Appendix D

Human Health Risk Assessment

Storrie and Rich Fire Areas Invasive Plant Treatment Project Appendix D

2

Introduction

The herbicide treatments proposed under the Storrie and Rich Fire Areas Invasive Plant Treatment Project (Storrie IPT Project) present some risks to human health and safety. The purpose of this appendix is to present a summary of the potential risks to human health from the proposed herbicide treatments. This summary is a series of excerpts taken from the relevant risk assessments completed by Syracuse Environmental Research Associates (SERA 2004, 2007a, 2011a, 2011b, 2011c) with only minor edits for readability and document flow.

The application of aminopyralid, glyphosate, chlorsulfuron, triclopyr, and imazapyr, as proposed by the Storrie IPT Project, is expected to present a low risk to human health and safety. Based on the available information, the addition of the proposed surfactant and dye, would also pose a low risk to human health and safety. The incorporation of Best Management Practices (USDA 2012, included in Appendix A) would also reduce the level of exposure and associated risk to the health and safety of workers and members of the general public. This is based on the analysis included in the SERA risk assessments (SERA 2004, 2007a, 2011a, 2011b, 2011c).

Summary of Project Proposal Five herbicides (aminopyralid, glyphosate, chlorsulfuron, triclopyr, and imazapyr) are proposed for control of invasive plants within the Storrie IPT Project area. Proposed application rates and chemical treatment acreage caps for each stream reach are described in detail in the proposed action section of the EA. No more than 100 acres of chemical treatment per year would occur in the entire project area of approximately 90,000 acres. A non-ionic modified vegetable oil surfactant (such as Competitor® or an equivalent formulation) and a marker dye (such as Hi-Lite Blue® or an equivalent formulation) would also be used to increase the efficacy of the herbicide treatments.

The proposed applications would comply with all applicable state and federal regulations for the safe use of pesticides (including the label requirements). For example, applicators would be adequately trained, medical aid would be available, wash water and eye wash water would be on-site or nearby, and personal protective equipment would be used (e.g. eye protection, gloves, long-sleeved shirt, and long pants). Best Management Practices for pesticide application, including a spill contingency plan, would be implemented.

Chapter 2 provides details of the herbicide treatment design elements that are proposed.

Hazard Analysis The hazards associated with using aminopyralid, glyphosate, chlorsulfuron, triclopyr, and imazapyr have been determined through comprehensive reviews of available toxicological studies; these reviews, which are compiled in a group of risk assessments completed by Syracuse Environmental Research Associates (SERA) (SERA 2004, 2007a, 2011a, 2011b, 2011c) under contract with the Forest Service, are incorporated by reference into this risk assessment. Copies of these risk assessments are included in the project record. The following sections includes relevant portions of the hazard analysis provided in the most recent SERA risk assessments (SERA 2004, 2007a, 2011a, 2011b, 2011c).

A note specific to impurities and metabolites - virtually no chemical synthesis yields a totally pure product. Technical grade herbicides, as with other technical grade products, undoubtedly

Appendix D Mt. Hough Ranger District, Plumas National Forest

3

contain some impurities. The U.S. Environmental Protection Agency (EPA) defines the term impurity as “…any substance…in a pesticide product other than an active ingredient or an inert ingredient, including un-reacted starting materials, side reaction products, contaminants, and degradation products” (40 CFR 158.153(d)). To some extent, concern for impurities in technical grade herbicides is reduced by the fact that the existing toxicity studies on these herbicides were conducted with the technical grade product. Thus, if toxic impurities are present in the technical grade product, they are likely to be encompassed by the available toxicity studies on the technical grade product. An exception to this general rule involves carcinogens, most of which are presumed to act by non-threshold mechanisms. Because of the non-threshold assumption, any amount of a carcinogen in an otherwise non-carcinogenic mixture is assumed to pose some carcinogenic risk.

As with contaminants, the potential effect of metabolites on a risk assessment is often encompassed by the available in vivo toxicity studies under the assumption that the toxicological consequences of metabolism in the species on which toxicity studies are available will be similar to those in the species of concern (human in this case). Uncertainties in this assumption are encompassed by using an uncertainty factor in deriving the reference dose (RfD) and may sometimes influence the selection of the study used to derive the RfD.

Unless otherwise specifically referenced, all of the information in the following sections was taken directly from the most recent SERA risk assessment (SERA 2004, 2007a, 2011a, 2011b, 2011c).

Exposure Assessment This exposure assessment examines the potential health effects to two groups of people that are most likely to be exposed to aminopyralid, glyphosate, chlorsulfuron, triclopyr, or imazapyr; workers and members of the public. Workers include applicators, supervisors, and other personnel directly involved in the application of herbicides. The public includes other Forest Service personnel, visitors, or nearby residents who could be exposed through herbicide drift, contact with sprayed vegetation, by drinking water that contains herbicide residue, or by eating contaminated vegetation (such as berries or foliage), game, or fish.

In these analyses, data are displayed for three different exposure scenarios: typical, lower, and upper. The upper level represents a conservative estimate of a worst-case scenario resulting from the highest application rate, lowest dilution rate, and largest number of acres treated per day. This approach is used to encompass as broadly as possible the range of potential exposures.

Workers Pesticide applicators are the individuals who are most likely to be exposed to a pesticide during the application process. For purpose of this analysis, two different types of worker exposure assessments were considered: general and accidental/incidental. General exposure scenarios were used to analyze exposure resulting from normal use (i.e. handling and application) of the chemicals (SERA 2007b). Accidental and incidental exposure scenarios were used to analyze specific types of exposures associated with mischance or mishandling of a chemical (SERA 2007b).

Exposure rates for workers are calculated using a number of factors that include: proposed application rates, dilution rates, estimated hours worked per day, number of acres treated per hour and human dermal absorption rates. As described in SERA (2007b), worker exposure rates

Storrie and Rich Fire Areas Invasive Plant Treatment Project Appendix D

4

are expressed in units of milligrams (mg) of absorbed dose per kilogram (kg) of body weight per pound of chemical handled (mg/kg/lb applied).

General Public Under normal conditions, members of the general public should not be exposed to substantial levels of aminopyralid, glyphosate, chlorsulfuron, triclopyr, or imazapyr. Nonetheless, exposure scenarios can be constructed for the general public, depending on various assumptions regarding application rates, dispersion, canopy interception, and human activity. Several highly conservative scenarios are utilized to characterize this risk.

The two types of exposure scenarios developed for the general public include acute exposure and longer-term or chronic exposure. All of the acute exposure scenarios are primarily accidental. They assume that an individual is exposed to the compound either during or shortly after its application. Specific scenarios are developed for direct spray, dermal contact with contaminated vegetation, and consumption of contaminated fruit, vegetation, water, and fish. Most of these scenarios should be regarded as extreme, some to the point of limited plausibility (SERA 2007b). The longer-term or chronic exposure scenarios parallel the acute exposure scenarios for the consumption of contaminated fruit, vegetation, water, and fish but are based on estimated levels of exposure for longer periods after application.

Imazapyr (SERA 2011b)

OVERVIEW Imazapyr is an effective herbicide for the control of both terrestrial and aquatic vegetation. Under some conditions, terrestrial applications of imazapyr could damage nontarget terrestrial vegetation. While imazapyr is an effective terrestrial herbicide, the exposure scenarios developed in the current risk assessments lead to a wide range of HQs, some of which are far below the level of concern and others which exceed the level of concern substantially. This apparent ambiguity relates to the attempt made in the exposure assessments to encompass a wide range of potential exposures associated with different weather patterns and other site-specific variables. Thus, for applications of imazapyr in areas where potential effects on nontarget plants are a substantial concern, refinements to the exposure scenarios for nontarget plants would be appropriate.

While adverse effects on plants may be anticipated, there is no basis for asserting that applications of imazapyr will pose any substantial risk to humans or other species of animals. The U.S. EPA/OPP classifies imazapyr as practically non-toxic to mammals, birds, honeybees, fish, and aquatic invertebrates. This classification is clearly justified. None of the expected (non-accidental) exposures to these groups of animals raise substantial concern; indeed, most accidental exposures raise only minimal concern. The major uncertainties regarding potential toxic effects in animals are associated with the lack of toxicity data on reptiles and amphibians.

PROGRAMS DESCRIPTION Imazapyr is a nonselective herbicide used to control a variety of grasses, broadleaf weeds, vines, and brush species. In Forest Service programs, imazapyr is used primarily in the Southern United States for invasive plant control, conifer release, and site preparation. Previous Forest Service risk assessments on imazapyr consider only four BASF formulations: Arsenal, Arsenal AC (applicators concentrate), Chopper, and Stalker, all of which contain imazapyr as the isopropylamine salt.

Appendix D Mt. Hough Ranger District, Plumas National Forest

5

Imazapyr is now off patent, and numerous formulations are available both from BASF and other companies. The current risk assessment explicitly considers 16 formulations of imazapyr but is intended to encompass all formulations of the isopropylamine salt of imazapyr registered for forestry or other related applications including applications for the control of emergent aquatic vegetation.

While imazapyr formulations can be used in pre-emergence applications, the most common and effective applications are post-emergent when the vegetation to be controlled is growing vigorously. The most common methods of ground application involve backpack (selective foliar) and boom spray (broadcast foliar) operations. Cut surface treatment methods may also be used in Forest Service programs involving imazapyr. Boom spray applications are used primarily in rights-of-way management.

Imazapyr is not used extensively in agriculture, except for applications to corn in the central United States. At least in the southeastern United States, forestry applications of imazapyr may be greater than agricultural applications. This also appears to be the case in California, although these forestry uses in California are not directly associated with Forest Service programs. Current information on the use of imazapyr in forestry applications in other regions of the United States is sparse.

HAZARD IDENTIFICATION The toxicity of imazapyr is relatively well-characterized in experimental studies conducted on mammals. Most of the available information on the toxicity of imazapyr to mammals comes from standard studies submitted to the U.S. EPA/OPP in support of the registration of imazapyr. Full copies of these studies, which are considered proprietary, were not available for the preparation of the current Forest Service risk assessment (Sera 2011b); however, these studies were available during the preparation of the previous Forest Service risk assessment (SERA 2004b). Furthermore, most of the important mammalian studies associated with potential risks to humans have been reviewed and are summarized in the risk assessment prepared by the Health Effects Division of the U.S. EPA/OPP. Some clinical case reports of intentional (attempted suicide) or accidental ingestion of large amounts of Arsenal are reported in the open literature. The reported signs and symptoms of imazapyr poisoning include vomiting, impaired consciousness, and respiratory distress requiring intubation. There are no reports of human fatality due to imazapyr ingestion.

Although the mode of action regarding the toxicity of imazapyr to humans or other mammals is unclear, this lack of understanding is at least partially a reflection of the apparently low and essentially undetectable acute and chronic systemic toxicity of imazapyr. The acute oral LD50 of unformulated imazapyr is greater than 5000 mg/kg, and the chronic dietary NOAEL for imazapyr is 10,000 ppm in dogs, rats, and mice. In the dog, this dietary concentration is equivalent to a daily dose of 250 mg/kg/day. In the other species, the equivalent daily doses are greater than 250 mg/kg/day. An adequate number of multi-generation reproductive and developmental studies were conducted with imazapyr, none of which indicates adverse effects on reproductive capacity or normal development. Also, the results of assays for carcinogenicity and mutagenicity are consistently negative. Accordingly, U.S. EPA categorizes the carcinogenic potential of imazapyr as Class E: evidence of non-carcinogenicity.

Increased food consumption is reported in chronic toxicity studies in which imazapyr was added to the diets of male and female mice as well as female rats. It is unclear whether this effect is attributed to toxicity or to an increase in the palatability of the chow. The weight of evidence

Storrie and Rich Fire Areas Invasive Plant Treatment Project Appendix D

6

suggests that imazapyr is not directly neurotoxic. Moreover, the available data do not suggest that systemic toxic effects are plausible after dermal or inhalation exposures to imazapyr. Finally, while the available data are limited, there is no basis for asserting that either the metabolites of imazapyr or the impurities or adjuvants in the formulated products are likely to have an impact on the risk assessment.

Imazapyr and imazapyr formulations can be mildly irritating to the eyes and skin. From a practical perspective, irritation to the eyes or skin would most likely to be associated with the application of this compound only if proper personal protection practices are not followed.

EXPOSURE ASSESSMENT For terrestrial applications, worker exposures are modeled for backpack spray, broadcast ground spray, and aerial spray. In non-accidental scenarios involving the normal application of imazapyr, central estimates of exposure for workers are approximately 0.013 mg/kg/day for backpack applications, 0.02 mg/kg/day for ground broadcast applications, and 0.015 mg/kg bw/day for aerial spray. Upper ranges of exposures are approximately 0.08 mg/kg/day for backpack and aerial applications and 0.15 mg/kg/day for ground broadcast applications. All of the accidental exposure scenarios for workers involve dermal exposures. The accidental exposure scenarios lead to dose estimates which are less than those associated with the general exposure levels estimated for workers. This point reflects the limited exposure periods (i.e., 1 minute and 1 hour) used for the accidental exposure scenarios. For terrestrial applications, the upper bound estimate of the absorbed dose is about 0.03 mg/kg bw, if contaminated gloves are worn for 1 hour. If contaminated gloves were worn for an 8-hour workday, the absorbed dose would be about 0.24 mg/kg bw, which is higher than any of the dose estimates for general (non-accidental) exposure scenarios.

For the general public, acute non-accidental exposure levels associated with terrestrial applications range from very low (e.g., ≈9x10-6 mg/kg/day) to 1.35 mg/kg bw at the unit application rate of 1.0 lb a.e./acre. The upper bound of exposure of 1.35 mg/kg bw is associated with the consumption of contaminated vegetation. The other acute exposure scenarios lead to lower and often much lower dose estimates. The lowest acute exposure levels are associated with swimming in or drinking contaminated water. Of the accidental exposure scenarios, the greatest exposure levels are associated with the consumption of contaminated water by a small child, for which the upper bound dose is about 2 mg/kg bw/day. For aquatic applications, the consumption of contaminated terrestrial vegetation is not a relevant route of exposure. The highest non-accidental exposure scenario for aquatic applications is associated with the consumption of contaminated water for which the upper bound of the estimated dose is about 0.04 mg/kg bw/day.

The chronic or longer-term exposure levels are much lower than the estimates of corresponding acute exposures. For terrestrial applications, the highest longer-term exposure levels are associated with the consumption of contaminated vegetation, and the upper bound for this scenario is about 0.6 mg/kg/day, which is followed by the scenario for the longer-term consumption of contaminated fruit with an upper bound of 0.09 mg/kg/day. The lowest longer-term exposure levels are associated with the consumption of contaminated fish.

DOSE-RESPONSE ASSESSMENT The dose-response assessment for imazapyr is relatively straightforward, and the toxicity data base is reasonably complete and unambiguous. The U.S. EPA/OPP derived a chronic RfD of 2.5

Appendix D Mt. Hough Ranger District, Plumas National Forest

7

mg/kg/day using a dog NOAEL of 250 mg/kg/day and an uncertainty factor of 100. The NOAEL selected by the U.S. EPA appears to be the most appropriate and is supported by additional NOAELs in rats and mice as well as a number of studies on potential reproduction and developmental effects. Consistent with the approach taken in U.S. EPA in the most recent human health risk assessment, no acute RfD is derived in the current Forest Service risk assessment and the chronic RfD of 2.5 mg/kg/day is used to characterize the risks of both acute and longer-term exposures. Because doses clearly associated with adverse effects have not been identified and because none of the hazard quotients exceeds the level of concern, considerations of dose-severity relationships cannot be made and are not necessary.

RISK CHARACTERIZATION The quantitative risk characterization in both the human health and in the ecological risk assessment is based on the hazard quotient (HQ), which is defined as the anticipated exposure divided by the toxicity value. For both workers and members of the general public, the chronic RfD of 2.5 mg a.e./kg bw/day is used to characterize risks associated with both acute and longer-term exposures. As discussed in the exposure assessment, all exposure assessments for terrestrial applications are based on the unit application rate of 1.0 lb a.e./acre.

In most Forest Service risk assessments, an HQ of 1 is defined as the level of concern. As discussed in the dose-response assessment, imazapyr is somewhat unusual in that doses of imazapyr which may cause adverse effects have not been determined. Thus, the interpretation of HQs that exceed a value of 1 would be unclear. This is not a practical concern in this risk assessment on imazapyr because none of the HQs exceed a value of 1 at an application rate of 1 lb a.e./acre and no exposures substantially exceed the HQ of 1 at the maximum application rate of 1.5 lb a.e./acre. Consequently, there is no basis for asserting that imazapyr is likely to pose any identifiable risks associated with systemic toxic effects to either workers or members of the general public.

Irritation to the eyes can result from exposure to concentrated solutions of imazapyr. From a practical perspective, eye irritation is likely to be the only overt toxic effect as a consequence of mishandling imazapyr, and these risks are likely to be greatest for workers handling concentrated solutions of imazapyr during cut surface treatments. The potential for eye irritation can be minimized or avoided by prudent industrial hygiene practices, including, exercising care to reduce splashing and wearing goggles, during the handling of the compound.

Aminopyralid (Source: SERA 2007a)

OVERVIEW Aminopyralid is a new herbicide that has been registered by the U.S. EPA for the control of invasive weeds. The control of invasive weeds is a major component in programs conducted by both the USDA/Forest Service and the National Park Service (NPS). Both of these organizations have begun using aminopyralid in weed management programs and both organizations are considering expanding the use of aminopyralid in other weed management programs.

The U.S. EPA has judged that aminopyralid appears to be a reduced risk herbicide. This judgment by the U.S. EPA is supported by the current risk assessment. Aminopyralid is an effective herbicide. As with any effective herbicide applied to terrestrial weeds, adverse effects in

Storrie and Rich Fire Areas Invasive Plant Treatment Project Appendix D

8

nontarget terrestrial plants are plausible. There is no indication, however, that adverse effects on workers, members of the general public or other nontarget animal species are likely.

This assessment of aminopyralid is tempered by the lack of information on aminopyralid in the open literature. All of the information on the toxicity of aminopyralid comes from studies that have been submitted to the U.S. EPA in support of aminopyralid registration. While these studies have been reviewed and the bulk of these studies appear to have been appropriately designed, conducted and reported, the available information on aminopyralid is much less diverse than the information that is available on herbicides that have been used for many years and for which the open literature is rich and varied. This situation will exist for any new herbicide or other new pesticide.

PROGRAM DESCRIPTION Two formulations of aminopyralid are specifically considered in this risk assessment: Milestone and Milestone VM. Both of these formulations contain the triisopropanolamine (TIPA) salt of aminopyralid (40.6 % w a.i./v, equivalent to 21.1% a.e. or 2 lbs a.e./gal). These formulations contain no inert ingredients other than water and triisopropanolamine.

Hazard Identification Because aminopyralid is a new herbicide, no information is available in the published literature on the toxicity of aminopyralid to humans or other mammalian species. The only information on aminopyralid that is available for assessing potential hazards in humans is a series of toxicity studies that have been submitted to and evaluated by the U.S. EPA’s Office of Pesticides in support of the registration for aminopyralid.

Although the mechanism of action of aminopyralid and other pyridine carboxylic acid herbicides is fairly well characterized in plants, the mechanism of action of aminopyralid in mammals is not well characterized. The weight-of-evidence suggests that aminopyralid may not have any remarkable systemic toxic effects. The effects that are most commonly seen involve effects on the gastrointestinal tract after oral exposure and these may be viewed as portal of entry effects rather than systemic toxic effects. The location of these effects within the gastrointestinal tract appears to vary among species with the ceca being the most common site of action in rats and the stomach being the most common site of action in dogs and rabbits. Mice do not seem to display any remarkable gastrointestinal effects after oral doses of aminopyralid. The reason for these differences among species is not clear but may simply reflect differences in methods of exposure (gavage versus dietary) and/or differences in anatomy.

In one acute oral toxicity study in rats using the aminopyralid TIPA formulation, lacrimation and cloudy eyes were noted in all test animals on the first day of the study but not on subsequent days. Clouding of the eyes is an unusual effect that has not been noted in other studies on aminopyralid, either the acid or the TIPA salt. The significance of this observation, if any, is unclear.

Aminopyralid is rapidly absorbed and excreted and is not substantially metabolized in mammals. As a consequence of rapid absorption and excretion, gavage and dietary exposures probably lead to very different patterns in the time-course of distribution in mammals. The oral LD50 of aminopyralid has not been determined because aminopyralid does not cause any mortality at the dose limits set by the U.S. EPA for acute oral toxicity studies – i.e., up to 5,000 mg/kg bw. Similarly, subchronic and chronic toxicity studies have failed to demonstrate any clear signs of

Appendix D Mt. Hough Ranger District, Plumas National Forest

9

systemic toxic effects. Developmental studies involving gavage administration, however, have noted signs of incoordination in adult female rabbits. The incoordination was rapidly reversible and did not persist past the day of dosing. Two chronic oral bioassays have been conducted; one in mice and the other in rats, and a 1-year feeding study is available in dogs. Based on the results of the chronic bioassays as well as the lack of mutagenic activity in several mutagenicity screening assays, there is no basis for asserting that aminopyralid is a carcinogen. Similarly, based on the chronic bioassays and several additional subchronic bioassays in mice, rats, dogs, and rabbits, there is no basis for asserting that aminopyralid will cause adverse effects on the immune system or endocrine function. The potential for effects on the nervous system is less clear. Aminopyralid has also been subject to several bioassays for developmental toxicity and one multi-generation study for reproductive performance. No adverse effects on offspring have been noted in these studies other than decreased body weight in offspring that is associated with decreased food consumption and decreased body weight in adult females.

Exposure Assessment For workers applying aminopyralid, three types of application methods are modeled: directed ground spray, broadcast ground spray, and aerial spray. In non-accidental scenarios involving the normal application of aminopyralid, central estimates of exposure for workers are approximately 0.001 mg/kg/day for aerial and backpack workers and about 0.002 mg/kg/day for broadcast ground spray workers. Upper ranges of exposures are approximately 0.012 mg/kg/day for broadcast ground spray workers and 0.006 mg/kg/day for backpack and aerial workers. All of the accidental exposure scenarios for workers involve dermal exposures. Except for the scenario involving a spill on the lower-legs for 1 hour (an upper bound dose of 0.003 mg/kg/event), the accidental exposures lead to dose estimates that are substantially lower than the general exposure levels estimated for workers. This is not uncommon and it reflects the fact that the general exposure estimates are based on field studies of workers in which accidental and/or incidental events such as spills probably occurred and in some cases were specifically noted to occur.

For the general public, acute levels of exposures range from minuscule (e.g., 1x10-8 mg/kg/day) to about 0.4 mg/kg bw at the typical application rate of 0.078 lb a.e./acre. The upper bound of exposure, 0.4 mg/kg bw, is associated with the consumption of contaminated water by a child shortly after an accidental spill. This exposure scenario is highly arbitrary. The upper bound of the dose associated with the consumption of contaminated vegetation, a more plausible but still extreme exposure scenario, is about 0.1 mg/kg bw. The other acute exposure scenarios lead to much lower dose estimates – i.e., ranging from near zero to about 0.042 mg/kg for the accidental direct spray of a child. The lowest acute exposures are associated with swimming in or drinking contaminated water.

The modeled chronic or longer-term exposures are much lower than the corresponding estimates of acute exposures. The highest longer-term exposures are associated with the consumption of contaminated vegetation and the upper bound for this scenario is about 0.027 mg/kg/day. This is followed by the scenario for the longer-term consumption of contaminated fruit with an upper bound of 0.003 mg/kg/day. As with the acute exposures, the lowest longer-term exposures are associated with the consumption of surface water.

Dose-Response Assessment The Office of Pesticide Programs of the U.S. EPA has derived a chronic RfD of 0.5 mg/kg/day for aminopyralid. This RfD is based on a chronic rat NOAEL of 50 mg/kg/day and an uncertainty

Storrie and Rich Fire Areas Invasive Plant Treatment Project Appendix D

10

factor of 100. The Office of Pesticide Programs has also derived an acute RfD of 1 mg/kg bw/day based on a NOAEL from a reproduction study of about 100 mg/kg/day. In deriving both of these RfD values, the U.S. EPA used an uncertainty factor of 100, a factor of 10 for extrapolating from animals to humans and a factor of 10 for extrapolating to sensitive individuals within the human population. Both of these RfD values are based on NOAELs for the most sensitive endpoint in the most sensitive species and studies in which LOAEL values were identified. In addition, both of the NOAEL values are supported by other studies. Thus, the RfD values recommended by the U.S. EPA are adopted directly in the current risk assessment.

Risk Characterization The risk characterization for both workers and members of the general public is reasonably simple and unambiguous: based on a generally conservative and protective set of assumptions regarding both the toxicity of aminopyralid and potential exposures to aminopyralid, there is no basis for suggesting that adverse effects are likely in either workers or members of the general public even at the maximum application rate that might be used in Forest Service or NPS programs.

For workers, no exposure scenarios, acute or chronic, exceeds the RfD at the upper bound of the estimated dose associated with the highest application rate of 0.11 lb a.e./acre. The hazard quotients for directed ground spray, broadcast ground spray, and aerial applications are below the level of concern by factors of 33 to 200 over the range of application rates considered in this risk assessment.

For members of the general public, upper bounds of hazard quotients at the highest application rate are below a level of concern by factors of 100 to 125,000 for longer term exposures. For one accidental exposure scenario, the consumption of contaminated water by a child immediately after an accidental spill of aminopyralid into a small pond, the hazard quotient is 0.6, approaching the level of concern (1.0). This is an intentionally extreme exposure scenario that typically leads to the highest hazard quotient in pesticide risk assessments similar to the current assessment on aminopyralid. The upper bounds of acute exposure scenarios for contaminated vegetation or fruit are below the level of concern by factors of 10 to 50. Acute non-accidental exposure scenarios for members of the general public that involve contaminated water are below the level of concern by factors of about 140 to 14,000.

The risk characterization given in this risk assessment is qualitatively similar to that given by the U.S. EPA: no risks to workers or members of the general public are anticipated. The current risk assessment derives somewhat higher hazard quotients than those in the U.S. EPA human health risk assessment because the current risk assessment uses a number of extreme exposure scenarios that are not used by the U.S. EPA.

Triclopyr (SERA 2011c)

OVERVIEW The triethylamine salt (TEA) of triclopyr is used in Forest Service programs primarily for conifer or hardwood release, Invasive plant control, site preparation, and rights-of-way management.

Potential risks associated with terrestrial applications are greatest for workers as well as women consuming vegetation contaminated with triclopyr. The central estimates of the HQs indicate that workers will not be subject to hazardous levels of triclopyr during applications of triclopyr

Appendix D Mt. Hough Ranger District, Plumas National Forest

11

TEA at the unit application rate of 1 lb a.e./acre. At the upper bounds of the estimated exposures for all application methods, the HQs for triclopyr TEA (HQs = 1.6 to 3) exceed the level of concern (HQ=1), based on the chronic RfD. For a young woman consuming contaminated vegetation, the upper bound HQ is 27 for acute exposures and 6 for longer-term exposures. In addition, some of the central estimates of exposure to triclopyr or TCP involving a young woman consuming contaminated vegetation or fruit also exceed the level of concern. All of these HQs apply to an application rate of 1 lb a.e./acre and will scale proportionately to the application rate. Because triclopyr has been shown to cause adverse developmental effects in mammals, the high HQs associated with terrestrial applications are of particular concern in terms of the potential for adverse reproductive outcomes in humans. Adverse developmental effects in experimental mammals have been observed, however, only at doses that cause frank signs of maternal toxicity. The available toxicity studies suggest that overt and severe toxicity would not be associated with any of the upper bound HQs and this diminishes concern for reproductive effects in humans.

Qualitatively, the risk characterization for ecological effects is parallel in many respects to the risk characterization for human health effects. At an application rate of 1 lb a.e./acre, HQs exceed the level of concern for exposures involving the consumption of contaminated vegetation by mammals and birds. HQs are greatest for large mammals. As with the human health risk assessment, the high HQs suggest the potential for adverse effects, but not overt toxic effects, in large mammals. Based on a very cursory probabilistic assessment, exposures of mammalian wildlife that would be associated with upper bound HQs are probably rare occurrences.

With the exception of aquatic plants, substantial risks to nontarget species (including humans) associated with the contamination of surface water are low, relative to risks associated with contaminated vegetation.

Program Description Triclopyr is used in Forest Service programs primarily for conifer and/or hardwood release, invasive plant control, site preparation, and rights-of-way management. Two forms of triclopyr are used commercially as herbicides: the triethylamine salt (TEA) and the butoxyethyl ester (BEE). The butoxyethyl ester (BEE) form is not proposed for use in this project. The TEA formulations include several 44.4% liquid formulations and one 14% granular formulation. Several TEA formulations are labeled for aquatic applications.

The most common application method for triclopyr is backpack (selective) foliar applications. Other application methods include ground broadcast foliar application, several non-broadcast application methods (i.e., basal bark, cut stump, and streamline basal bark), and aerial application.

Formulations of triclopyr TEA may be applied at rates of up to 9 lb a.e./acre. While the full range of labeled application rates are considered in this risk assessment, the typical application rate in Forest Service programs is 1 lb a.e./acre and rarely exceeds 6 lb a.e./acre.

Based on Forest Service use statistics for 2004, about 12,500 lbs of triclopyr are used annually in Forest Service programs, and most of this use occurs in the southeastern region of the United States. The use of triclopyr in Forest Service programs represents only about 1% of the agricultural use of triclopyr.

Storrie and Rich Fire Areas Invasive Plant Treatment Project Appendix D

12

HAZARD IDENTIFICATION The toxicity of triclopyr to mammals is relatively well characterized in numerous standard acute, subchronic, and chronic toxicity studies as well as developmental and reproduction studies required by the U.S. EPA/OPP for pesticide registration. In mammals, the toxicity studies that yield the most sensitive endpoints—i.e., the signs of toxicity that occur at the lowest doses—for triclopyr involve developmental and reproductive effects. For both developmental and reproductive effects, however, adverse effects on offspring, most of which are indicative of delayed growth rather than frank abnormalities, occur at doses associated with maternal toxicity.

Based on histopathology and clinical chemistry data from standard acute, subchronic and chronic toxicity studies on triclopyr, the liver and kidneys are the primary target organs. Like most weak acids, triclopyr is excreted primarily in the kidney by an active transport process. At very high doses, this process may become saturated causing triclopyr to reach toxic levels. At sufficiently high doses, triclopyr may cause toxic effects, including death. Nonetheless, triclopyr has a low order of acute lethal potency. There is no information suggesting that triclopyr causes direct adverse effects on the nervous system, endocrine system, or immune function.

Standard bioassays for carcinogenicity were conducted in both rats and mice. In male rats and mice, no statistically significant dose-related trends in tumor incidence were apparent. Based on pair-wise comparisons (i.e., control group vs an exposed group), statistically significant increases were observed for some tumor types, including benign and/or malignant pheochromocytomas combined as well as skin fibromas, in rats but not mice. In female rats and mice, there was a statistically significant dose-related increase in mammary gland adenocarcinomas. The U.S. EPA/OPP reviewed these studies and determined that the evidence for carcinogenicity is marginal and did not recommend a quantitative dose-response assessment for the carcinogenicity of triclopyr. The current risk assessment defers to this decision.

The major metabolite of triclopyr in both mammals and the environment is 3,5,6-trichloro-2- pyridinol, commonly abbreviated as TCP. Although TCP does not have the phytotoxic potency of triclopyr, this compound is toxic to mammals as well as other species. Based on RfDs derived by the U.S. EPA/OPP (Section 3.3), TCP is more toxic than triclopyr to mammals, and as discussed further in the ecological risk assessment, it is also more toxic than triclopyr to aquatic animals. Consequently, exposures to TCP and its toxicity are considered explicitly in the current risk assessment.

EXPOSURE ASSESSMENT For terrestrial applications, all exposure assessments are based on a unit application rate of 1 lb a.e./acre.

For workers involved in terrestrial applications of triclopyr, three types of application methods are modeled: directed foliar (backpack), broadcast ground spray, and aerial spray. The exposure assessment for workers is substantially different from that in the previous Forest Service risk assessment (SERA 2003). In the previous risk assessment, standard worker exposure rates (mg/kg bw per lb/acre) were used. Reservations with this approach were expressed based on the backpack study by Spencer et al. (2000 as referenced in SERA 2011b) ; however, another backpack study by Middendorf (1992a as referenced in SERA 2011b) involving basal stem applications suggested that the standard worker exposure rates for triclopyr were appropriate. Since the 2003 risk assessment, a backpack foliar study conducted by Krieger et al. (2005 as referenced in SERA 2011b), and an earlier backpack foliar study by Middendorf (1992b as

Appendix D Mt. Hough Ranger District, Plumas National Forest

13

referenced in SERA 2011b) were identified. All of these studies involve forestry applications of triclopyr BEE and were sponsored by the Forest Service. Taken together, the studies by Middendorf (1992b as referenced in SERA 2011b), Spencer et al. (2000 as referenced in SERA 2011b), and Krieger et al. (2005 as referenced in SERA 2011b) clearly suggest that workers involved in backpack foliar applications of triclopyr BEE will be subject to substantially greater exposures than would be anticipated based on the standard methods used in most Forest Service risk assessments. Consequently, the study by Middendorf (1992b as referenced in SERA 2011b) is used directly for estimating exposures to workers involved in backpack foliar applications. This study is also used to adjust exposure rates for workers involved in ground boom and aerial applications of triclopyr BEE formulations. There are no detailed exposure studies of workers applying triclopyr TEA formulations in the available literature. While that lack of worker exposure studies involving triclopyr TEA adds uncertainty to this risk assessment, the differences in dermal absorption rates (which are well-documented for triclopyr BEE) suggest that no adjustments for the worker exposure rates for applications of triclopyr TEA are necessary. Consequently, for applications of triclopyr TEA the standard worker exposure rates used in most Forest Service risk assessments are maintained. The differences in exposures estimated for workers involved in applications of triclopyr TEA and triclopyr BEE have a substantial impact on the risk characterization. Triclopyr BEE is not proposed for use in this project.

Under normal circumstances, members of the general public should not be exposed to substantial levels of triclopyr as a result of Forest Service activities. Nonetheless, several highly conservative scenarios are developed for this risk assessment. For terrestrial applications of triclopyr, the greatest exposures are associated with the acute and longer-term consumption of contaminated fruit and vegetation. This is typical of any pesticide exposure following foliar application. Exposures associated with dermal contact and the consumption of water (except for an accidental spill) are considerably lower.

DOSE-RESPONSE ASSESSMENT The U.S. EPA/OPP has derived acute and chronic RfDs for both triclopyr and 3,5,6-trichloro-2-pyridinol (TCP), and these RfDs are adopted without modification. As a general practice in Forest Service risk assessments, the RfDs derived by the U.S. EPA/OPP are used because they generally provide a level of analysis, review, and resources that far exceed those that are or can be conducted in the support of most Forest Service risk assessments. In addition, it is desirable for different agencies and organizations within the federal government to use concordant risk assessment values.

The acute and chronic RfDs for triclopyr are 1 and 0.05 mg/kg bw/day, respectively. Both RfDs are based on NOAELs in rats, and both use an uncertainty factor of 100. The acute RfD is based on a developmental study in which no effects were noted at 100 mg/kg bw/day but severe maternal toxicity was noted at 300 mg/kg bw/day. The chronic RfD is based on a two-generation reproduction study in rats in which no adverse effects were noted at 5 mg/kg bw/day but effects on the kidney were noted at 25 mg/kg bw/day. Because of concerns for the reproductive and developmental toxicity of triclopyr, the acute RfD is not used to assess risks to women of childbearing age. For this group, the chronic RfD is used to assess the risks associated with both acute and longer-term exposures.

Storrie and Rich Fire Areas Invasive Plant Treatment Project Appendix D

14

The acute and chronic RfDs for TCP are lower than those for triclopyr. For TCP, the acute RfD is 0.025 mg/kg bw/day and the chronic RfD is 0.012 mg/kg bw/day. The acute RfD is based on a developmental study in rabbits in which birth defects were noted at a dose of 100 mg/kg bw/day but no adverse effects were observed at 25 mg/kg bw/day. An uncertainty factor of 1000 is applied to the NOAEL to derive the acute RfD. Unlike the case with triclopyr, however, the acute RfD is applied only to women of childbearing age. The chronic RfD is based on a chronic study in dogs in which the NOAEL was 12 mg/kg bw/day. As with the acute RfD, the chronic RfD is derived using an uncertainty factor of 1000.

RISK CHARACTERIZATION The risk characterization for terrestrial applications of triclopyr is more complex and requires greater discussion.

For workers involved in terrestrial applications of triclopyr, the risk characterization is qualitatively similar to the previous Forest Service risk assessment on triclopyr. At the typical application rate of 1 lb a.e./acre, the central estimates of the HQs indicate that workers will not be subject to hazardous levels of triclopyr during applications of triclopyr TEA. At the upper bounds of the estimated exposures for all application methods, the HQs for triclopyr TEA (HQs = 1.6 to 3) exceed the level of concern based on the chronic RfD. Based on the acute RfD, no HQs substantially exceed the level of concern. The HQs based on the acute RfD, however, would only apply to male workers. All HQs for workers will increase linearly with the application rate.

For members of the general public, the only non-accidental exposure scenarios of concern involve the consumption of contaminated fruit or vegetation with consequent exposures to triclopyr and 3,5,6-trichloro-2-pyridinol (TCP), the primary metabolite of triclopyr. At an application rate of 1 lb a.e./acre, the upper bound HQ of 27 for triclopyr in the acute exposure scenario for the consumption of contaminated vegetation by a young woman exceeds the upper bound HQs for occupational exposures. In addition, some of the central estimates of exposure to triclopyr or TCP involving a young woman consuming contaminated vegetation or fruit also exceed the level of concern. Relative to the risks associated with the consumption of contaminated fruit or vegetation, risks associated with other exposure scenarios are marginal.

Because triclopyr has been shown to cause adverse developmental effects in mammals, the high HQs associated with terrestrial applications are of particular concern in terms of the potential for adverse reproductive outcomes in females. Adverse developmental effects in experimental mammals have been observed, however, only at doses that cause frank signs of maternal toxicity. No epidemiology studies or case reports have been encountered that associate human exposures to triclopyr with either frank signs of toxicity or developmental effects. In addition, the available toxicity studies suggest that overt and severe toxicity would not be associated with any of the upper bound HQs. This diminishes concern for reproductive effects in females. Conversely, an epidemiology study on Forest Service personnel conducted by OSHA noted a marginally significant increase in the odds ratios for miscarriages among women in the Forest Service who reported using herbicides. While this analysis does not implicate triclopyr or any other herbicide as a causative agent in miscarriages, the lack of epidemiology studies focused on females of reproductive age with documented exposures to triclopyr adds uncertainty to the risk characterization for terrestrial applications of triclopyr.

Appendix D Mt. Hough Ranger District, Plumas National Forest

15

Glyphosate (2011a) This risk assessment on glyphosate is dominated by three considerations: the extensive literature available on glyphosate, the availability of numerous glyphosate formulations, and the use of surfactants either as components in glyphosate formulations or as adjuvants added to glyphosate formulations prior to application. There are obvious, and in many cases substantial, differences among the toxicities of technical grade glyphosate, glyphosate formulations that do not contain a surfactant, and some glyphosate formulations that contain polyoxyethyleneamine (POEA) surfactants. While the available information does not permit formulation-specific toxicity values, an attempt is made to discriminate between less toxic and more toxic formulations, when possible.

PROGRAMS DESCRIPTION Glyphosate is an herbicide used in Forest Service programs primarily in conifer release, site preparation, and Invasive plant control. The Forest Service identified more than 50 formulations which are explicitly considered in the current risk assessment. This risk assessment, however, is structured to consider any current or future glyphosate formulation registered for applications used in Forest Service programs.

The formulations of glyphosate identified by the Forest Service contain the ammonium, dimethylamine, isopropylamine, or potassium salts of glyphosate. Some formulations contain only one of these salts of glyphosate as an aqueous solution. Other formulations contain surfactants. The product labels for many formulations of glyphosate that do not contain a surfactant indicate that a surfactant must be added to the field solution prior to application. Some formulations that contain a surfactant indicate that other nonionic surfactants may be added to the field solution prior to application. In addition to surfactants, other additives to field solutions of glyphosate include ammonium sulfate, dyes, and drift reducing agents.

The most common application method for glyphosate in Forest Service programs is backpack applied directed foliar sprays. Other application methods used occasionally include broadcast foliar ground applications, cut stem applications, and direct application to emergent aquatic vegetation. Some glyphosate formulations are registered for aerial application. This project does not propose any aerial applications.

Based on the most recent Forest Service use reports, the typical glyphosate application rate is about 2 lb a.e./acre, with most terrestrial applications using rates ranging from 0.5 to 4 lbs a.e./acre. The agricultural use of glyphosate in the United States is greater than Forest Service use by a factor of over 2900. Thus, there is no reason to believe that Forest Service programs will contribute substantially to general concentrations of glyphosate nationally.

HAZARD IDENTIFICATION The toxicity data on technical grade glyphosate are extensive, including both a standard set of toxicity studies submitted to the U.S. EPA/OPP in support of the registration of glyphosate as well as a robust open literature consisting of numerous diverse in vivo and in vitro studies, including some studies in humans. As with any complex collection of studies, the studies on technical grade glyphosate may be subject to differing interpretations. The preponderance of the available data, however, clearly indicates that the mammalian toxicity of glyphosate is low, and very few specific hazards can be identified. Oral doses that exceed around 300 mg/kg bw, glyphosate may cause signs of toxicity, including decreased body weight, changes in certain biochemical parameters in blood as well as tissues, and inhibition of some enzymes (i.e., P450)3 involved in

Storrie and Rich Fire Areas Invasive Plant Treatment Project Appendix D

16

the metabolism of both endogenous and exogenous compounds. At doses from about 1000 to 5000 mg/kg bw, glyphosate can cause death. The most sensitive endpoint for glyphosate—i.e., the adverse effect occurring at the lowest dose—involves developmental effects; accordingly, the EPA-derived RfDs for glyphosate are based on developmental effects. These adverse developmental effects, which consist primarily of delayed development, occur only at doses causing signs of maternal toxicity. There is no indication that glyphosate causes birth defects.

The hazard identification for glyphosate formulations is much less clear. In most Forest Service pesticide risk assessments, the active ingredient is the agent of primary concern, and consideration for ingredients in the formulations is limited to a brief discussion in the section on Adjuvants and Other Ingredients. In the current Forest Service risk assessment, however, the way in which the formulation ingredients other than glyphosate are handled is much different. Many glyphosate formulations include surfactants, and the toxicity of these surfactants is of equal or greater concern to the risk assessment than is the toxicity of technical grade glyphosate. Consequently, as justified by the available data, the hazard identification is subdivided into sections that address the toxicity of glyphosate, the toxicity of glyphosate formulations, and/or the toxicity of the surfactants.

Because surfactants appear to be agents of concern, a central issue in the current Forest Service risk assessment involves differences in surfactants among the glyphosate formulations used by the Forest Service as well as glyphosate formulations for which toxicity data are available in the open literature. The term POEA (an acronym for polyoxyethyleneamine) is commonly used to designate surfactants used in some glyphosate formulations. POEA, however, is not a single surfactant. POEA surfactants are mixtures. Because the constituents in the surfactants are considered proprietary (trade secrets or Confidential Business Information), detailed information about the constituents is not publically available. The POEA surfactant used in one glyphosate formulation may be different from the POEA surfactant used in other glyphosate formulations, even among formulations provided by the same manufacturer. Thus, it is not clear whether the toxicity studies conducted on POEA surfactants are applicable to all or any of the glyphosate formulations currently in use.

The difference or potential difference in the composition of surfactants used in various formulations of glyphosate has a practical impact on the hazard identification for the current Forest Service risk assessment. Several studies conducted outside of the United States on glyphosate formulations which are not used domestically report adverse effects of concern, including potential effects on endocrine function in rats and signs of genotoxicity in humans. In the absence of comparable studies on glyphosate formulations manufactured and used in the United States, the extent to which this information is relevant to U.S. formulations of glyphosate is unclear.

EXPOSURE ASSESSMENT All exposure assessments are based on unit application rate of 1 lb a.e./acre. The consequences of varying this application rate are considered in the risk characterization.

For workers applying glyphosate, three types of application methods are modeled: directed foliar (backpack), broadcast ground spray, and aerial spray. In non-accidental scenarios involving the normal application of glyphosate, central estimates of exposure for workers are approximately 0.015 mg/kg bw/day for aerial, 0.022 mg/kg bw/day for ground broadcast, and 0.013 mg/kg bw/day for directed foliar applications. Upper ranges of exposures are approximately 0.08 mg/kg bw/day for aerial, 0.15 mg/kg bw/day for ground broadcast, and 0.08 mg/kg bw/day for directed

Appendix D Mt. Hough Ranger District, Plumas National Forest

17

foliar applications. All of the accidental exposure scenarios for workers involve dermal exposures. Because glyphosate is not readily absorbed by the dermal route and because the accidental dermal exposure scenarios involve relatively brief periods of time, the estimated doses are much lower than those associated with general exposures over the course of a workday.

For the general public, acute levels of exposures range from minuscule e.g., 1x10-10 mg/kg/day, the lower bound for swimming in contaminated water, to about 2 mg/kg bw. The upper bound of exposure, 2 mg/kg bw, is associated with the consumption of contaminated water by a child shortly after an accidental spill. This exposure scenario is highly arbitrary. The upper bound of the dose associated with the consumption of contaminated vegetation, a more plausible but still extreme exposure scenario, is about 1.4 mg/kg bw. The other acute exposure scenarios lead to much lower dose estimates.

The chronic or longer-term exposure levels are much lower than the estimates of corresponding acute exposures. The highest longer-term exposure levels are associated with the consumption of contaminated vegetation, and the upper bound for this scenario is about 0.2 mg/kg/day, which is followed by the scenario for the longer-term consumption of contaminated fruit with an upper bound of 0.03 mg/kg/day. As with the acute exposures, the lowest longer-term exposures are associated with the consumption of surface water.

DOSE RESPONSE ASSESSMENT The current Forest Service risk assessment adopts the RfD of 2 mg/kg bw/day which is based on a NOAEL of 175 mg/kg bw/day from a developmental study in rabbits (U.S. EPA/OPP 1993a,c, 2000 as referenced in SERA 2011a). Relative to other similar criteria which are available from the U.S. EPA and WHO, the RfD derived by U.S. EPA/OPP (1993a,c, 2000 as referenced in SERA 2011a) is preferable because it is based on a study that defines both a NOAEL and a LOAEL. The other available exposure criteria are based on free standing NOAELs—i.e., studies that do not define an adverse effect level. Using an RfD derived by the EPA is standard practice in most Forest Service risk assessments.

The U.S. EPA RfDs are used because they generally provide a level of analysis, review, and resources that far exceed those that are or can be conducted in the support of most Forest Service risk assessments. In addition, it is desirable for different agencies and organizations within the federal government to use concordant risk assessment values.

RISK CHARACTERIZATION The quantitative risk characterization is expressed as the hazard quotient (HQ). For both general and accidental exposures, the HQ is calculated as the estimated doses in units of mg/kg bw for acute exposures or units of mg/kg bw/day for longer-term exposures divided by the RfD of 2 mg/kg/day (U.S. EPA/OPP 1993a,b as referenced in SERA 2011a). The RfD is derived from a developmental study and applied to both acute and longer-term exposures.

For both workers and members of the general public, the RfD of 2 mg a.e./kg bw/day is used to characterize risks associated with acute and longer-term exposure levels. All exposure assessments are based on the unit application rate of 1 lb a.e./acre. Because the HQs are based on the RfD, an HQ of 1 or less suggests that exposures are below the level of concern. HQs greater than 1 indicate that the exposure exceeds the level of concern.

Storrie and Rich Fire Areas Invasive Plant Treatment Project Appendix D

18

Based on the HQ method, concern for workers is minimal. At the highest labeled application rate for terrestrial applications, about 8 lbs a.e./acre, the highest HQ is 0.6, the upper bound of the HQ for workers involved in ground broadcast applications.

For members of the general public, the only non-accidental exposure scenario of concern is for acute exposure involving the consumption of contaminated vegetation shortly after glyphosate is applied. For this exposure scenario, the HQ reaches a level of concern (HQ=1) at an application rate of about 1.4 lbs a.e./acre. At the maximum labeled application rate of about 8 lbs a.e./acre, the resulting HQ value would be about 5.6 with a corresponding dose of about 10.8 mg/kg bw.

Apart from the standard HQ method, there are additional concerns, including a report of systemic toxicity in California workers involved in glyphosate applications. In addition, two studies indicate a potential for chromosomal damage in South American populations exposed to glyphosate formulated with surfactants from aerial sprays at application rates in the range of those used in Forest Service programs. While these studies are not used quantitatively in the current Forest Service risk assessment, they suggest a potential for health effects that are not identified or confirmed using the standard HQ method.

It has been USFS practice to defer to the US Environmental Protection Agency (US EPA) unless there is a compelling reason to do otherwise. At this point, there is not yet a compelling reason to adopt the IARC’s classification since all the technical details are not yet available from IARC and since US EPA’s and our analyses would indicate a different conclusion. As stated, a new risk assessment from US EPA is expected later this year which will undoubtedly consider the IARC’s classification. If the US EPA accepts the IARC recommendation, then the USFS would consider an update to the glyphosate RA and for purposes of existing NEPA documents, such a reclassification would be considered ‘new information’ (Bakke 2015).

Chlorsulfuron (SERA 2004) OVERVIEW

Adverse effects on some nontarget plant species, both terrestrial and aquatic, are plausible unless measures are taken to limit exposure. For terrestrial plants, the dominant factor in the risk characterization is the potency of chlorsulfuron relative to the application rate – i.e., the typical application rate considered in this risk assessment is over 6000 times higher than the noobserved effect concentration (NOEC) in direct spray assays of the most sensitive nontarget species. The highest application rate that may be considered in Forest Service programs is over 25,000 times the NOEC and about a factor of 2 above the NOEC in tolerant species. Given these relationships, damage to nontarget plant species after ground broadcast applications could extend to distances of greater than 900 feet from the application site. This risk characterization applies only to ground broadcast applications. When used in directed foliar applications (i.e., backpack), offsite drift could be reduced substantially but the extent of this reduction cannot be quantified.

Damage to aquatic plants, particularly macrophytes, is likely to be less substantial but still noteworthy. At the typical application rate, peak concentrations of chlorsulfuron in water could result in damage to aquatic macrophytes – i.e., hazard quotients ranging from 1.2 to about 24 based on an EC50 for growth inhibition. Thus, if chlorsulfuron is applied in areas where transport to water containing aquatic macrophytes is likely, it would be plausible that detectable damage could be observed. Aquatic algae do not appear to be as sensitive to chlorsulfuron and the

Appendix D Mt. Hough Ranger District, Plumas National Forest

19

hazard quotient is only modestly above the level of concern based on an acute NOEC. Thus, it is not clear if any substantial damage would be likely in aquatic algae.

Just as there is little reason to doubt that adverse effects on some plant species are plausible, there is no clear basis for suggesting that effects on humans or wildlife (terrestrial or aquatic animals) are likely or would be substantial. In workers involved in ground broadcast applications, the level of concern is modestly exceeded at the highest application rate (0.25 lb/acre) and the level of concern is reached at an application rate of 0.14 lb/acre.

PROGRAM DESCRIPTION Chlorsulfuron is recommended for preemergent and early postemergent control of many annual, biennial, and perennial broadleaf weeds. Three formulations of chlorsulfuron are available in the United States: Telar® DF and Glean®, which are produced by Dupont, and Corsair ™, which is produced by Riverdale. Chlorsulfuron is formulated as a dry flowable granule that is mixed with water and applied as a spray. All three formulations contain 75% (w/w) chlorsulfuron and 25% (w/w) inert ingredients. Telar DF and Corsair are labeled for non-crop, industrial use and Glean is labeled for agricultural use. None of the formulations are specifically registered for forestry use.

Chlorsulfuron is used in Forest Service programs only for the control of invasive plants. The most common methods of ground application for chlorsulfuron involve backpack (selective foliar) and boom spray (broadcast foliar) operations. The Forest Service does not use aerial applications for chlorsulfuron. For this risk assessment, the typical rate of 0.056 lbs/acre is used, with a range of 0.0059 to 0.25 lbs/acre. This range is based on lowest and highest labeled application rates recommended on the manufacturer’s label.

The Forest Service used approximately 33 lbs of chlorsulfuron in 2002, the most recent year for which use statistics are available.

Hazard Identification In experimental mammals, the acute oral LD50 for chlorsulfuron is greater than 5000 mg/kg, which indicates a low order of toxicity. Acute exposure studies of chlorsulfuron and chlorsulfuron formulations give similar results, indicating that formulations of chlorsulfuron are not more toxic than chlorsulfuron alone. The most common signs of acute, subchronic, and chronic toxicity are weight loss and decreased body weight gain. The only other commonly noted effects are changes in various hematological parameters and general gross pathological changes to several organs. None of these changes, however, suggest a clear or specific target organ toxicity. Appropriate tests have provided no evidence that chlorsulfuron presents any reproductive risks or causes malformations or cancer. Results of all mutagenicity tests on chlorsulfuron are negative. The inhalation toxicity of chlorsulfuron is not well documented in the literature. Results of a single acute inhalation study indicate that chlorsulfuron produces local irritant effects. Chlorsulfuron is mildly irritating to the eyes and skin, but does not produce sensitizing effects following repeated dermal exposure.

Limited information is available on the toxicokinetics of chlorsulfuron. The kinetics of absorption of chlorsulfuron following dermal, oral or inhalation exposure are not documented in the available literature. Chlorsulfuron does not appear to concentrate or be retained in tissues following either single or multiple dose administration. Chlorsulfuron exhibits first order elimination kinetics, with an estimated half-life in rats of < 6 hours. In all mammalian species studied, chlorsulfuron and its metabolites are extensively and rapidly cleared by a combination

Storrie and Rich Fire Areas Invasive Plant Treatment Project Appendix D

20

of excretion and metabolism. The primary excretory compartment for chlorsulfuron and its metabolites is the urine, with smaller amounts excreted in the feces. Most of the chlorsulfuron excreted in urine is in the form of the parent compound. Studies on the toxicity of chlorsulfuron metabolites have not been conducted.

As discussed in the exposure assessment, skin absorption is the primary route of exposure for workers. Data regarding the dermal absorption kinetics of chlorsulfuron are not available in the published or unpublished literature. For this risk assessment, estimates of dermal absorption rates – both zero order and first order – are based on quantitative structure-activity relationships.

These estimates of dermal absorption rates are used in turn to estimate the amounts of chlorsulfuron that might be absorbed by workers, which then are used with the available doseresponse data to characterize risk. The lack of experimental data regarding dermal absorption of chlorsulfuron adds substantial uncertainties to this risk assessment. Uncertainties in the rates of dermal absorption, although they are substantial, can be estimated quantitatively and are incorporated in the human health exposure assessment.

Exposure Assessment Exposure assessments are conducted for both workers and members of the general public for the typical application rate of 0.056 lb/acre. The consequences of using the maximum application rate that might be used by the Forest Service, 0.25 lb/acre, are discussed in the risk characterization.

For workers, three types of application methods are generally modeled in Forest Service risk assessments: directed ground, broadcast ground, and aerial. Neither Telar nor Corsair, however, are registered for aerial application and estimates of exposures for workers involved in aerial application are not used in the risk characterization. Central estimates of exposure for ground workers are approximately 0.0007 mg/kg/day for directed ground spray and 0.001 mg/kg/day for broadcast ground spray. Upper range of exposures are approximately 0.0045 mg/kg/day for directed ground spray and 0.0085 mg/kg/day for broadcast ground spray. All of the accidental exposure scenarios for workers involve dermal exposures and all of these accidental exposures lead to estimates of dose that are either in the range of or substantially below the general exposure estimates for workers.

For the general public, the range of acute exposures is from approximately 0.0000002 mg/kg associated with the lower range for dermal exposure from an accidental spray on the lower legs to 0.09 mg/kg associated with the upper range for consumption of contaminated water by a child following an accidental spill of chlorsulfuron into a small pond. High dose estimates are also associated with consumption of contaminated fruit (approximately 0.01 mg/kg) and fish (approximately 0.008 mg/kg for subsistence populations). For chronic or longer term exposures, the modeled exposures are much lower than for acute exposures, ranging from approximately 0.000000001 mg/kg/day associated with the lower range for the normal consumption of fish to approximately 0.004 mg/kg/day associated with the upper range for consumption of contaminated fruit.

Dose-Response Assessment The Office of Pesticide Programs of the U.S. EPA has derived a chronic RfD of 0.05 mg/kg/day for chlorsulfuron. This RfD is based on a chronic rat NOAEL of 5 mg/kg/day (Wood et al. 1980b as

Appendix D Mt. Hough Ranger District, Plumas National Forest

21

referenced in SERA 2004) and an uncertainty factor of 100. In the same study, the LOAEL was 25 mg/kg/day and the effect noted was a weight loss and decreased weight gain. No frank signs of toxicity were seen at this or higher dose levels. This NOAEL for chronic toxic effects is below the NOAEL of 25 mg/kg/day for reproductive effects (Wood et al. 1981a as referenced in SERA 2004). Thus, doses at or below the RfD will be below the level of concern for reproductive effects. For acute/incidental exposures, the U.S. EPA uses an acute NOAEL of 75 mg/kg/day with an uncertainty factor of 300 resulting in an acute RfD of 0.25 mg/kg/day [75 mg/kg/day / 300]. Both of these values are used in the current risk assessment for characterizing risks associated with exposures to chlorsulfuron.

Risk Characterization For both workers and members of the general public, typical exposures to chlorsulfuron do not lead to estimated doses that exceed a level of concern. For workers, the upper range of hazard quotients is below the level of concern for backpack and aerial applications but somewhat above the level of concern for ground broadcast applications at the highest application rate. For ground broadcast applications, the level of concern is reached at an application rate of 0.14 lb/acre. For members of the general public, the upper limits for hazard quotients are below a level of concern except for the accidental spill of a large amount of chlorsulfuron into a very small pond. Even this exposure scenario results in only a small excursion above the acute RfD and is not likely to be toxicologically significant, because of the short duration of exposure relative to those considered in the derivation of the RfD.

Mild irritation to the skin and eyes can result from exposure to relatively high levels of chlorsulfuron. From a practical perspective, eye or skin irritation is likely to be the only overt effect as a consequence of mishandling chlorsulfuron. These effects can be minimized or avoided by prudent industrial hygiene practices during the handling of the compound.

Cumulative Effects Cumulative effects from the proposed herbicides may result from (a) repeated exposure to one particular chemical or (b) simultaneous exposure to a particular chemical and other agents that may cause the same effect or effects by the same or similar modes of action.

In terms of repeated exposure to one particular chemical, the analysis of chronic exposure scenarios discussed in this risk analysis specifically addresses the potential long-term cumulative impacts associated with aminopyralid, imazapyr, glyphosate, chlorsulfuron, and triclopyr. This risk assessment determined that there is a low likelihood of cumulative adverse effects associated with long-term or repeated exposures to the proposed chemicals.

Since these herbicides persist in the environment for a relatively short time (generally less than one year), do not bio-accumulate, and are rapidly eliminated from the body, additive doses from re-treatments in subsequent years are not anticipated. Based on the re-treatment schedule proposed, it is possible that residues from the initial herbicide application could still be detectable during subsequent re-treatments the following year, but these plants would represent a low risk to humans as they would show obvious signs of herbicide effects and would be undesirable for collection.

It is conceivable that workers or members of the public could be exposed to herbicides as a result of treatments on surrounding public or private forestlands. Where individuals could be

Storrie and Rich Fire Areas Invasive Plant Treatment Project Appendix D

22

exposed by more than one route, the risk of such cases can be quantitatively characterized by simply adding the hazard quotients for each exposure scenario.

Additional sources of pesticide exposure include use of herbicides and fungicides on adjacent private timberlands or home use by a worker or member of the general public.

The U.S. EPA has developed the theoretical maximum residue contribution (TMRC), which can be used to consider the cumulative effects associated with use of these herbicides outside of the Storrie IPT Project. The TMRC is an estimate of maximum daily exposure to chemical residues that a member of the general public could be exposed to from all published and pending uses of a pesticide on a food crop. Adding the TMRC to this project’s chronic dose estimates can be used as an estimate of the cumulative effects of this project with theoretical background exposure levels of these herbicides. The result of doing this doesn’t change the risk conclusions based on the project-related HQ values.

Cumulative effects can also be caused by the interaction of different chemicals with a common metabolite or a common toxic action; however, the herbicides this analysis has been demonstrated to share a common metabolite.

Inert Ingredients The approach used in USDA (1989, as referenced in USDA 2008), the SERA Risk Assessments (SERA 2004, 2007a, 2011a, 2011b, 2011c), and this analysis to assess the human health effects of inert ingredients and full formulations has been to: (1) compare acute toxicity data between the formulated products (including inert ingredients) and their active ingredients alone; (2) disclose whether or not the formulated products have undergone chronic toxicity testing; and (3) identify, with the help of EPA and the chemical companies, ingredients of known toxicological concern in the formulated products and assess the risks of those ingredients.

Researchers have studied the relationships between acute and chronic toxicity and while the biological end-points are different, relationships do exist and acute toxicity data can be used to give an indication of overall toxicity (Zeise, et al. 1984, as referenced in USDA 2008). The court in NCAP v. Lyng, 844 F.2d 598 (9th Cir 1988) decided that this method of analysis provided sufficient information for a decision maker to make a reasoned decision. In SRCC v. Robertson, Civ.No. S-91-217 (E.D. Cal., June 12, 1992) and again in CATs v. Dombeck, Civ. S-00-2016 (E.D. Cal., Aug 31, 2001) the district court upheld the adequacy of the methodology used in USDA (1989, as referenced in USDA 2008) for disclosure of inert ingredients and additives.

Since most information about inert ingredients is classified as “Confidential Business Information” (CBI) the Forest Service asked EPA to review the thirteen herbicides for the preparation of USDA 1989 (includes glyphosate) and the commercial formulations and advise if they contained inert ingredients of toxicological concern (Inerts List 1 or 2) (USDA 1989, as referenced in USDA 2008). The EPA determined that there were no inerts on List 1 or 2. In addition, the CBI files were reviewed in the development of the most recent SERA risk assessments (SERA 2007a, 2011a, 2011b). Information has also been received from the companies who produce the herbicides and spray additives.

Comparison of acute toxicity (LD50 values) data between the formulated products (including inert ingredients) and their active ingredients alone shows that the formulated products are generally less toxic than their active ingredients (SERA 2007a, 2011a, 2011b, USDA 1989, as referenced in USDA 2008).

Appendix D Mt. Hough Ranger District, Plumas National Forest

23

The sole inert ingredient listed for the formulations of aminopyralid, most likely to be used in the Storrie IPT Project is water (SERA 2007a). The Northwest Coalition for Alternatives to Pesticides (NCAP) obtained information on the identity of the inerts in Arsenal AC (a formulation of imazapyr) from U.S. EPA, under the Freedom of Information Act. This listing is no longer posted on the NCAP web site; however, the information was reviewed in the SERA (2004a) risk assessment. The only inert other than water listed at NCAP site was glacial acetic acid (CAS No. 64-19-7). Dilute acetic acid is an approved food additive. Acetic acid is a major component of vinegar and is a U.S. EPA List 4B inert. The triclopyr formulations used by the Forest Service contain two inerts which are classified as toxic, ethanol (Garlon 3A) and kerosene (Garlon 4). Garlon 4 is not proposed for use in this project. Both of these agents are neurotoxic. The toxicity of ethanol, which is used in formulations such as Garlon 3A, is extremely well characterized in humans, and the hazards of exposure include intoxication from acute exposure as well as liver cirrhosis and fetal alcohol syndrome. For chronic exposure, the alcohol contained in Garlon 3A will not be of toxicological significance because of the rapid breakdown of alcohol in the environment and the relatively high levels of alcohol associated with chronic alcohol poisoning. Similarly, alcohol is not likely to pose an acute toxic hazard. Approximately 15 mL of alcohol is contained in 1 oz of an alcoholic beverage containing 50% alcohol (100 proof) [0.5 ⋅ 1 oz ⋅ 29.6 mL/oz ≃ 14.8 mL]. This level may cause mild intoxication in sensitive individuals. Each mL of Garlon 3A contains 0.01 mL of ethanol. Therefore, 1480 mL, or approximately 1.5 L, of Garlon 3A must be consumed to equal the amount of alcohol contained in 1 oz of an alcoholic beverage. The same amount of Garlon 3A contains 540,000 mg a.e. of triclopyr [1.5 L ⋅ 360,000 mg a.e./L]. For a 70 kg man, this dose would equal approximately 770 mg a.e./kg, which is similar to the LD50 for rats. As discussed in the dose-response section, this estimate may be a reasonable approximation of a lethal dose for triclopyr in humans. Thus, compared with the active ingredient, which is triclopyr, the amount of ethanol in Garlon 3A does not appear to be toxicologically significant in terms of potential systemic toxicity. Nonetheless, ethanol is an effective solvent. Some formulations of triclopyr TEA have been associated with severe eye irritation. While somewhat speculative, these irritant effects could be due, at least in part, to ethanol.

Table: Disclosed Inerts in Triclopyr Formulations

Formulation Name [1] % a.i. Inert (CAS No. if specified) Amount

TEA Salt

Garlon 3A 44.4% Ethanol (64-17-5)

NOS

2.1%

50.5%

Renovate 3 44.4% Ethanol (64-17-5)

NOS

2.1%

50.5%

Renovate OTF (granular) 10 to 30% Proprietary Fiber

Proprietary Clay

Proprietary Salt

Titanium dioxide (13463-67-7)

30 to 60%

5 to10%

5 to10%

0.1 to 1%

Tahoe 3A 44.4% Other (including ethanol) 55.6%

Triclopyr 3A 44.4% Ethylenediaminetetraacetic acid

[EDTA] (64-02-8)[2]

Storrie and Rich Fire Areas Invasive Plant Treatment Project Appendix D

24

While these formulated products have not undergone chronic toxicity testing like their active ingredients, the acute toxicity comparisons, the EPA review, and our examination of toxicity information on the inert ingredients in each product leads us to conclude that the inert ingredients in these formulations do not significantly increase the risk to human health and safety over the risks identified for the active ingredients.

Additives Additives (also known as adjuvants) are mixed with an herbicide solution to improve the performance of the spray mixture by either enhancing the activity of the herbicide’s active ingredient or by offsetting problems associated with application, such as water or wind factors (Bakke 2007). The two additives proposed for use in the Storrie IPT Project are: an esterified vegetable oil surfactant (e.g., Competitor® or an equivalent formulation) to facilitate and enhance the spreading and penetrating properties of the herbicides and a marker dye (e.g., Hi-light® Blue or an equivalent formulation) to allow for the identification of plants that have been treated.

Additives are not under the same registration guidelines as are pesticides; therefore much of the information that describes the active ingredients in additives is considered confidential business information (CBI). The EPA does not register or approve the labeling of spray additives, although the California Department of Pesticide Regulation (DPR) does require the registration of those that are considered to increase the action of the pesticide it is used with. All additives are generally field tested by the manufacturer in combination with several different herbicides and weed species, and under a number of different environmental conditions (Bakke 2007).

The most common risk factor associated with the use of the proposed additives is skin or eye exposure. This risk can be minimized through good industrial hygiene practices (i.e. personal protective eyewear and gloves) while utilizing these products. Overall, the additives proposed for use within the Storrie IPT Project are not expected to pose an adverse risk to the health and safety of workers or members of the general public. This is based on information provided on the product labels as well as in the discussion contained in Bakke (2007) in which the two additives proposed for use under this project are discussed and some acute toxicity data presented. The following provides further discussion of the additives analyzed for the Storrie IPT Project.

Competitor® (or an Equivalent Formulation) Product labels contain “signal words” (caution, warning, danger, and poison) which indicate the product’s relative toxicity to humans. The signal word is assigned using a combination of acute toxicity studies and the toxicity of each of the product’s components (Tu et al. 2001). Competitor® has been assigned a “caution” signal word and the label indicates that improper use may cause irritation to the skin and eyes.

The main ingredient in Competitor® is an esterified vegetable oil. It also contains two emulsifiers, sorbitan alkylpolyethoxylate ester and dialkyl polyethoxylene glycol. Vegetable oil surfactants are gaining in popularity due to their capability to increase herbicide absorption and spray retention (Bakke 2007). The U.S. Food and Drug Administration (FDA) considers methyl and ethyl esters of fatty acids produced from edible fats and oils to be food grade additives (21 CFR 172.225). However, because of the lack of exact ingredient statements on these surfactants, it is not always clear whether the oils used meet the U.S. FDA standard.

Appendix D Mt. Hough Ranger District, Plumas National Forest

25

Hi-light® Blue (or an Equivalent Formulation). Hi-Light® Blue dye is not required to be registered as a pesticide; therefore there is no signal word included on the label. However, according to Bakke (2007), this product would likely have a “caution” signal word if required to identify one. The label does indicate that this product is mildly irritating to the skin and eyes. Hi-Light® Blue is commonly used in toilet bowl cleaners and as a colorant for lakes and ponds (SERA 1997). This dye is water-soluble, contains no listed hazardous substances, and is considered virtually non-toxic to humans (SERA 1997, Bakke 2007) .The effect of use on non-target terrestrial and aquatic species is unknown; however the use of this dye use has not resulted in any known problems (Bakke 2007).

The use of Hi-Light® Blue in the proposed herbicide formulations would result in almost no increased risk to the health and safety of the workers or members of the general public. In fact, the use of dye in herbicide application can reduce likelihood and risk of exposure by facilitating avoidance of treated vegetation.

Synergistic Effects Synergistic effects are those effects resulting from exposure to a combination of two or more chemicals that are greater than the sum of the effects of each chemical alone (additive). Refer to USDA (1989, as referenced in USDA 2003) for a detailed discussion on synergistic effects.

It is not anticipated that synergistic effects would be seen with the additives proposed in the Storrie IPT Project. Based on a review of several recent studies, there is no demonstrated synergistic relationship between herbicides and surfactants (Abdelghani et al 1997; Henry et al 1994; Lewis 1992; Oakes and Pollak 1999, 2000 as referenced in Bakke 2007).

Although the combination of surfactant and herbicide might indicate an increased rate of absorption through the skin, a review of recent studies indicates this is not often true (Ashton et al 1986; Boman et al 1989; Chowan and Pritchard 1978; Dalvi and Zatz 1981; Eagle et al 1992; Sarpotdar and Zatz 1986; Walters et al 1993, 1998; Whitworth and Carter 1969 as referenced in Bakke 2007). For a surfactant to increase the absorption of another compound, the surfactant must affect the upper layer of the skin. Without some physical effect to the skin, there will be no change in absorption as compared to the other compound alone. The studies indicate that in general non-ionic surfactants have less of an effect on the skin, and hence absorption, then anionic or cationic surfactants. Compound specific studies indicate that the alkylphenol ethoxylates generally have little or no effect on absorption of other compounds. In several studies, the addition of a surfactant actually decreased the absorption through the skin. It would appear that there is little support for the contention that the addition of surfactants to herbicide mixtures would increase the absorption through the skin of these h