of irrigation water, using the following equations. This calculation assumes that two inches (0.167 ft) of irrigation water is required for optimum plant growth. The calculations are as follows: 43,560 ft 2 /A* ft irrigation water= 7,274 ft 3 for 2 inches of irrigation water/a 7,274 ft 3 irrigation water/a* liter/ft 3 =205, liters of irrigation water/a 205, liter of irrigation water/a *EEC :g/l = fomesafen :g/a (fomesafen :g/a)/ (10 6 ) = fomesafen grams/a*1lb/454 grams=fomesafen lbs ai/a. 22

23 Based on two inches of irrigation and the SCIGROW estimate, the application rate of fomesafen is estimated at lbs ai/a. Using the concentrations of 1 mg/l and 17mg/L (from the groundwater study) as outer bounds, concentrations of fomesafen in irrigation water could range from lbs ai/a Soil Accumulation Because of the persistence of fomesafen in soil, a screening level assessment was conducted to quantify the accumulation of fomesafen residues in soil. A first-order decay model (A=A o e -kt ) was used to estimate fomesafen soil concentrations. The time period in the model (t) was set to 730 days to represent alternate years applications. The upper 90 th percentile of the mean half-life (t 1/2 =428 days; k= days -1 ) was used to represent the microbial mediated decay rate of fomesafen in soil. The starting concentration (A 0 ) was set at the label recommended application rate of lbs ai/a for aerial applications and 0.5 lbs ai/a for ground applications. The modeling scenario assumes that 100% of fomesafen residue is applied to the soil as recommended for a preemergent application. The model scenario also assumes that microbial degradation is the only route of dissipation from the application site. These assumptions are expected to exaggerate predicted formesafen soil concentrations. Figure 2 illustrates the fomesafen concentrations in soil reach a plateau after approximately 10 years regardless of the application rate. Application rates of lbs/a can theoretically result in a maximum fomesafen concentration of 0.14 mg/kg. Higher application rates of 0.5 lbs ai/a can theoretically result in a maximum fomesafen concentration of 0.19 mg/kg Fomesafen Soil Conc. (mg/kg) Year Figure 2 - Estimate of Fomesafen Loading in the Surface Soil (0-15 cm depth) from alternate year applications of lbs/a (solid line) and 0.5 lbs/a (dotted line) 23

24 3.3 Bird and Mammal Exposure (TREX) EFED estimates exposure of birds and mammals using the Terrestrial Exposure Model (TREX). TREX uses the Kenaga nomagram to determine pesticide residues on several categories of food items, then calculates the potential dose an organism might receive from ingesting contaminated items using allometric equations. Dose estimates in the tables 8 and 9 are based on the Kenaga upper bound dose and the assumptions that the organism exclusively eats one type of food item and forages only in the treated and/or overspray areas. For birds, dose estimates for the 0.2 lb ai/a application rate range from 0.87 mg/kg bwt (1000g frugivores, granivores, and insectivores) to 54.7 mg/kg bwt (20 g herbivores). At the 0.37 lb ai/a application rate, estimated doses range from 1.64 (1000g frugivores, granivores, and insectivores) to 102 (1000g fruit and pods). Dose estimates for the 0.49 lb ai/a application rate range from 2.14 mg/kg bwt (1000g frugivores, granivores, and insectivores) to 134 mg/kg bwt (20 g herbivores). Table 11 Bird Dose Estimates Kenaga Upper Bound Dose (mg/kg bwt) Feeding Categories Small (20 g) Medium (100 g) Large (1000 g) 0.2 lb ai/a Appllication Rate (Alternative) Short grass Tall grass Broadleaf plants/small insects Fruits/pods/seeds/large insects lb ai/a Application Rate Short grass Tall grass Broadleaf plants/small insects Fruits/pods/seeds/large insects lb ai/a Application Rate Short grass Tall grass Broadleaf plants/small insects Fruits/pods/seeds/large insects

25 For mammals dose estimates for the 0.2 lb ai/a application rate range from 0.10 mg/kg bwt (1000g granivore) to 45.8 mg/kg bwt (20 g short grass). At the 0.37 lb ai/a application rate, estimated doses range from 0.19 (1000g granivore) to 85.8 (20 g short grass). Dose estimates for the 0.49 lb ai/a application rate range from 0.25 mg/kg bwt (1000g granivore) to 112 mg/kg bwt (20 g short grass). Table 12 Mammal Dose Estimates Kenaga Upper Bound Dose (mg/kg bwt) Feeding Categories Small (15 g) Medium (35 g) Large (1000 g) 0.2 lb ai/a Appllication Rate (Alternative) Herbivores/Insectivores Short grass Tall grass Broadleaf plants/small insects Fruits/pods/seeds/large insects Granivores Fruits/pods/seeds/large insects lb ai/a Application Rate Herbivores/Insectivores Short grass Tall grass Broadleaf plants/small insects Fruits/pods/seeds/large insects Granivores Fruits/pods/seeds/large insects lb ai/a Application Rate Herbivores/Insectivores Short grass Tall grass Broadleaf plants/small insects Fruits/pods/seeds/large insects Granivores Fruits/pods/seeds/large insects

26 3.3 Terrestrial Plant Exposure (TerrPlant & AgDrift) Currently, EFED uses the TerrPlant Model (Version 1.2.1) to evaluate exposure of terrestrial plants to pesticides applied on agricultural fields. In cases where spray drift may be of concern in the risk assessment EFED also uses the AgDrift model TerrPlant TerrPlant has two basic exposure scenarios. The first is an adjacent upland area, which is exposed to the pesticide via drift and dissolved concentrations in sheet runoff. The second is an adjacent semi-aquatic (wetland) area, which is exposed to the pesticide via drift and to dissolved concentrations in channelized runoff. Drift is calculated as a percentage of the application rate (1% for ground, and 5% for aerial, airblast, or spray chemigation) and is not adjusted for distance from the application site. The amount of dissolved pesticide in the runoff component is estimated based on solubility of the active ingredient. TerrPlant estimates are shown in Table 13. Table 13 Terrestrial Plant Exposure Application Method Total Loading (Runoff +Drift) (lb ai/a) Drift EEC (lb ai/a) Upland areas Wetland areas All areas Use at lb ai/a Aerial Ground Use at 0.50 lb ai/a Aerial Ground AgDrift Because of concerns about effects on non-target plants located in the overspray or spray drift areas, EFED elected to perform an analysis of potential deposition of fomesafen using AgDrift modeling software. AgDrift was developed using extensive fieldmeasured data sets, and provides a method of estimating deposition of the compound of concern at a specified distance away from the application source. Deposition is heavily dependent on the method of application and droplet size. A Tier I analysis, which is driven primarily by these two variables, was conducted for both ground and aerial applications. More sophisticated (Tier II) analyses also incorporate atmospheric parameters, but were not deemed necessary for this screening level risk assessment. Because terrestrial plants, especially dicotyledons (dicots) are extremely sensitive to fomesafen, deposition associated with both proposed methods of application was evaluated. For ground applications, two alternative lower rates were also assessed. The lower alternative rates (0.375 lb ai/a and 0.2 lb ai/a) were selected based on 1) the lower proposed rate for aerial application, and 2) the AgDrift predicted deposition of fomesafen on foliage directly below the application source for the aerial application rate. Because the registrant proposed the aerial rate, EFED presumes it will be efficacious for the target pests. Detailed information on input parameters is included in Appendix E. 26

27 Point deposition was estimated for both ground (Table 14) and aerial (Table 15) applications. Table 14 Estimated Point Deposition of Fomesafen from AgDrift (Tier I, Ground Application, Low Boom, 90% Percentile Estimate) Distance from Fraction Point Deposition (lb ai/a) Application Source (ft) of Applied Proposed Rate (0.50 lb ai/a) Alternative Rate (0.375 lb ai/a) Alternative Rate (0.20 lb ai/a) Table 15 Estimated Point Deposition of Fomesafen from AgDrift (Tier I, Aerial Application, 90% Percentile Estimate) Point Deposition Distance from Application Source (ft) Fraction of Applied (lb ai/a) Proposed Rate (0.375 lb ai/a) 27

28 4. Effects Assessment Toxicity endpoints are established based on data generated from guideline studies submitted by the registrant, and from open literature studies that meet the criteria for inclusion into the ECOTOX database maintained by EPA/ORD. EFED policy is to use the most sensitive endpoint for each taxa evaluated. In aquatic systems, taxa evaluated include aquatic plants, invertebrates, and fish. Fish serve as a surrogate for aquatic-phase amphibians. Where data are available, separate endpoints are used for freshwater and estuarine/marine organisms. In terrestrial systems, taxa evaluated include birds and mammals. Bird endpoints are generally derived from guideline studies on bobwhite quail and/or mallard duck. Bird data is used as a surrogate for reptiles and terrestrial-phase amphibians. Mammal data is derived from guideline studies conducted on laboratory rats, mice, or rabbits. 4.1 Aquatic Guideline Data Fomesafen was originally registered for use in the 1980s. Guideline studies from that time were available for aquatic invertebrates and fish, both freshwater and marine/estuarine. Although some of the studies were conducted on formulated product, and would not be acceptable under current standards, they were classified as core or supplemental under the guidelines at the time they were submitted. When necessary, endpoints were re-calculated and/or data were converted to express toxicity on the basis of active ingredient. Details of conversion are included in Appendix E. Aquatic plant data were submitted by the registrant (upon request by EFED), during the development of this risk assessment. Although the Data Evaluation Review (DER) process has not yet been completed for these studies, they have been provisionally classifed as Supplemental, and the toxicity data has been incorporated into the assessment. Overall, fomesafen is slightly toxic to practically nontoxic to invertebrates and practically non-toxic to fish on an acute basis (Table 16). Chronic data were also available, and are presented in Table

29 Table 16 Acute Aquatic Data from Registrant-submitted Studies Species LC 50 (ppm) 95% C.I. (ppm) Freshwater Organisms Green alga 1 (Selenastrum capricornutum) Water flea (Daphnia magna) Rainbow Trout (Onchorynchus mykiss) Estuarine/ Marine organisms Marine diatom 1 (Skeletonema costatum) Mysid shrimp (Mysidopsis bahia) Sheepshead minnow (Cyprinodon varigetus) 0.12 (biomass) 376 (practically nontoxic) 126 (practically nontoxic) 1.51 (biomass) 25 (slightly toxic) >163 (practically nontoxic) NOAEC (ppm) ND ND ND >163 Classification (MRID) Supplemental ( ) Technical Core 2, 3 (163169) Formulation Core 2, 3 (103023) Formulation Supplemental ( ) Technical Core 2 (135647) Technical Core 2, 3 (135651) Formulation 1 Provisional data and classification, pending final review. 2 Data are from studies originally reviewed and classified in 1984, some of which used formulated product. 3 For purposes of this risk assessment, test concentrations were adjusted for percent a.i. if necessary, and endpoints were re-calculated using TOXANAL software. ND-not determined. Table 17 Chronic Aquatic Data from Registrant-submitted Studies Species NOAEC LOAEC (ppm) (ppm) Endpoints Affected Freshwater Organisms Water flea (Daphnia magna) Estuarine/ Marine organisms Mysid shrimp (Mysidopsis bahia) Sheepshead minnow 2 (Cyprinodon varigetus) Reduced growth, Total # of offspring Parental mortality Reduced larval survival Classification 1 (MRID) Core (135642) Formulation Core (135648) Formulation Core (135644) Formulation 1 Data are from studies originally reviewed and classified in 1984, some of which used formulated product. 2 For purposes of this risk assessment, test concentrations were adjusted for percent a.i. 4.2 Aquatic Data from ECOTOX The ECOTOX database was accessed, and no toxicity data for fomesafen were located. 29

30 4.3 Terrestrial Plant Guideline Data Terrestrial plant guideline studies were submitted during the development of this risk assessment. Data are shown below (Table 18), but are considered provisional pending final data evaluation review. Fomesafen is effective, both pre- and post-emergent, against a variety of plants, although dicots appear to be more sensitive than monocots for both endpoints. The product is marketed as a control for broad-leafed weeds. In some cases, calculated EC 25 s were below the concentrations tested, so a NOAEC was not determined. The most sensitive endpoint, used in the risk assessment, is the vegetative vigor EC 25 for radish ( lb ai/a). Table 18 Terrestrial Plant Guideline Data Common Class EC Species 25 name (lb ai/a) NOAEC (lb ai/a)) Vegetative Vigor Raphanus sativus Radish D Echinochloa crus-galli Barnyard grass M Seedling emergence Lycopersicon Tomato esculentum D ND Allium cepa Onion M ND 1 Provisional classification, pending final data evaluation review. Classification 1 (MRID) Supplementary ( ) Supplementary ( ) Efficacy data (MRID , Appendix E) were part of the data package submitted. The efficacy data included pre-emergence and post-emergence treatment of 24 plant species, at two concentrations (0.25 and 1.0 kg ai/ha). The two concentrations bracket the currently proposed rates (0.42 and 0.54 kg ai/ha). The plant species tested included both monocots (11 species) and dicots (13 species). Both crop (7 species) and non-crop (17 species) plants were evaluated. With the exception of soybeans, all plants tested experienced >20% damage when treated pre-emergence, with a significant number (65%) experiencing >80% damage when treated with the lower concentration (0.25 kg ai/ha). Applied post-emergence, fomesafen is slightly less effective, with damage typically in the 0-40% range for monocots and 40-80% range for dicots. The report did not specify how damage was quantified. 30