Chlopyrifos_WSM_Documentation_Final_Report.pdf

SERA TR-056-11-02-02b

Chlorpyrifos:

WorksheetMaker Workbook Documentation

Final Report

Submitted to:

Dr. Harold Thistle

USDA Forest Service Forest Health Technology Enterprise Team

180 Canfield St.

Morgantown, WV 26505 Email: hthistle@fs.fed.us

USDA Forest Service Contract: AG-3187-C-12-0009 USDA Forest Order Number: AG-3187-D-14-0146

SERA Internal Task No. 56-11

Submitted by:

Patrick R. Durkin

Syracuse Environmental Research Associates, Inc.

8125 Solomon Seal

Manlius, New York 13104

E-Mail: SERA_INC@msn.com Home Page: www.sera-inc.com

July 29, 2015 mailto:hthistle@fs.fed.us http://www.sera-inc.com/ ii

Table of Contents

LIST OF TABLES ......................................................................................................................... iii LIST OF APPENDICES ................................................................................................................ iii REVISION NOTE ......................................................................................................................... iii ACRONYMS, ABBREVIATIONS, AND SYMBOLS ................................................................ iv

1. INTRODUCTION

1.1. General Information

1.2. Chemical Specific Information

1.2.1. Information Sources

1.2.2. Forest Service Use at JH Stone Nursery

2. CHEMICAL/PHYSICAL PROPERTIES

3. HUMAN HEALTH

3.1. Hazard Identification

3.2. Exposure Assessment

3.2.1. Workers

3.2.2. General Public

3.3. Dose-Response Assessment

3.4. Risk Characterization

3.4.1. Workers

3.4.2. General Public

4. ECOLOGICAL EFFECTS

4.1. Hazard Identification

4.2. Exposure Assessment

4.3. Dose-Response Assessment

4.3.1. Terrestrial Organisms

4.3.1.1. Mammals

4.3.1.2. Birds

4.3.1.3. Other Terrestrial Organisms

4.3.2. Aquatic Organisms

4.3.2.1. Fish

4.3.2.2. Amphibians

4.3.2.3. Aquatic Invertebrates

4.3.2.4. Aquatic Plants

4.4. Risk Characterization

4.4.1. Terrestrial Organisms

4.4.1.1. Mammals

4.4.1.2. Birds

4.4.1.4. Other Organisms and Indirect Effects................... Error! Bookmark not defined.

iii

4.4.2. Aquatic Organisms

5. REFERENCES

5.1. Chemical Specific References

5.2. Standard Reference

LIST OF TABLES

Table 1: Chemical and Physical Properties Table 2: Derivation of Worker Exposure Rates Table 3: Inputs and Outputs for FIRST Simulations Table 4: Summary of toxicity values used in human health risk assessment Table 5: Summary of toxicity values used in ERA

LIST OF APPENDICES

Appendix 1: FIRST Runs

REVISION NOTE

This report is a modification to SERA TR-056-1-02-01a, Chlorpyrifos:

WorksheetMaker Workbook Documentation: PARTIAL (pending Forest Service input) which was sent to the Forest Service with queries on November 26, 2014.

Changes to this report are based on the following comments from the Forest Service:

Notes from John Justin and Shawna Bautista in an EXCEL workbook (Mancozeb SAMPLE Workbook-SB-JJ.xlsm) via an email from Shawna Bautista dated 6/2/2015.

Comments from John Justin in a Microsoft Word file (Query Template for WSM Input Effort.docx) via an email from Shawna Bautista dated 12/30/2014.

The most recent set of comments (i.e., the EXCEL workbook from 6/2/2015) is given precedence when the feedback is inconsistent.

iv

ACRONYMS, ABBREVIATIONS, AND SYMBOLS

ACGIH American Conference of Governmental Industrial Hygienists AChE acetylcholine AEL adverse-effect level a.i. active ingredient a.k.a. also known as a.s. active substance AOEL acceptable occupational exposure limit ATSDR Agency for Toxic Substances and Disease Registry BCF bioconcentration factor bw body weight calc calculated value CBI confidential business information ChE cholinesterase CI confidence interval cm centimeter CNS central nervous system COC crop oil concentrates DAA days after application DAT days after treatment DER data evaluation record d.f. degrees of freedom EC emulsifiable concentrate ECx concentration causing X% inhibition of a process EC25 concentration causing 25% inhibition of a process EC50 concentration causing 50% inhibition of a process ECOTOX ECOTOXicology (database used by U.S. EPA/OPP) EFED Environmental Fate and Effects Division (U.S. EPA/OPP) ERA ecological risk assessment ExToxNet Extension Toxicology Network F female FH Forest Health FIFRA Federal Insecticide, Fungicide and Rodenticide Act FIRST FQPA Index Reservoir Screening Tool FQPA Food Quality Protection Act g gram GLP Good Laboratory Practices ha hectare HED Health Effects Division (U.S. EPA/OPP) HHRA human health risk assessment HIARC Hazard Identification and Assessment Review Committee (part of U.S. EPA/OPP/HED) HQ hazard quotient HRAC Herbicide Resistance Action Committee IARC International Agency for Research on Cancer IRED Interim Reregistration Eligibility Decision IRIS Integrated Risk Information System v ka absorption coefficient ke elimination coefficient kg kilogram Ko/c organic carbon partition coefficient Ko/w octanol-water partition coefficient Kp skin permeability coefficient L liter lb pound LC50 lethal concentration, 50% kill LD50 lethal dose, 50% kill LOAEL lowest-observed-adverse-effect level LOC level of concern LR50 50% lethal response [EFSA/European term] m meter M male mg milligram mg/kg/day milligrams of agent per kilogram of body weight per day mL milliliter mM millimole mPa millipascal, (0.001 Pa) MITC methyl isothiocyanate MOS margin of safety MRID Master Record Identification Number MSDS material safety data sheet MSO methylated seed oil MW molecular weight NOAEL no-observed-adverse-effect level NOEC no-observed-effect concentration NOEL no-observed-effect level NOS not otherwise specified N.R. not reported OPP Office of Pesticide Programs ppm parts per million RBC red blood cell RED re-registration eligibility decision RfD reference dose SERA Syracuse Environmental Research Associates TCP 3,5,6-trichloro-2-pyridinol T.G.I.A. Technical grade active ingredient UF uncertainty factor U.S. United States USDA U.S. Department of Agriculture U.S. EPA U.S. Environmental Protection Agency USGS U.S. Geological Survey WHO World Health Organization

1. INTRODUCTION 1

1.1. General Information 2

This document supports the development of a WorksheetMaker EXCEL workbook for the 3 subject pesticides. As detailed in SERA (2011a), WorksheetMaker is a utility that automates the 4 generation of EXCEL workbooks that accompany Forest Service risk assessments, and these 5 EXCEL workbooks are typically generated in the development of Forest Service risk 6 assessments (SERA 2014a). 7 The development of full Forest Service risk assessments, however, is resource intensive. For 9 some pesticides used in only relatively small amounts and/or only in few locations, the 10 development of full Forest Service risk assessments is not feasible. Nonetheless, the Forest 11 Service may be required to develop risk analyses supported by WorksheetMaker EXCEL 12 workbooks. To meet this need, an MS Word utility was developed to facilitate the addition of 13 pesticides and pesticide formulations into the Microsoft Access database used by 14 WorksheetMaker (SERA 2011b). With this addition, WorksheetMaker can be used to generate 15 EXCEL workbooks typical of those that accompany Forest Service risk assessments. 16 The current document is designed to serve as documentation for the application of this general 18 method for the pesticide discussed in Section 1.2. The major difference between this approach to 19 using WorksheetMaker and the typical use of WorksheetMaker in the development of Forest 20 Service risk assessments involves the level of documentation and the sources used in developing 21 the documentation. While standard Forest Service risk assessments involve a relatively detailed 22 review and evaluation of the open literature and publically available documents from the U.S. 23 EPA, as discussed further in Section 1.2, the current assessment relies primarily on secondary 24 sources with minimal independent evaluation of the data. 25

1.2. Chemical Specific Information 27

1.2.1. Information Sources 28

The primary motivation for the development of this report involves the assessment of pesticide 29 use at the J. Herbert Stone Nursery in Central Point, Oregon. While reference to this Forest 30 Service facility is made in the current report as needed, the design of this report is intended to 31 support other activities within the Forest Service with only minor modification—i.e., changes in 32 input parameters related to pesticide use in WorksheetMaker. 33 The current document concerns chlorpyrifos. Most of the information on chlorpyrifos was 35 identified at the U.S. EPA’s Pesticide Chemical Search website 36 (http://iaspub.epa.gov/apex/pesticides/f?p=CHEMICALSEARCH:1:5025550978400::NO:1::) 37 using the search term “Chlorpyrifos”. Chlorpyrifos is a well-studied insecticide. U.S. EPA’s 38 Pesticide Chemical Search website lists 22 regulatory action documents on chlorpyrifos, 453 39 cleared science reviews, and five E-Dockets for the period from 2007 to 2012. TOXLINE 40 (http://toxnet.nlm.nih.gov) contains 8401 open literature citations for chlorpyrifos. 41 For the current documentation, information on human health effects is taken primarily from the 43 preliminary human health risk assessment for the registration review of chlorpyrifos (U.S. 44 http://iaspub.epa.gov/apex/pesticides/f?p=CHEMICALSEARCH:1:5025550978400::NO:1 http://toxnet.nlm.nih.gov/

EPA/OPP/HED 2011a). The Interim Reregistration Eligibility Decision for Chlorpyrifos (U.S. 1 EPA/OPP 2002), the EPA’s human health risk assessment for the RED (U.S. EPA/OPP/HED 2 2000), and the EPA’s assessment of human health effects related to volatilization (U.S. 3 EPA/OPP/HED 2013) were also consulted. Information on ecological effects is taken from the 4 EPA’s assessment of the potential effects of chlorpyrifos on threatened and endangered species 5 (U.S. EPA/OPP/EFED 2009). Information on chemical and physical properties and the 6 environmental fate of chlorpyrifos is taken from the EPA documents, as discussed further in 7 Section 2. 8

1.2.2. Forest Service Use at JH Stone Nursery 9

The JH Stone Nursery specified two formulations of chlorpyrifos—i.e., DuraGuard ME (20.0% 10 chlorpyrifos) for greenhouse applications or Dursban 50W (50.0% chlorpyrifos) for field grown 11 crops. The application rate for field crops is specified as 2 lbs./acre with an application volume 12 of 30 gallons per acre. A total of 5 acres will be treated each year with single applications 13 involving between 0.5 and 2.5 acres. The Forest Service also indicated that up to 5 applications 14 may be made at an application interval as short as 14 days. 15 Dursban 50W may be applied by backpack (directed foliar) or tractor mounted spray boom. The 17 application rate of 2 lbs/acre appears to pertain to the Dursban 50W formulation. Thus, the 18 application rate in terms of a.i. is taken as 1 lb a.i./acre. 19 The JH Stone Nursery specified that personal protective equipment will include coveralls over 21 long sleeve shirt and pants, chemical resistant gloves, chemical resistant apron when mixing or 22 loading, chemical resistant footwear plus socks, and a NIOSH approved respirator with any 23 N,R,P, or HE filter. This array of personal protective equipment is required on the product labels 24 for Dursban 50W and DuraGuard ME. 25 As noted in Section 3.2.1, a key input for worker exposure assessments is the amount of the 27 pesticide that a worker would handle in 1 day. The Forest Service indicated (in the annotated 28 workbook from 6/2/2015) that the maximum exposure would involve treating 2.5 acres per hour 29 for 2 hours of application. Thus, the maximum amount handled by a worker would be 5 lbs 30 a.i./day [1 lb a.i./acre x 2.5 acres/hour x 2 hours/day]. Applications could be made with either 31 tractor mounted spray booms or directed foliar backpack spray. Modifications to the workbook 32 necessary to assess backpack applications are given in Section 3.2.1. 33 The WorksheetMaker workbook that accompanies this risk assessment is based on an October 7, 35 2008 label for Dursban 50W obtained from Syngenta web site 36 (http://www.syngentacropprotection.com/pdf/labels/daconilweastik0100.pdf). Based on 37 information from the Forest Service (the illustration in JH Stone Nursery 38 Information-v3 Shawna Aug 21.docx), the JH Stone nursery is about 220 acres in 39 size. Using a 5-acre treated area, a proportion of about 0.023 [5 acres ÷ 220 acres ≈ 0.0227] of 40 the nursery area would be treated. As discussed further in Section 3.2.2, this proportion is used 41 to modify the water contamination rates. 42 http://www.syngentacropprotection.com/pdf/labels/daconilweastik0100.pdf

2. CHEMICAL/PHYSICAL PROPERTIES 2

Chlorpyrifos has been in use in the United States since 1965 (U.S. EPA/OPP 2006). This 3 insecticide was developed by Dow Chemical Company, and current formulations of Dursban and 4 Lorsban are currently registered to Dow AgroSciences. Chlorpyrifos is off-patent, and 5 formulations such as DuraGuard are supplied by other manufacturers. From 1987 to 1998, the 6 annual use of chlorpyrifos in the United States ranged from 21 to 24 million pounds with about 7 11 million pounds used in non-agricultural sites (U.S. EPA/OPP 2006). Chlorpyrifos has been 8 available in over 400 formulations (U.S. EPA/OPP 2006); however, only 137 formulations are 9 currently active in the United States (Kegley et al. 2014). The U.S. EPA registration review 10 program operates on a 15-year cycle. Chlorpyrifos has been under registration review since 11 2009 and is not scheduled for completion until 2015. 12 Table 1 summarizes the chemical and physical properties of chlorpyrifos. As discussed further 14 in Section 3 (Human Health) and Section 4 (Ecological Effects), chlorpyrifos is an inhibitor of 15 acetylcholinesterase (e.g., U.S. EPA/OPP 2009). As reviewed in U.S. EPA/OPP/EFED (2009), 16 aerobic and anaerobic metabolism is the major route of degradation for chlorpyrifos in the 17 environment. As summarized in Table 1, the aerobic soil half-lives for chlorpyrifos range from 18 11 to 180 days and the anaerobic soil half-lives range from 39 to 51 days. While the ranges of 19 half-lives, particularly those for aerobic soil metabolism may be viewed as somewhat wide, high 20 variability in rates of soil metabolism are common and often reflect differences in the nature and 21 density of microbial populations in the soil. As also summarized in Table 1, the data on the 22 binding of chlorpyrifos to soils also display substantial variability with Koc values ranging from 23 360 to 31,000 mL/g. As discussed further in Section 3.2.2, the variability in soil binding and soil 24 metabolism are the primary factors in the estimates of surface water concentrations of 25 chlorpyrifos. No data are available on the aquatic metabolism of chlorpyrifos. Following the 26 standard approach used by U.S. EPA/OPP, the rates of aerobic and anaerobic aquatic metabolism 27 are estimated as twice the half-lives for the aerobic and anaerobic soil half-lives. Chlorpyrifos is 28 relatively lipophilic with reported Kow values of about 50,000 to 90,000. As with most lipophilic 29 compounds, chlorpyrifos will bioconcentrate in fish and has a reported BCF of about 2700 in 30 rainbow trout. 31 As also summarized in Table 1, the major environmental metabolite of chlorpyrifos is 3,5,6-33 trichloro-2-pyridinol (TCP). TCP is also a major environmental metabolite of triclopyr that is 34 discussed extensively in the Forest Service risk assessment on triclopyr (SERA 2011). If 35 chlorpyrifos and triclopyr are applied in the same location, the impact of both pesticides on the 36 formation of TCP may require additional consideration. Unlike the case with triclopyr, 37 chlorpyrifos is much more toxic than TCP and is not a metabolite of major concern in 38 applications of chlorpyrifos in terms of either human health (U.S. EPA/OPP/HED 2000, p. 2) or 39 ecological effects (U.S. EPA/OPP/EFED 2009, p. 10). Consequently, TCP is not addressed 40 further in the current document. 41

3. HUMAN HEALTH 3

3.1. Hazard Identification 4

While full Forest Service risk assessments provide a detailed discussion of the available toxicity 5 data on the pesticide under consideration, this approach is not taken in the current document, in 6 the interest of economy. 7 Chlorpyrifos acts by reversibly inhibiting acetylcholinesterase (AChE). Malathion is another 9 insecticide that inhibits AChE, and the Forest Service risk assessment on malathion (SERA 10 2008) provides a relatively detailed discussion of biological mechanisms and toxicity of AChE 11 inhibition. In short, acetylcholine is a neurotransmitter, a compound that facilitates transmission 12 of neural impulses between nerve cells as well as the activation of muscle and other effector cells 13 by nerve cells. Normally, the acetylcholine is rapidly degraded to inactive agents (acetate ion 14 and choline) by AChE. Inhibitors of AChE lead to the accumulation of acetylcholine in synapses 15 (i.e., spaces between cells) which causes a continuous stimulation of the cholinergic system and 16 may lead to paralysis because of nerve cell fatigue. Depending on the degree of AChE 17 inhibition, a broad spectrum of clinical effects may be induced ranging from mild signs of 18 toxicity (e.g., salivation or lacrimation) to convulsions and death (SERA 2008). 19 As discussed in U.S. EPA/OPP/HED (2000), the inhibition of acetylcholinesterase is a 21 mechanism of action common to numerous insecticides (e.g., azinphos methyl, chlorpyrifos-22 methyl, diazinon, dichlorvos, dicrotophos, dimethoate, disulfoton, methamidophos, 23 methidathion, monocrotophos, oxydemeton methyl, phorate, phosmet, and pirimiphos-methyl). 24 As also detailed in U.S. EPA/OPP/HED (2000), the EPA assesses cumulative effects of 25 inhibitors of AChE in situations where exposures to multiple AChE inhibitors may occur. Based 26 on a list of pesticides used at the J. Herbert Stone Nursery (i.e., JHSN_Pesticide_List_ 27 REVISED-08_07_14 (1).xlsx), chlorpyrifos is the only AChE inhibitor used at this 28 nursery. Consequently, there is no apparent need to assess cumulative effects involving the 29 inhibition of AChE. 30

3.2. Exposure Assessment 31

3.2.1. Workers 32

As discussed in SERA (2014b), the Forest Service risk assessments use a standard set of worker 33 exposure rates (Table 14 in SERA 2014b). As discussed in SERA (2014b), worker exposure 34 rates for directed foliar applications are also used for greenhouse applications, and the rates are 35 adjusted as necessary for dermal absorption. 36 U.S. EPA/OPP/HED (2000, p. 3) uses a dermal absorption rate of 3% for chlorpyrifos. This rate 38 is intended to reflect the proportion of chlorpyrifos absorbed over an 8-hour work day and, 39 therefore, corresponds to a first-order dermal absorption rate coefficient (ka) of 0.00375 hour-1 40 [0.03 ÷ 8 hours]. As detailed in Worksheet B03b of the WorksheetMaker workbook that 41 accompanies this report, the central estimate of the first-order dermal absorption rate based on 42 algorithms commonly used in Forest Service risk assessments (SERA 2014a, Section 3.1.3.2.2) 43 is about 0.0041 hour-1 with a 95% confidence interval of about 0.0015 to 0.012 hour-1. Rounding 44 to one significant digit, the rate of 0.00375 hour-1 and 0.0041 hour-1 are identical—i.e., 0.004 1 hour-1. Consequently, the first-order dermal absorption rates from Worksheet B03b are used in 2 the current analysis to reflect the underlying uncertainty in the estimates using the 95% 3 confidence interval. These first-order dermal absorption rates of 0.0041 hour-1 (0.0015 to 0.012 4 hour-1) from Worksheet B03b are entered into Worksheet B01 and are used for all exposure 5 assessments that require estimates of the first-order dermal absorption rates. 6 For general exposures to workers, estimates of worker exposure rates are required for ground 8 broadcast applications of Dursban 50W (Section 1.2.2). As discussed in Section 2, the 9 WorksheetMaker workbook that accompanies this risk assessment is based on ground broadcast 10 applications. As detailed in Table 14 of SERA (2014b), the reference chemical for this 11 application method is 2,4-D with a first-order dermal absorption rate of 0.00066 hour-1. As 12 discussed above, central estimates of the dermal absorption rates used for chlorpyrifos in this 13 analysis are 0.0015 hour-1. The central estimate of the worker exposure rate for ground 14 broadcast applications of 2,4-D is 0.0001 with a 95% prediction interval of 0.000002 to 0.005 15 mg/kg bw per lb a.i. handled. To account for differences in the estimated dermal absorption of 16 chlorpyrifos and 2,4-D, the worker exposure rates for 2,4-D are multiplied by the ratio of the 17 dermal absorption rate coefficient for chlorpyrifos to the corresponding value for 2,4-D—i.e., 18

0.0014 ÷ 0.00066 ≈ 2.1212…. These calculations are detailed in Table 2-1. Following the 19 convention in SERA (2014b), the resulting worker exposure rates are rounded to one significant 20 digit, and the occupational exposure rates for ground broadcast applications of chlorpyrifos are 21 estimated as 0.0002 (0.000004 to 0.01) mg/kg bw per lb handled. These worker exposure rates 22 are entered in Worksheet C01 of the EXCEL workbook that accompanies this risk assessment. 23 Note that Table 2 is a compound table. Table 2-1 gives the worker exposure rates for ground 24 broadcast applications. Table 2-2 gives the worker exposure rates for directed foliar backpack 25 applications as discussed further below. 26 The Forest Service indicated that Dursban 50W may be applied by backpack applications and 28 DuraGuard ME may be used in greenhouse applications. As discussed in SERA (2014b, Section 29 3.2.3.4.2), worker exposure rates for greenhouse applications may be based on the rates for 30 backpack applications. As a convenience for the worker exposure assessment of backpack and 31 greenhouse applications, Table 2-2 gives the derivation of worker exposure rates for backpack 32 applications. As detailed in Table 14 of SERA (2014b), three reference chemicals with 33 corresponding worker exposure rates are given for backpack and greenhouse applications—i.e., 34 glyphosate (ka = 0.00041 hour-1), 2,4-D (ka = 0.00066 hour-1), and triclopyr BEE (ka = 0.0031 35 hour-1). To minimize extrapolation, triclopyr BEE is used as the reference chemical for 36 chlorpyrifos. As noted above, the central estimate of the first-order dermal absorption rate 37 coefficient for chlorpyrifos is taken as 0.0041hour-1. The central estimate of the worker exposure 38 rate for triclopyr BEE is 0.01 with a 95% prediction interval of 0.002 to 0.06 mg/kg bw per lb a.i. 39 handled. To account for differences in the estimated dermal absorption of chlorpyrifos and 40 triclopyr BEE, the worker exposure rates are multiplied by the ratio of the dermal absorption rate 41 coefficient for chlorpyrifos to the corresponding value for triclopyr BEE [0.0014 ÷ 0.0031 ≈ 42 0.4516]. As detailed in Table 2-2, the resulting estimates of the worker exposure rates for 43 chlorpyrifos are about 0.00451 (0.000903 to 0.027096) mg/kg bw per lb handled. Following the 44 convention in SERA (2014b), the resulting worker exposure rates are rounded to one significant 45 digit—i.e., 0.005 (0.0009 to 0.03) mg/kg bw per lb handled. 46

As discussed in Section 1.2.2, the Forest Service indicated that workers will handle up to 5 lbs 1 a.i./day. This amount is used in Worksheets C01 to estimate doses for workers. 2 In addition to general exposures, four standard accidental exposure scenarios discussed in SERA 4 (2014a, Section 3.2.2.2) are also considered, and they are detailed in Worksheets C02a,b and 5 C03a,b. 6

3.2.2. General Public 7

As detailed in SERA (2014a, Section 3.2.3), Forest Service risk assessments provide a standard 8 set of exposure scenarios for members of the general public. These exposure scenarios are 9 applicable to standard forestry applications of pesticides and these scenarios are included in the 10 WorksheetMaker workbook that accompanies this document. The applicability of these 11 scenarios to nursery applications, particularly applications within greenhouses, is probably 12 minimal. 13 While most of the exposure scenarios given in the WorksheetMaker workbook are standard for 15 Forest Service risk assessments, one notable exception is the surface water modelling. Full 16 Forest Service risk assessments typically estimate concentrations of a pesticide in surface water 17 using GLEAMS-Driver (SERA 2014a, Section 3.2.3.4.3). In the interest of economy, the current 18 analysis uses FIRST (FQPA Index Reservoir Screening Tool). FIRST is a Tier 1 model 19 developed by the U.S. EPA to estimate concentrations of pesticide in surface water and details of 20 the FIRST model are available at http://www.epa.gov/oppefed1/models/water/first_description.htm. 21 The input parameters and the estimated surface water concentrations of chlorpyrifos are 23 summarized in Table 3. The output files from FIRST are given in Appendix 1. Most of the 24 chemical specific inputs are taken from Table 1. As with standard GLEAMS-Driver modeling, a 25 unit application rate of 1 lb a.i./acre is used. The results from the modeling are entered into 26 Worksheet B04Rt as water contamination rates—i.e., mg/L per lb a.i./acre applied. These 27 concentrations are adjusted for the functional application rate discussed in Section 2. Note that 28 the number of applications and the application interval are sensitive input parameters for FIRST. 29 Based on information from the JH Stone Nursery, a total of 5 applications with an application 30 interval of 14 days are used. Decreasing the number of applications or increasing the application 31 interval will lead to lower estimates of chlorpyrifos concentrations in water. 32 One very important input parameter for FIRST is the proportion of the watershed that is treated. 34 As indicated in Table 2, the FIRST modeling was conducted using a proportion of 1.0—i.e., the 35 entire watershed is treated. Based on inputs from the Forest Service (Section 2), the raw output 36 values from FIRST in Table 2 are reduced using a proportion of 0.023 as derived in 37 Section 1.2.2. These values are labeled as “Adjusted Output” in Table 3 and are entered into 38 Worksheet B04Rt in units of mg a.i./L per lb a.i./acre and rounded to two significant places. 39 Forest Service risk assessments also include an accidental spill scenario. Typically, Forest 41 Service risk assessments use a spill volume of 100 (20 to 200) gallons. As discussed in 42 Section 1.2.2, chlorpyrifos will be applied to only 5 acres at an application volume of 43 30 gallons/acre. Thus, the maximum amount available for an accidental spill would be 44 150 gallons. Thus, the upper bound of the spill volume is Worksheet A01 is set to 150 gallons 45 rather than the typical upper bound of 200 gallons. 46 http://www.epa.gov/oppefed1/models/water/first_description.htm

Another major set of standard exposure scenarios for members of the general public involves the 2 consumption of contaminated vegetation. These exposure scenarios are described fully in SERA 3 (2014a, Section 3.2.3.7). These exposure scenarios were developed for forestry applications of 4 pesticides to areas with edible vegetation that might be consumed by the general public. As 5 discussed further in Section 3.4.2 (risk characterization for members of the general public), these 6 exposure scenarios may be of limited relevance to a nursery environment. 7

3.3. Dose-Response Assessment 8

The dose-response assessment for chlorpyrifos is summarized in Table 4 from the most recent 9 EPA human health risk assessment (U.S. EPA/OPP/HED 2011a). Three toxicity values are used 10 in the WorksheetMaker workbook: an acute RfD of 0.0036 mg/kg bw/day, a chronic RfD of 11

0.0003 mg/kg bw/day, and an intermediate RfD of 0.0015 mg/kg bw/day. All of the toxicity 12 values are based on cholinesterase (ChE) inhibition in red blood cells using an uncertainty factor 13 of 100. For all three toxicity values, the uncertainty factor of 100 is based on a factor of 10 for 14 interspecies extrapolation and a factor of 10 for intraspecies variation (i.e., differences in 15 sensitivity among individuals). The intermediate RfD is based on a standard NOAEL. The acute 16 and chronic RfDs, however, are based on benchmark doses, specifically doses associated with a 17 10% inhibition of ChE in red blood cells. As detailed in U.S. EPA (2012b), benchmark doses 18 are treated in a manner similar to NOAELs and are particularly useful for responses such as ChE 19 inhibition for which detailed dose/response data are available. Benchmark doses are beneficial 20 in that they more fully account for the response data at several doses rather than relying on a 21 single experimental dose (i.e., the NOAEL). 22 Additional details on the analyses of the neurotoxicity studies are presented in U.S. 24 EPA/OPP/HED (2011b). 25 No chronic (i.e., lifetime) exposures are anticipated at the JH Stone Nursery. Consequently, the 27 acute RfD of 0.0036 mg/kg bw is used for the risk characterization of acute/single exposure 28 scenariosm and the intermediate RfD of 0.0015 is used for exposures that may occur over the 29 course of several months (e.g., longer-term dietary exposures). 30

3.4. Risk Characterization 31

3.4.1. Workers 32

The risk characterization for workers is given in Worksheet E02 of the WorksheetMaker EXCEL 33 workbook that accompanies this report. 34 All of the accidental exposure scenarios exceed the level of concern at the upper bounds. The 36 exposure scenarios for wearing contaminated gloves lead to much higher HQs (up to 800) than 37 the accidental exposure scenarios involving spills onto the hands (maximum HQ of 1.3) or lower 38 legs (maximum HQ of 3). These exposure scenarios, particularly those for contaminated gloves, 39 underscore the need for proper and careful handling of chlorpyrifos. In the event of an 40 accidental exposure, aggressive measures to limit exposure are clearly warranted, and medical 41 follow-up would be advisable in the event of any signs of neurological impairment. 42

For non-accidental exposure scenarios, HQs are given for both acute (1-day) exposures based on 1 the acute RfD of 0.0036 mg/kg bw/day and for intermediate exposures (several months) based on 2 the intermediate toxicity value of 0.0015 mg/kg bw/day. No chronic exposures are anticipated, 3 given the limited use of chlorpyrifos at the JH Stone Nursery. The HQs for acute exposures are 4

0.3 (0.006 to 14) and the HQs for intermediate exposures are 0.7 (0.01 to 33). The most 5 reasonable interpretation of these HQs is that chlorpyrifos may be applied with no substantial 6 risks to workers at typical (i.e., central estimates) exposure levels. Poor handling practices or 7 other factors that would increase exposure could lead to unacceptable exposure levels for 8 workers. 9 The EPA uses a substantially different methodology in assessing worker exposure (SERA 2009, 11 Section 4.1). A direct and detailed comparison of the most recent EPA risk characterization for 12 workers to the risk characterization given in Worksheet E02 cannot be made. Nonetheless, the 13 qualitative assessment from EPA is similar to that given in the current analysis: 14

Of the 305 exposure scenarios assessed 134 had risk estimates that did not 16 exceed the level of concern at some level of personal protection (i.e. ARIs 17 are > 1). Ninety-one (91) exposure scenarios had risk estimates not of 18 concern when engineering controls were considered. The remaining 80 19 scenarios resulted in risk estimates of concern (i.e. ARIs are < 1) at all 20 levels of personal protection and engineering controls considered. 21

U.S. EPA/OPP/HED (2011a, p. 83). 22 Note that an ARI (Aggregate Risk Index) of less than 1 is analogous to an HQ of greater than 1. 24 In other words, the EPA notes that some applications of chlorpyrifos can result in risks to 25 workers that are a concern. This conclusion is essentially identical to that given in the current 26 analysis. 27

3.4.2. General Public 28

The HQs for members of the general public are summarized in Worksheet E04 for the 29 WorksheetMaker EXCEL workbook that accompanies this report. 30

All accidental exposure scenarios (i.e., an accidental spray of a child or a woman or an accidental 31 spill of chlorpyrifos into a small body of water) lead to HQs that exceed the level of concern 32 (HQ=1). Even the lower bounds of the HQs (the least severe accidents considered) exceed the 33 level of concern, except for the direct spray of a woman’s legs (HQ=0.6), which approaches a 34 level of concern. The highest HQs are associated with the upper bounds for the consumption of 35 contaminated fish by members of the general public (upper bound HQ=1823) or subsistence 36 populations (upper bound HQ=8882). The HQs require little interpretation. As with worker 37 exposures, accidents involving members of the general public could lead to exposure levels that 38 grossly exceed acceptable levels. Also as with workers, aggressive measures to limit exposure 39 are clearly warranted, and medical follow-up could be advisable in the event of any signs of 40 neurological impairment. 41

As also summarized in Worksheet E04, many non-accidental acute and longer-term exposure 1 scenarios lead to HQs that exceed the level of concern. The higher HQs are associated with the 2 consumption of contaminated vegetation—i.e., HQs of up to 375. The relevance of these HQs to 3 a nursery application of any pesticide would depend on the public access to the vegetation and 4 whether the treated vegetation treated is edible. These factors should be considered on a case-5 by-case basis. 6

The HQs associated with the consumption of contaminated fish are lower than those for 7 contaminated vegetation (i.e., an upper bound HQ of 25 for the acute consumption of 8 contaminated fish). These HQs, however, may be of greater practical concern. Given the nature 9 of the FIRST modeling (Section 3.2.2), the estimated surface water concentrations would 10 represent expected contamination of surface water. If individuals might take fish from 11 contaminated water in the area of a nursery, more refined site-specific modeling may be justified. 12

4. ECOLOGICAL EFFECTS 1

4.1. Hazard Identification 2

As with the hazard identification for human health (Section 3.2), the hazard identification for 3 ecological effects is highly abbreviated in the current document. The overall database for 4 ecological effects is discussed in detail in U.S. EPA/OPP/EFED (2009), the most recent 5 ecological risk assessment on chlorpyrifos. Specific toxicity values for different groups of 6 receptors are discussed in Section 4.3. 7 Mechanistically, the inhibition of AChE is the endpoint of concern in most groups of ecological 9 receptors. This endpoint, however, is not relevant to terrestrial plants; accordingly, standard 10 toxicity studies on terrestrial plants were not conducted. Based on incident data, however, U.S. 11 EPA/OPP/EFED (2009, p. 155) indicates a qualitative concern that chlorpyrifos may damage 12 terrestrial plants, but not to a significant degree (p. 161). 13

4.2. Exposure Assessment 14

As detailed in SERA (2014a), most Forest Service risk assessments are based on a series of 15 exposure assessments for mammals, birds, terrestrial invertebrates, plants, and aquatic 16 organisms. These exposure assessments are detailed in the WorksheetMaker book that 17 accompanies this document. Details of and the rational for these exposure assessments are given 18 Section 4.2 of SERA (2014a) and are not discussed further in the current assessment. One 19 exception, however, involves terrestrial plants. Because standard toxicity studies are not 20 available on terrestrial plants, standard exposure assessments for terrestrial plants are not given 21 in the WorksheetMaker workbook that accompanies this document. 22

4.3. Dose-Response Assessment 23

The dose response assessments for nontarget organisms are summarized in Table 5 and discussed 24 in the following subsections on different groups of receptors. 25

4.3.1. Terrestrial Organisms 26

4.3.1.1. Mammals 27

As discussed in Section 3.3 and summarized in Table 4, the acute oral RfD is based on a 10% 28 benchmark dose of 0.36 mg/kg bw in rats for ChE inhibition in red blood cells. Following 29 standard practice in Forest Service risk assessments (SERA 2014a), the basis for the human 30 health risk assessment is used as the basis for assessing effects in mammalian wildlife. Thus, the 31 NOAEL of 0.36 mg/kg bw is used to characterize risks to mammals following short-term oral 32 exposures to chlorpyrifos. Similarly, the NOAEL of 0.03 mg/kg bw/day from a study in 33 pregnant rats (which is the basis for the chronic oral RfD) is used to assess the consequences of 34 longer-term oral exposures of mammalian wildlife to chlorpyrifos. 35

4.3.1.2. Birds 36

In the most recent EPA ecological risk assessment, U.S. EPA/OPP/EFED (2009, Table 4.5, p. 37

106) cites three toxicity values for birds: an acute gavage LD50 of 5.62 mg/kg bw in the common 38 grackle (MRID 40378401), an acute dietary LC50 of 136 ppm in mallard ducks (MRID 39 00095007), and a chronic dietary NOAEC of 25 ppm in mallard ducks with an LOAEC of 60 40 ppm based on a reduced number of eggs and reduced body weight in both sexes of the parental 1 birds. 2 Forest Service risk assessments typically express toxicity values for birds in units of mg/kg bw 4 rather than dietary concentrations. In converting dietary concentrations to mg/kg bw doses, 5 approximate food consumption rates in acute dietary studies are about 0.4 kg food/kg bw for 6 mallards and 0.3 kg food/kg bw for quail. These food consumption rates are from standard 7 studies using very young birds. Approximate food consumption rates during reproduction 8 studies are about 0.07 kg food/kg bw. Using these conversion factors, the acute dietary LC50 of 9 136 ppm in mallards corresponds to a dose of 54.4 mg/kg bw [136 mg/kg food x 0.4 kg food/kg 10 bw ≈ 54.4 mg/kg bw] and the longer-term dietary concentration of 25 ppm corresponds to a dose 11 of 1.75 mg/kg bw [25 mg/kg food x 0.07 kg food/kg bw/day ≈ 1.75 mg/kg bw/day]. 12 The acute gavage LD50 of 5.62 mg/kg bw is substantially below the estimated acute dietary LD50 14 of 54.4 mg/kg bw. This pattern is not unusual and may reflect a higher bioavailability and/or 15 higher peak body burdens in gavage studies, relative to dietary studies. As a conservative 16 approach, the lower acute gavage LD50 of 5.62 mg/kg bw is used as the basis for the dose-17 response assessment in birds. Following the risk presumption approach used by U.S. EPA for 18 terrestrial species (discussed in SERA 2014a, Section 4.3.2 and Table 19), the acute NOAEC for 19 birds is estimated as 0.562 mg/kg bw—i.e., the LD50 is divided by a factor of 10 to approximate 20 the NOAEC. 21 As discussed above, the longer-term dietary NOAEC is estimated as 1.75 mg/kg bw/day. This 23 estimated longer-term NOAEC is higher than the estimated acute NOAEC by a factor of about 24 3 [1.75 mg/kg bw/day ÷ 0.562 mg/kg bw ≈ 3.114]. It is not sensible to use a chronic NOAEC 25 that is higher than the acute NOAEC. As noted above, the acute toxicity value is based on a 26 gavage study. Dietary studies are probably more relevant than gavage studies for the exposure 27 scenarios considered in this document. Thus, the estimated longer-term dietary NOAEC is 28 estimated as 1.75 mg/kg bw/day is applied to the risk characterization for acute and longer-term 29 exposure of birds. 30 Note that the above approach is not the most conservative—i.e., using the lower estimated acute 32 NOAECs for both acute and chronic exposures would result in lower toxicity values. 33 Nonetheless, preference is given to experimental toxicity values (in this case the chronic toxicity 34 values) rather than inherently conservative approximations. 35

4.3.1.3. Other Terrestrial Organisms 36

The only other toxicity value for terrestrial organisms discussed in U.S. EPA/OPP/EFED (2009) 37 is an acute contact LD50 of 0.059 µg/bee (MRID 05001991), which is equivalent to 0.000059 38 mg/bee. Using a typical body weight of 116 mg (equivalent to 0.000116 kg) for a worker bee 39 (SERA 2014a, Section 4.2.3.1), the dose of 0.000059 mg/bee is equivalent to about 0.51 mg/kg 40 bw [0.000059 mg/bee ÷ 0.000116 kg bw ≈ 0.5086 mg/kg bw]. As with birds and for the same 41 reason (i.e., the EPA risk presumption approach), the estimated LD50 of 0.51 mg/kg bw is 42 divided by 10 to approximate an NOAEC of 0.051 mg/kg bw. 43

4.3.2. Aquatic Organisms 1

As discussed in the following sections, chlorpyrifos is classified as very highly toxic to aquatic 2 animals, and U.S. EPA/OPP/EFED (2009) reports toxicity values in units of µg/L. In the 3 WorksheetMaker workbook that accompanies this report, toxicity values are expressed in units 4 of mg/L. In the following sections, toxicity values are generally discussed in units of µg/L to 5 maintain consistency with the EPA source document. All toxicity values used in the 6 WorksheetMaker workbook, however, are also expressed in units of mg/L (typically in 7 parentheses) for clarity. 8

4.3.2.1. Fish 9

Following standard practice, U.S. EPA/OPP/EFED (2009, Table 4.1) summarizes toxicity data 10 on freshwater and marine/estuarine fish. Unless there are obvious and substantial differences in 11 the sensitivity of freshwater and saltwater fish, Forest Service risk assessments generally focus 12 on identifying the most sensitive and most tolerant species of fish. Based on the data 13 summarized in U.S. EPA/OPP/EFED (2009), no such substantial differences are apparent. The 14 reported acute LC50 values are 1.8 µg/L for freshwater fish (bluegill sunfish) and 0.70 µg/L for 15 saltwater fish (tidewater silverside). Following the risk presumption approach used by U.S. EPA 16 for aquatic species (discussed in SERA 2014a, Section 4.3.2 and Table 19), acute LC50 values 17 are divided by a factor of 20 to approximate the NOAEC. Thus, the estimated acute NOAECs 18 would be 0.09 µg/L for bluegills and 0.035 µg/L for silversides. 19 U.S. EPA/OPP/EFED (2009) also reports chronic NOAEC values of 0.28 µg/L for Atlantic 21 silversides and 0.57 µg/L for fathead minnows. As with birds (Section 4.3.1.2), the estimated 22 acute NOAECs are below the experimental chronic NOAECs. Consequently, the chronic 23 NOAECs are used to characterize risks from both acute and chronic exposures. While using the 24 lower estimated acute NOAECs might be viewed as a more conservative approach, preference is 25 given to the experimental NOAECs (i.e., the chronic values) rather than intentionally 26 conservative approximations of the acute NOAECs. 27 There is not a substantial difference between the two chronic values—i.e., 0.57 µg/L ÷ 0.28 µg/L 29 ≈ 2.03. In the absence of additional data on the chronic effects of chlorpyrifos in fish, however, 30 the lower NOAEC of 0.28 µg/L (0.00028 mg/L) is used for potentially sensitive species of fish 31 and the higher NOAEC of 0.57 (0.00057 mg/L) is used for potentially tolerant species of fish. 32

4.3.2.2. Amphibians 33

Two toxicity values are available on aquatic-phase amphibians in U.S. EPA/OPP/EFED (2009): 34 an acute LC50 of 0.6 µg/L and a chronic LOAEC of 0.1 µg/L. Both of these toxicity values are 35 for the African clawed frog tadpole (Xenopus laevis). Again using the risk presumption 36 approach from U.S. EPA for aquatic species (discussed in SERA 2014a, Section 4.3.2 and Table 37 19), the acute LC50 value is divided by a factor of 20 to approximate an NOAEC of [0.6 µg/L ÷ 38 20 = 0.03 µg/ or 0.00003 mg/L]. 39 LOAEL values are not used directly in Forest Service risk assessments. Following the general 41 approach discussed in SERA (2014a, Section 3.3 and Table 15), the chronic LOAEC of 0.1 µg/L 42 is divided by a factor of 10 to approximate an NOAEC of 0.01 µg/L or 0.00001 mg/L. 43

4.3.2.3. Aquatic Invertebrates 1

As summarized in U.S. EPA/OPP/EFED (2009, Table 4.1), for aquatic invertebrates, acute LC50 2 values are available in a freshwater species (i.e., 0.07 µg/L for Daphnia magna) and a saltwater 3 species (0.035 µg/L in mysid shrimp, Americamysis bahia). Dividing these toxicity values by a 4 factor of 20, the approximated acute NOAEC values are 0.0035 µg/L for Daphnia magna and 5

0.00175 µg/L for mysid shrimp. 6 For Daphnia magna, the reported chronic NOAEC is 0.025 µg/L (0.000025 mg/L). As with the 8 toxicity values for fish, preference is given to the experimental chronic NOAEC, and this toxicity 9 value is used to characterize risks associated with both acute and longer term exposures in 10 presumably tolerant species of aquatic invertebrates. 11 For mysids, a chronic LOAEC of 0.0046 µg/L is available. Following the same approach used 13 with amphibians (Section 4.3.2.2), the LOAEC of 0.0046 µg/L is divided by a factor of 10 to 14 approximate a chronic NOAEC of 0.00046 µg/L (0.00000046 mg/L). This estimated chronic 15 NOAEC is below the estimated acute NOAEC of 0.00175 µg/L for mysid shrimp. 16 Consequently, the estimated acute NOAEC of 0.00175 µg/L (0.00000175 mg/L) is used to 17 characterize risks of acute exposures, and the estimated chronic NOAEC 0f 0.00046 µg/L 18 (0.00000046 mg/L) is used to characterize risks of longer-term exposures in presumably 19 sensitive species of aquatic invertebrates. 20

4.3.2.4. Aquatic Plants 21

The only toxicity data on aquatic plants in U.S. EPA/OPP/EFED (2009, Table 4.1 and p. 161) is 22 a reported 120-hour EC50 of 140 µg/L based on reduced cell density in a species of freshwater 23 green alga (Pseudokirchneriella subcapitata). Dividing this EC50 by 20, the NOAEC is 24 approximated as 7 µg/L (0.007 mg/L). As would be expected based on the known mechanism of 25 action of chlorpyrifos, the EC50 in plants is substantially higher than EC50 values in aquatic 26 animals—i.e., higher than the EC50 of 0.035 µg/L in sensitive aquatic invertebrates by a factor of 27 4000 [140 µg/L ÷ 0.035 µg/L]. 28 No data on the toxicity of chlorpyrifos to aquatic macrophytes is presented in U.S. 30

EPA/OPP/EFED (2009). 31

4.4. Risk Characterization 32

4.4.1. Terrestrial Organisms 33

4.4.1.1. Mammals and Birds 34

The HQs for mammals and birds are summarized in Worksheet G02a (mammals) and Worksheet 35 GO2b (birds) of the WorksheetMaker workbook that accompanies this risk assessment. As 36 summarized in Table 5, the NOAECs for mammals are much less than those for birds. While 37 this leads to much lower HQs for mammals, relative to birds, the qualitative risk characterization 38 is similar for both groups of organisms. 39 The exposure scenarios for the consumption of contaminated water do not lead to HQs that 41 exceed the level of concern, even in the event of an accidental spill (HQ=1). Some of the upper 42 bound estimates of exposures associated with the consumption of contaminated fish exceed the 1 level of concern for mammals (maximum HQ of 39) and birds (maximum HQ of 3). 2 Most of the upper bound HQs for the consumption of contaminated vegetation, grass, and fruit 4 exceed the level of concern for both mammals (maximum HQ of 1382) and birds (maximum HQ 5 of 976). Even at the lower bounds of the HQs, exceedances occur for mammals (maximum HQ 6 of 30) and birds (maximum HQ of 22). 7 While the EPA uses a somewhat different approach to estimating exposures in mammals and 9 birds from that used in Forest Service risk assessments (SERA 2009, Section 4.3, Table 5), the 10 risk characterization for mammals and birds given in the current report is qualitatively similar to 11 that in the EPA’s most recent ecological risk assessment on chlorpyrifos: 12

…the acute and chronic avian and mammalian dose and dietary-based 14 RQs estimated with T-REX exceed the LOCs listed species for all uses of 15 chlorpyrifos, including granular and seed treatment uses. 16

U.S. EPA/OPP/EFED (2009, Section 5.5.1.2, p. 150). 17 Note that T-REX is a spreadsheet tool used by EPA which, similar to WorksheetMaker, is used 19 to conduct exposure assessments and RQs (Risk Quotients). RQs are analogous to HQs used in 20 Forest Service risk assessments. 21 The severe risk characterization for birds is supported by field incidents in which birds were 23 killed following applications of chlorpyrifos with an association of “probably” or “highly 24 probably” (U.S. EPA/OPP/EDED 2009, p. 154). 25 As discussed in Section 3.4.2, the high HQs for humans associated with the consumption of 27 contaminated vegetation may have limited practical relevance, because the general public may 28 have limited access to the JH Stone Nursery, and it does not seem likely that the Forest Service 29 would be cultivating edible/agricultural crops. The extent to which these reservations would 30 apply to mammalian wildlife and birds is not clear. 31

4.4.1.2. Invertebrates 32

The HQs for the direct spray or drift exposures of the honeybee (with or without foliar 33 interception) are given in Worksheet G09. As would be expected with an effective insecticide, 34 the direct spray of honeybees with chlorpyrifos leads to a very high HQ (1345). Depending on 35 the extent of foliar interception, HQs can exceed the level of concern (HQ=1) at distances of up 36 to 500 feet downwind. As detailed in SERA (2011a, Section 3.3.2), drift modeling should be site 37 specific, and more elaborate estimates of drift could be made on a case-by-case basis. 38 As with the risk characterization for birds, the severe risk characterization for bees is supported 40 by field incidents in which bees were killed following applications of chlorpyrifos with an 41 association of “probably” or “highly probably” (U.S. EPA/OPP/EDED 2009, p. 154). 42

4.4.2. Aquatic Organisms 43

The HQs for aquatic organisms are summarized in Worksheet G03 for accidental, acute non-44 accidental, and longer-term exposure scenarios.