Pesticides on Tobacco: Federal Activities to Assess Risks and	 
Monitor Residues (26-MAR-03, GAO-03-485).			 
                                                                 
Pesticides play a significant role in increasing production of	 
tobacco, food, and other crops by reducing the number of	 
crop-destroying pests. However, if used improperly, pesticides	 
can have significant adverse health effects. GAO was asked to (1)
identify the pesticides commonly used on tobacco crops and the	 
potential health risks associated with them, (2) determine how	 
the Environmental Protection Agency (EPA) assesses and mitigates 
health risks associated with pesticides used on tobacco, and (3) 
assess the extent to which federal agencies regulate and test for
pesticide residues on tobacco.					 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-03-485 					        
    ACCNO:   A06439						        
  TITLE:     Pesticides on Tobacco: Federal Activities to Assess Risks
and Monitor Residues						 
     DATE:   03/26/2003 
  SUBJECT:   Contaminated foods 				 
	     Food inspection					 
	     Health hazards					 
	     Pesticides 					 
	     Tobacco industry					 
	     Toxic substances					 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-03-485

Report to the Ranking Minority Member, Committee on Government Reform,
House of Representatives

United States General Accounting Office

GAO

March 2003 PESTICIDES ON TOBACCO

Federal Activities to Assess Risks and Monitor Residues

GAO- 03- 485

In the 1990s, domestic growers commonly used 37 pesticides approved for
use on tobacco by EPA. Most of these pesticides were also used on food
crops. When used in ways that deviate from conditions set by EPA, many of
these pesticides can cause moderate to severe respiratory and neurological
damage* and may result in death. Moreover, animal studies suggest that
some of these pesticides may cause birth defects or cancer.

Under its pesticide registration program, EPA evaluates toxicity and other
data to assess health risks to workers and the public from exposure to
pesticides* and risks to smokers from exposure to residues in smoke. These
assessments have identified a range of risks that required such mitigation
as limiting where and how the pesticide may be used, prohibiting use in
certain states, and requiring workers to wear respirators and chemical-
resistant clothing. On the other hand, EPA has concluded that low levels
of residues in tobacco smoke do not pose short- term health concerns
requiring mitigation. EPA does not assess intermediate or long- term risks
to smokers because of the severity of health effects linked to use of
tobacco products themselves.

While EPA regulates the specific pesticides that may be used on tobacco
and other crops and specifies how the pesticides may be used, it does not
otherwise regulate residues of pesticides approved for use on tobacco.
USDA, however, is required by the Dairy and Tobacco Adjustment Act to test
imported and domestic tobacco for residues of pesticides not approved by
EPA for use on tobacco that federal officials believe are

used in other countries. By helping ensure that other countries do not use
highly toxic pesticides that U. S. tobacco growers may not use, federal
regulation of pesticide residues on tobacco addresses trade equity as well
as health and environmental issues. However, USDA has not reevaluated the
list of pesticides for which it tests since 1989, even though EPA has
cancelled tobacco use for over 30 pesticides since then.

USDA Inspectors Take a Tobacco Sample for Laboratory Testing

Pesticides play a significant role in increasing production of tobacco,
food, and other crops by reducing the number of crop- destroying

pests. However, if used improperly, pesticides can have significant
adverse health effects. GAO was asked to (1) identify the pesticides
commonly used on

tobacco crops and the potential health risks associated with them, (2)
determine how the Environmental Protection Agency (EPA) assesses and
mitigates

health risks associated with pesticides used on tobacco, and (3) assess
the extent to which federal agencies regulate and test for pesticide
residues on tobacco.

GAO recommends that the Secretary of the Department of Agriculture direct
the Administrators of the Agricultural

Marketing Service and the Farm Service Agency to periodically review and
update the pesticides on tobacco for which they set residue limits and
test imported and domestic tobacco. Commenting on a draft of this

report, EPA officials said GAO accurately characterized the agency*s risk
assessment process for pesticides used on tobacco, and Department of
Agriculture officials agreed with GAO*s recommendation to periodically
review and update the pesticides for which the department sets residue
limits and tests tobacco.

www. gao. gov/ cgi- bin/ getrpt? GAO- 03- 485. To view the full report,
including the scope and methodology, click on the link above. For more
information, contact John B. Stephenson at (202) 512- 3841 or stephensonj@
gao. gov. Highlights of GAO- 03- 485, a report to the

Ranking Minority Member, Committee on Government Reform, House of
Representatives

March 2003

PESTICIDES ON TOBACCO

Federal Activities to Assess Risks and Monitor Residues

Page i GAO- 03- 485 Pesticides on Tobacco Letter 1 Results in Brief 2
Background 4 Pesticides Commonly Used on Tobacco Have Potential Short- and

Long- Term Adverse Health Effects 7 EPA Concludes that Health Risks of
Pesticide Residues on Tobacco Are Minimal but Requires Mitigation for
Risks from Other Exposures 14 Federal Regulation of Pesticide Residues on
Tobacco Is Limited 26 Conclusions 32 Recommendation for Executive Action
33 Agency Comments 33 Appendix I Objectives, Scope, and Methodology 35

Appendix II Pesticide Use on Tobacco and Other Crops 37

Appendix III Germany, Italy, and Spain Have Adopted Regulatory Limits for
Pesticide Residues on Tobacco 39

Appendix IV USDA Tests Domestic Tobacco in the Loan Stock Program 42

Appendix V GAO Contacts and Staff Acknowledgments 44

Tables

Table 1: Pesticides Commonly Used on Domestic Tobacco, 1990- 98 8 Table 2:
Pesticide Use on Tobacco, 1990- 98 9 Table 3: Organochlorine,
Organophosphate, and Carbamate Pesticides Commonly Used on Tobacco in the
1990s 11 Table 4: Margins of Exposure for Five Pesticides Approved for Use
on Tobacco 26 Table 5: USDA*s Residue Limits for Pesticides on Tobacco 29
Contents

Page ii GAO- 03- 485 Pesticides on Tobacco

Table 6: Pesticide Use on Tobacco and All Crops, 1990- 98 37 Table 7:
Residue Limits Adopted by Germany, Italy, and Spain for Pesticides
Commonly Used on Tobacco in the United States during the 1990s 39 Figure

Figure 1: EPA*s Tiered Approach to Assessing Health Risks of Exposure to
Residues on Tobacco 17 Abbreviations

1,3- D 1,3- dichloropropene 2,4- D 2,4- dichlorophenoxyacetic acid 2,4,5-
T 2,4,5- trichlorophenoxyacetic acid AMS Agricultural Marketing Service
CORESTA Cooperation Centre for Scientific Research Relative to

Tobacco DBCP Dibromocloropropane DDE Dichlorodiphenyldichloroethylene DDT
Dichlorodiphenyltrichloroethane EDB Ethylene dibromide EPA Environmental
Protection Agency FDA Food and Drug Administration FFDCA Federal Food,
Drug, and Cosmetic Act FIFRA Federal Insecticide, Fungicide, and
Rodenticide Act FQPA Food Quality Protection Act FSA Farm Service Agency
HCB Hexachlorobenzene

Page iii GAO- 03- 485 Pesticides on Tobacco

NCFAP National Center for Food and Agricultural Policy NCSU North Carolina
State University ppm parts per million TDE Tetrachlorodiphenylethane TTR
total toxic residue USDA U. S. Department of Agriculture

This is a work of the U. S. Government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. It may contain
copyrighted graphics, images or other materials. Permission from the
copyright holder may be necessary should you wish to reproduce copyrighted
materials separately from GAO*s product.

Page 1 GAO- 03- 485 Pesticides on Tobacco

March 26, 2003 The Honorable Henry A. Waxman Ranking Minority Member
Committee on Government Reform House of Representatives

Dear Mr. Waxman: As you know, pesticides are used regularly on food and
nonfood crops, such as tobacco, to control a range of unwanted animal,
plant, and microbial pests. 1 Trace amounts of pesticides, called
residues, remain on tobacco and other crops after treatment. Typically,
the residue levels on tobacco decline as the leaves are harvested, dried,
and further processed into consumer products, and still further when the
tobacco is burned. However, varying residue levels may remain. As a
result, human exposure to pesticide residues on tobacco may occur when
residues remaining in cigarette smoke are inhaled. While much is known
about the significant health risks of using tobacco products, limited
information exists on the extent to which the use of pesticides on tobacco
may increase the considerable health risks associated with tobacco use
itself.

By controlling pests that reduce crop yields, pesticides can provide more
abundant supplies of fruits, vegetables, and other crops. Nonetheless,
pesticides are generally designed to be toxic to living organisms and thus
can have significant adverse health effects if used improperly. The
Environmental Protection Agency (EPA) determines whether and under what
conditions pesticides can be used in the United States without posing an
unreasonable risk to human health or the environment. Pesticides that meet
EPA*s requirements are granted licenses or *registrations,* which permit
their distribution, sale, and use according to specific directions and
requirements identified on the labels. In addition, the U. S. Department
of Agriculture (USDA) and the Food and Drug Administration (FDA) monitor
crops for certain pesticide residues. 1 This report generally uses the
term *pesticide* to refer to the toxic compounds, also called

active ingredients, that are contained in pesticide products. Pesticide
products typically include at least one active ingredient as well as inert
ingredients. United States General Accounting Office Washington, DC 20548

Page 2 GAO- 03- 485 Pesticides on Tobacco

In response to your request for information on how the federal government
addresses the public health implications of pesticides on tobacco, this
report describes (1) the pesticides commonly used on tobacco and the
potential health risks associated with them; (2) how EPA assesses and
mitigates health risks associated with pesticides used on tobacco; and (3)
how, and the extent to which, federal agencies regulate and monitor
pesticide residues on tobacco. Several pesticide use surveys conducted or
sponsored by the federal government provide information on the types and
amounts of pesticides commonly used on tobacco and other crops in the
1990s. These data, available through 1998, estimate average annual
agricultural use of pesticides, excluding such uses as pest control in
greenhouses. To determine how EPA assesses and mitigates health risks
associated with pesticides used on tobacco, we reviewed, among other
things, studies and documentation related to 13 pesticides commonly used
on tobacco that EPA evaluated under its reregistration program between
1994 and 2002. For more details on our scope and methodology, see appendix
I.

Surveys conducted during the 1990s indicate that tobacco producers in the
United States commonly used 37 pesticides approved for such use by EPA,
most of which were also approved for use on food crops. When used in

ways that deviate from the conditions set by EPA, many of these pesticides
can cause moderate to severe respiratory and neurological damage* and may
result in death. Further, animal studies suggest that some of these
pesticides may cause birth defects and cancer. About half of the
pesticides used on tobacco work primarily by preventing the normal flow of
nerve impulses to muscles and are among those most often

implicated in poisonings, injuries, and illnesses. In humans, symptoms
appear within minutes to hours after exposure and range from tightness in
the chest, headache, nausea, and dizziness to death from respiratory
failure.

EPA bases its assessments of the health risks to both workers and the
general population from exposure to the pesticides that are used on
tobacco and other crops on its evaluation of a wide range of toxicity,
residue, and other data. Workers are exposed through mixing and applying
pesticides, and the general population is exposed through pesticide

products used in the home and in public places and through pesticide
residues in food and water. EPA also assesses the health risks to smokers
from exposure to pesticide residues that remain in cigarette smoke by
analyzing data on the toxicity of specific pesticides and the residue
levels that remain on tobacco and in tobacco smoke. EPA*s assessments of
risks Results in Brief

Page 3 GAO- 03- 485 Pesticides on Tobacco

to workers and the public from exposure to pesticides that are used on
tobacco and other crops have identified a range of potential adverse
health effects. To mitigate such effects, EPA has set special limitations
on where

and how some pesticides may be used, such as requiring workers who use
them to wear respirators and chemical- resistant clothing, prohibiting
their use in certain states to avoid high pesticide levels in groundwater
used for drinking, and not permitting certain uses at all. On the other
hand, EPA has generally concluded that the low levels of residues measured
in tobacco smoke do not pose short- term health concerns. EPA does not
assess the additional risk of either intermediate- or long- term exposure
to pesticide residues in smoke because of the severity and quantity of
health effects associated with the use of tobacco products themselves. In

addition, the agency does not include short-, intermediate-, or long- term
exposure to residues on tobacco in its assessments of total exposures to
the pesticides. Officials and experts with whom we spoke generally agreed
that pesticide residues on tobacco could incrementally increase health
risks, though some also said the known harm from using tobacco products
dwarfs any potential effect from exposure to pesticide residues in the
smoke.

While EPA regulates the specific pesticides that may be used on tobacco
and other crops and specifies how the pesticides it approves may be used,
EPA does not otherwise regulate the residues of pesticides approved for
use on tobacco and other nonfood crops. USDA, however, is required by the
Dairy and Tobacco Adjustment Act, as amended, to test domestic and
imported tobacco for pesticides not approved for use on tobacco by EPA. As
a result, federal regulation of pesticide residues on tobacco is limited
to selected pesticides that are not approved by EPA for such use in the
United States. USDA tests most imported tobacco, as well as the portion of
domestic tobacco the federal government acquires under the tobacco price
support program, for residues of 20 pesticides not approved for use

on tobacco that federal officials believe are used in some other
countries. Most of these pesticides, such as DDT, are highly toxic,
persist in the environment, and accumulate in the bodies of humans and
animals. 2 By helping to ensure that other countries do not use pesticides
that U. S. tobacco growers are not allowed to use, the federal regulation
of pesticide residues on tobacco addresses trade equity as well as health
and

2 Dichlorodiphenyltrichloroethane, known as DDT, was one of the most
widely used chemicals for controlling insect pests on crops after 1945.
Under the authority of EPA, all registrations of DDT have been cancelled,
prohibiting the use of the pesticide in the United States.

Page 4 GAO- 03- 485 Pesticides on Tobacco

environmental issues. USDA has not reevaluated since 1989 the pesticides
the department monitors in its tobacco pesticide residue testing program,
although EPA has subsequently cancelled tobacco uses for at least 30
pesticides not currently monitored by USDA. Consequently, USDA*s testing
program excludes some highly toxic pesticides that may still be used in
other countries. To better protect the public from residues not approved
for use on tobacco, we are recommending that USDA periodically reevaluate
the pesticides it includes in its testing program. Tobacco is a high-
value, pesticide- intensive crop. That is, tobacco is the

nation*s ninth highest valued crop, and in terms of the amount of
pesticide applied per acre, tobacco ranks sixth* behind potatoes,
tomatoes, citrus, grapes, and apples. In the United States, tobacco is
grown in 16 states, 2 of

which* Kentucky and North Carolina* produce about two- thirds of all
domestic tobacco. 3 Further, it is grown in over 100 countries. Until
recently, the United States was the world*s leading exporter of
unmanufactured tobacco; however, in 2001, it ranked third, behind Brazil

and Zimbabwe. The tobacco industry in the United States both exports
tobacco to Japan and Western Europe* principally Germany, the Netherlands,
Denmark, the United Kingdom, Belgium, Italy, and Spain* and imports
tobacco in increasing amounts from countries such as Brazil, Argentina,
Malawi, and Thailand. Furthermore, the United States is the second largest
producer of cigarettes in the world, following China. More than 90 percent
of the tobacco grown in the United States is used to manufacture
cigarettes, as is most tobacco produced in the world. The remainder is
used for chewing tobacco, snuff, cigars, and pipe tobacco. Tobacco types
are often defined by such characteristics as how the tobacco is cured
(flue-, air-, or sun- cured), as well as the color, size, and thickness of
the leaves. Different types of tobacco are used in the various tobacco
products. The tobacco component of cigarettes made in the United States
usually consists of flue- cured and burley tobacco blended with imported
oriental tobacco and small amounts of specialty tobaccos grown in Maryland
and Pennsylvania. 4 3 The 14 other states are Connecticut, Florida,
Georgia, Indiana, Maryland, Massachusetts,

Missouri, Ohio, Pennsylvania, South Carolina, Tennessee, Virginia, West
Virginia, and Wisconsin.

4 In addition to tobacco, cigarettes contain other ingredients (additives)
to enhance flavor and other qualities of the product. Background

Page 5 GAO- 03- 485 Pesticides on Tobacco

Although pesticides play a significant role in increasing production of
tobacco, food, and other crops by reducing the number of crop- destroying
pests, exposure to pesticides can harm humans. The potential for harm is

related to both the amount of a substance a person is exposed to* the
dose* and the toxicity of the chemical. For example, small doses of
aspirin can be beneficial to people, but at very high doses, this common
medicine can be deadly. Furthermore, in some individuals, even at very low
doses, aspirin may be lethal. The age and health status of an individual
can also affect the potential for harm. Children may be more susceptible
to harm because, for example, they eat more food, drink more water, and

breathe more air than adults per pound of body weight, resulting in
greater exposure. Generally, assessments of dose and response involve
considering the dose levels at which adverse effects are observed in test
animals and using these dose levels to calculate an equivalent dose in
humans.

In many cases, exposure to pesticides is through residues that remain on
crops following use of the pesticides. The amount of pesticide residue
that remains reflects, among other things, the amount of pesticide
applied, the

time lapsed since application, and the speed with which the pesticide
dissipates in the environment. Residue levels remaining on crops are also
affected by where the pesticides are applied, such as in the soil or on
the plant, and when they are used in the life cycle of the plant, such as
when the plant is a seedling or shortly before the plant is harvested.
Typically, residues on tobacco decline as the plant moves from field to
finished consumer product.

The primary federal requirements pertaining to the registration, sale, and
use of pesticides are in the Federal Insecticide, Fungicide, and
Rodenticide Act (FIFRA) and the Federal Food, Drug, and Cosmetic Act
(FFDCA), both as amended by the Food Quality Protection Act (FQPA).
Pesticides must generally be registered with EPA in order to be sold or
distributed. EPA will register a pesticide if it determines, among other
things, that the pesticide will not generally cause unreasonable adverse
effects on human health or the environment when used in accordance with
conditions specified on the label. Throughout this report we will focus on
EPA*s analysis of potentially harmful effects to human health, rather than
the environment.

In 1988 FIFRA was amended to require that EPA review pesticides initially
registered prior to November 1984* when less toxicity data were available*
to consider their health effects and to determine whether and how they
might continue to be registered. These reviews are designed to

Page 6 GAO- 03- 485 Pesticides on Tobacco

ensure that older pesticides meet contemporary health and safety standards
and that their risks are mitigated. Essentially, manufacturers of the
older pesticides must provide EPA with substantially the same toxicity,
chemistry, and other data as are now required to register a new pesticide.
5 EPA reviews of the older pesticides are called reregistrations. Most of
the pesticides used on tobacco during the 1990s were initially approved
before 1984 and therefore are subject to reregistration.

In addition, the FQPA amendments to FIFRA passed in 1996 require EPA to
reevaluate the amounts of pesticide residues allowed on or in food* known
as tolerances. EPA must ensure that there is a reasonable certainty that
no harm will result from all pesticide exposures from food and

nonfood uses for which there is reliable information. In doing so, unless
another safety factor is determined to be appropriate, EPA is required to
apply an additional 10- fold safety factor in setting tolerances to ensure
the safety of foods for children. EPA is also required to ensure that
there is reasonable certainty that no harm will result to children
specifically from *aggregate* exposure to a single pesticide* that is,
from all sources, such as lawn treatments, household uses, drinking water,
and food. EPA must

also consider available information concerning the cumulative effects on
children of pesticides that act in a similar harmful way (known as a
common mechanism of toxicity). To accomplish this requirement, EPA has
recently developed a method to evaluate the cumulative exposure of one
class of highly toxic pesticides* the organophosphates* from residues in
food and drinking water and from residential uses. EPA uses risk
assessment* the systematic, scientific description of

potential adverse effects from exposure to hazardous substances* to
evaluate the potential health impacts of a pesticide on humans and
determine what measures are needed to mitigate identified risks. The
product of a risk assessment is an identification of the various health
risks, along with quantitative and/ or qualitative statements regarding
the probability that an exposed population will be harmed and to what
degree. For example, EPA qualitatively classifies pesticides and other
toxic substances according to their potential to cause cancer using
descriptors such as *likely* or *suggestive evidence but not sufficient to
assess human carcinogenic potential.* In addition, for many carcinogens,
EPA develops a quantitative dose/ response health risk assessment that
estimates the

5 Typically, applicants for pesticide registrations are the manufacturers;
EPA calls the applicants *registrants.*

Page 7 GAO- 03- 485 Pesticides on Tobacco

health risks at varying exposures. For health effects other than cancer,
EPA may calculate what it terms a *reference dose* or, in the case of
exposure by inhalation, a *reference concentration,* which represents a
daily level of exposure that is unlikely to result in harm over a
lifetime. Alternatively, EPA may calculate a *margin of exposure,* which
is a ratio that shows how far the actual (or estimated) human exposure to
a substance is from levels that are harmful. In essence, evaluating and
managing the risk of exposure to a pesticide involves determining the
maximum safe level of exposure to the pesticide and assessing whether
expected actual exposure is below this maximum level. If expected actual
exposure levels exceed the maximum safe amount, EPA must determine the
best ways to reduce exposure.

According to federally sponsored surveys, during the 1990s tobacco
producers in the United States commonly used 37 of the pesticides approved
by EPA for such use. 6 As shown in table 1, most of the pesticides used on
tobacco were insecticides and herbicides, which control insect and plant
pests; others were fungicides, which combat fungal diseases, or plant
growth regulators; and a few had more than one use. 7 6 The National
Center for Food and Agricultural Policy, a private nonprofit, nonadvocacy
research organization, conducted two key surveys for USDA covering the
periods 1990- 93

(called the 1992 survey) and 1994- 98 (called the 1997 survey). See
appendix I for more detail on these surveys.

7 In this report, the term *insecticide* includes pesticides used to
control insects, spiders, and nematodes (worms). Pesticides Commonly Used
on Tobacco

Have Potential Short- and Long- Term Adverse Health Effects

Page 8 GAO- 03- 485 Pesticides on Tobacco

Table 1: Pesticides Commonly Used on Domestic Tobacco, 1990- 98 Primary
use( s) Pesticide

Insecticide Acephate, aldicarb, Bacillus thuringiensis, carbaryl,
carbofuran, chlorpyrifos, diazinon, disulfoton, endosulfan, ethoprop,
fenamiphos, fonofos, imidacloprid, malathion, methidathion, methomyl,
spinosad, trichlorfon Herbicide Benefin, clomazone, diphenamid,
isopropalin,

napropamide, pebulate, pendimethalin, sethoxydim, sulfentrazone Fungicide
Dimethomorph, mancozeb, mefenoxam, metalaxyl Plant growth regulator
Ethephon, flumetralin Plant growth regulator, herbicide Maleic hydrazide
Fumigant, insecticide Chloropicrin Fumigant, insecticide, herbicide Methyl
bromide Fungicide, insecticide, herbicide 1,3- dichloropropene (1,3- D)
Source: EPA, International Organization for Standardization, National
Center for Food and Agricultural Policy, and USDA. Note: GAO*s analysis of
EPA, International Organization for Standardization, National Center for
Food and Agricultural Policy, and USDA data.

Most of these pesticides were also widely used on food crops. The actual
number and amount of pesticides used on tobacco or other crops in any
given year vary depending on factors such as the weather and the specific
pests that become problematic. For example, the incidence of many plant
diseases is closely correlated to the amount of rainfall, resulting in
greater use of fungicides in years with high rainfall. In addition,
pesticide use

tends to change over time as pests develop resistance to the pesticides
and as use on tobacco is approved for new pesticides and cancelled for
older pesticides. 8 As table 2 shows, 10 pesticides identified in the 1997
survey as commonly used on tobacco were not identified in the earlier
survey. Two of these pesticides, dimethomorph and mancozeb, began to be
used in response to the appearance of a disease resistant to metalaxyl,
which declined in usage during the 1990s. In addition, during the years
included in the 1997 survey, tobacco use for 5 of the 7 pesticides no
longer reported as being used* diazinon, diphenamid, isopropalin,
methidathion, and

trichlorfon* was being cancelled. 9 In some cases, pesticide cancellations
resulted in the increased use of other pesticides. For example, by 1997
clomazone had replaced diphenamid and isopropalin as the pesticide of

8 During the 1990s, EPA cancelled tobacco use for approximately 34
pesticides. 9 When EPA cancels the use of a pesticide, the pesticide is
typically phased out over time.

Page 9 GAO- 03- 485 Pesticides on Tobacco

choice for controlling unwanted weeds, and imidacloprid was most commonly
used to control insect pests, leading to reduced use of acephate,
aldicarb, chlorpyrifos, ethoprop, and carbofuran. Manufacturers may
initiate cancellation of some or all uses of a pesticide, often for
economic reasons, 10 or EPA may cancel uses when the agency determines
that one or more uses pose unreasonable risks to human health or the
environment. For example, as required under the Clean Air Act, EPA has
been phasing out the use of methyl bromide on tobacco and a wide range of
other crops because it depletes the earth*s protective layer of ozone. 11
Methyl bromide use on tobacco decreased from about 5.4 million pounds

in 1992 to about 0.7 million pounds in 1997 because of EPA*s efforts and
changes in how tobacco producers raise seedlings. Specifically, producers
have begun to grow tobacco seedlings in greenhouses, where methyl bromide
is not generally used.

Table 2: Pesticide Use on Tobacco, 1990- 98 Pesticide

Pounds used on tobacco 1992 survey (1990- 93)

Pounds used on tobacco 1997 survey (1994- 98)

1,3- dichloropropene (1,3- D) 11,537,540 13,279,285 Chloropicrin 577,082
6,761,644 Maleic hydrazide 1,789,208 1,790,089 Acephate 1,570,457 871,899
Methyl bromide 5,356,748 685,026 Pendimethalin 321,931 473,718
Chlorpyrifos 685,554 406,822 Fenamiphos 257,142 379,841 Mancozeb a 356,811
Flumetralin a 352,742 Metalaxyl 371,645 271,368 Clomazone a 217,617
Ethoprop 438,274 182,321 Endosulfan a 172,766 Mefenoxam a 139,199 Pebulate
412,000 131,665 Ethephon 113,238 102,130

10 One reason manufacturers may choose to request cancellation of
pesticide registrations is to avoid costs associated with reregistering
pesticides for each use, such as the cost of providing EPA with data and
studies.

11 EPA*s efforts to phase out the use of methyl bromide in the 1990s were
consistent with international efforts to curtail its use under the
Montreal Protocol, a treaty signed by over 160 countries to control the
production and trade of ozone- depleting substances globally.

Page 10 GAO- 03- 485 Pesticides on Tobacco

Pesticide Pounds used on tobacco

1992 survey (1990- 93) Pounds used on tobacco

1997 survey (1994- 98)

Napropamide 191,840 92,622 Sulfentrazone a 69,073 Imidacloprid a 67,896
Aldicarb 159,044 59,719 Dimethomorph a 36,818 Methomyl 57,137 29,773
Malathion 7,549 15,437 Disulfoton 52,578 13,495 Sethoxydim a 9,579
Spinosad a 2,815 Carbaryl 16,487 2,057 Fonofos 12,798 16 Benefin 56,963 a

Bacillus thuringiensis b b Carbofuran 149,965 a Diazinon 53,670 a
Diphenamid 81,624 a Isopropalin 129,287 a Methidathion 68 a Trichlorfon
722 a

Total used on tobacco 24,400,552 26,974,241

Source: National Center for Food and Agricultural Policy. a Not identified
as being used in the survey.

b Neither survey estimated the use of Bacillus thuringiensis in pounds.

EPA determines the amounts and conditions under which a pesticide may be
used so that it will not pose unreasonable risks to workers or the general
population. Failure to comply with the conditions set by EPA could result
in a range of harmful effects. For example, 17 of the 37 pesticides
commonly used on tobacco in the 1990s belong to three chemical classes
that, at high doses, are known to cause adverse human health effects up to
and including death (see table 3).

Page 11 GAO- 03- 485 Pesticides on Tobacco

Table 3: Organochlorine, Organophosphate, and Carbamate Pesticides
Commonly Used on Tobacco in the 1990s

Chemical class Pesticide name Organochlorine Endosulfan Organophosphate
Acephate, chlorpyrifos, diazinon, disulfoton, ethoprop, fenamiphos,

fonofos, malathion, methidathion, trichlorfon Carbamate Aldicarb,
carbaryl, carbofuran, mancozeb, methomyl, pebulate Source: EPA, National
Center for Food and Agricultural Policy, and USDA. Note: GAO*s analysis of
EPA, National Center for Food and Agricultural Policy, and USDA data.

Although they do not all produce their toxic effects in the same way,
pesticides in these three classes* organochlorines, organophosphates, and
carbamates* act on the nervous system to prevent the normal flow of nerve
impulses to muscles that control both voluntary movement, such as walking,
and involuntary movement, such as breathing and heart beat. Pesticides in
all three classes are absorbed to varying degrees through inhalation,
ingestion, and skin contact. Exposure to amounts of these pesticides that
exceed levels set by EPA could result in immediate and lifethreatening
effects, such as respiratory failure, or conditions that do not appear
immediately, such as cancer. While EPA has concluded that most of these 17
pesticides do not cause birth defects, the agency has also concluded that
5 of them and a by- product of another may cause cancer. 12 Since the
1970s, EPA has severely restricted its approvals of

organochlorine pesticides, which include DDT, aldrin, and chlordane,
because of their potential to harm humans and the environment.
Organochlorine pesticides persist in the environment* some have remained
in soil for over 50 years* and accumulate in body tissue, particularly
fat. Organochlorine pesticides are associated with a range of adverse
health effects, including cancer and damage to the neurological and
reproductive systems. The one organochlorine pesticide still approved for
use on tobacco, endosulfan, is highly toxic when ingested or inhaled and
slightly toxic through contact with the skin. While EPA has determined
that it is unlikely to cause cancer as other members of this class do,
endosulfan, like all organochlorine pesticides, primarily affects the
nervous system. EPA has requested additional data from the manufacturer to
address its concerns that exposure to endosulfan could

12 EPA will assess the health effects of carbofuran in fiscal year 2003
during its reregistration review.

Page 12 GAO- 03- 485 Pesticides on Tobacco

harm the nervous system of developing fetuses. Organophosphate and
carbamate pesticides have largely replaced the organochlorine pesticides
in the United States.

While they break down quickly in the environment and do not accumulate in
body tissues, organophosphate pesticides are much more acutely toxic to
humans and animals than the persistent organochlorine pesticides they have
largely replaced. The primary cause of death from organophosphate
poisoning is respiratory failure, although cardiovascular symptoms, such
as decreased heart rate that progresses to cardiac arrest, usually occur
as well. In humans, additional symptoms from exposure to organophosphate

pesticides, which can develop during use or within minutes to hours after
exposure, include headache, nausea, dizziness, sweating, muscle twitching,
anxiety, and depression. Exposure by inhalation causes the most rapid
appearance of toxic symptoms. As a result, to minimize the potential for
harmful exposure of workers, EPA requires those who mix, use, or apply the
pesticides to have special training, use respirators, and wear chemical-
resistant clothing. Regarding the potential to cause cancer, EPA has
determined that 4 of the 10 organophosphate pesticides used on tobacco*
acephate, ethoprop, methidathion, and trichlorfon* may cause

cancer. In addition, EPA has concluded that 7 of the 8 organophosphate
pesticides it evaluated for their potential to cause birth defects would
not cause them but that the eighth* chlorpyrifos* may do so at very high
levels that may also harm the pregnant female. 13 Carbamates, which also
affect the central nervous system, produce

symptoms similar to those of organophosphate pesticides, although the
effects of carbamate poisoning tend to be of shorter duration and somewhat
easier to treat. The primary cause of death from carbamate poisoning is
respiratory failure. Of the six carbamate pesticides used on tobacco, EPA
has determined that one and a by- product always associated with another
may cause cancer; two are unlikely to cause cancer; data are insufficient
to determine the cancer- causing potential of one; and one will be
evaluated in fiscal year 2003. EPA has evaluated four of the carbamates

for their potential to cause birth defects: three do not and only minimal
evidence exists for the potential of the fourth to cause birth defects.
EPA has requested, but not yet received, data from the manufacturer on the

13 EPA has requested data to assess the potential of trichlorfon, which is
no longer approved for use on tobacco, to cause birth defects and has
terminated its assessment of fonofos because all uses of the pesticide
were cancelled.

Page 13 GAO- 03- 485 Pesticides on Tobacco

potential of one of the two remaining carbamate pesticides to produce
birth defects, and the agency will evaluate the health effects of the
other in fiscal year 2003.

The potential acute adverse health effects from the remaining 20
pesticides used on tobacco* representing 12 different chemical classes*
range from mild to severe. 14 For example, EPA found no known health
effects on mammals from exposure to Bacillus thuringiensis as it is
currently manufactured. Similarly, EPA has found that both maleic
hydrazide, a plant growth regulator and herbicide, and metalaxyl, a
fungicide, have low acute toxicity, and neither is thought to cause cancer
or birth defects. However, EPA has found that serious adverse health
effects could occur with high exposures to insecticides, such as
chloropicrin, 1,3- dichloropropene (1,3- D), and methyl bromide, which are
applied as fumigants and can be severely irritating to the eyes, skin, and
lungs. EPA has also found that poisoning from exposure to methyl bromide
may result in persistent neurological impairment. In general, because most
of the pesticides used on tobacco are widely

used on food and other crops, as well as in residential and other
settings, the exposure resulting from residues on tobacco represents a
small portion of total exposure to these pesticides. Specifically, 1997
survey data estimate that about 27 million pounds of the 37 pesticides
were used on tobacco, while the estimated use of these pesticides
nationally on all crops was 175 million pounds. Therefore, most of the
exposure to these pesticides stems from their use on other crops and in
other products, such as household insecticides. However, for some
pesticides* dimethomorph,

fenamiphos, flumetralin, maleic hydrazide, mefenoxam, and sulfentrazone*
more than 50 percent of their use in 1994 through 1998 was on tobacco.
Further, more than 80 percent of maleic hydrazide used and 100 percent of
flumetralin and sulfentrazone used were applied to tobacco. Appendix II
provides information on the amounts of the 37 pesticides used on (1)
tobacco and (2) domestic crops, as estimated in the 1992 and 1997 surveys.

14 One herbicide, clomazone, has not been classified chemically.

Page 14 GAO- 03- 485 Pesticides on Tobacco

To determine whether the use of individual pesticides can reasonably be
expected not to harm human health, EPA conducts health risk assessments
under its pesticide registration program. These risk assessments are based
on EPA*s evaluations of the results of numerous scientific studies and
tests that the agency requires pesticide manufacturers to carry out. EPA
also assesses the health risks to smokers from exposure to pesticides used
on tobacco by analyzing data on their

toxicity and the residue levels that remain on tobacco and in tobacco
smoke. Because pesticides are used extensively on crops, including
tobacco, and in home pesticide products, the risk assessments focus on
exposures of (1) workers who handle the pesticides and (2) the general
public, which is exposed to pesticides via residues on food or in drinking
water or from pesticide products used in and around the home and in public
places. EPA*s health risk assessments often identify risks to workers that
must be mitigated before EPA will approve the pesticide. The assessments
also identify risks to the general population that may also require
special limitations on how or where the pesticides may be used. EPA has
generally concluded that the low levels of residues measured in

tobacco smoke do not pose health concerns that require mitigation. While
EPA officials were generally able to provide us with copies of the studies
and evaluations we requested during our review, documentation of the
agency*s evaluation of the validity and reliability of the residue studies
was inconsistently available.

Under its pesticide registration program, EPA routinely assesses the
health risks of exposure to pesticides from residues in drinking water and
food and from pesticide use in the home, in public places, and at work.
The Health Effects Division of the Office of Pesticide Programs in EPA
develops its health risk assessments on the basis of a substantial body of
data, including toxicity, residue chemistry, and other data provided by
pesticide manufacturers, as well as other relevant information, such as
human and animal studies from the general scientific literature and
poisoning incident databases. The risk assessments focus on the potential
cancer and noncancer health risks associated with short- term (acute),

intermediate-, and long- term (chronic) exposures to pesticides from the
primary exposure routes* oral, inhalation, and contact with skin (dermal).
Noncancer health risks that EPA assesses include risk of birth defects,
reproductive impairments, damage to genetic material, and interference EPA
Concludes that

Health Risks of Pesticide Residues on Tobacco Are Minimal but Requires
Mitigation for Risks from Other Exposures

EPA Assesses Health Risks of Varied Exposures to Pesticides

Page 15 GAO- 03- 485 Pesticides on Tobacco

with the body*s endocrine system. 15 EPA*s health risk assessments are
subject to numerous reviews by a variety of committees, including the
agency*s Hazard Identification Science Assessment Review Committee, Cancer
Science Assessment Review Committee, and Reproductive and Developmental
Toxicity Science Assessment Review Committee. The health risk assessments
provide critical information to the pesticide

registration divisions on the human health component of risk management
decisions* such as whether to approve pesticides for use; what amounts may
be used; and what special restrictions, if any, may be needed.

To evaluate the levels of pesticides to which cigarette smokers might be
exposed from residues on tobacco, EPA reviews plant metabolism and residue
studies provided by manufacturers that identify the residues of
pesticides, and any harmful by- products 16 they may produce, that remain
on the crop after it has been treated. The plant metabolism studies reveal
how plants process a pesticide once it is applied and the relative amounts

of the pesticide and its by- products that remain after treatment* the
total toxic residue (TTR). The residue studies, called field trials,
quantify the levels of pesticide and by- product residues that remain on
plants grown under actual agricultural conditions that approximate the
expected *real life* environment. Such field trial data, which are
required for all

pesticides that will be used on food, may not always be required for
pesticides used on tobacco because EPA uses a *tiered* approach to
evaluate residues on tobacco. That is, for tobacco, the agency requires
additional residue data after the metabolism study only if it has shown
that the combined residue levels of the pesticide itself and any harmful
by- products exceed 0.1 parts per million (ppm)* the agency*s *threshold
of concern* for residues on tobacco. Thus, as figure 1 shows, EPA
generally requires plant metabolism studies for green tobacco and may
require data from field trials for both green and cured (aged) tobacco, 15
The body*s endocrine system produces hormones that help guide the
development, growth, reproduction, and behavior of animals, including
humans. Some chemicals can

interfere with the normal function of this complex system in ways that
mimic a natural hormone, thereby fooling the body into overresponding to a
hormone or blocking the effects of a hormone. Others may directly
stimulate or inhibit the endocrine system, leading

to overproduction or underproduction of hormones. Certain drugs are used
to intentionally cause some of these effects, such as birth control pills.
16 EPA requires metabolism studies to identify by- products of pesticides
that are of toxicological concern or that account for 10 percent or more
of the total radioactive

residues. (For metabolism studies, the pesticides have had radioactive
atoms attached to their molecular structure to allow tracking of the
pesticide through the plant.)

Page 16 GAO- 03- 485 Pesticides on Tobacco

depending upon the amount of residues that are identified. 17 In addition,
EPA may require pyrolysis studies that measure the residues in smoke when
tobacco treated with a pesticide is burned. Finally, EPA may require
additional residue studies to estimate potential exposure, even if the
residues are below 0.1 ppm, if it has concerns about the toxicity of a
pesticide.

17 The results of metabolism studies on food crops may be used to
determine the identity of the residues of concern.

Page 17 GAO- 03- 485 Pesticides on Tobacco

Figure 1: EPA*s Tiered Approach to Assessing Health Risks of Exposure to
Residues on Tobacco

a Residues are measured as TTR* the sum of the residue from the parent
pesticide and its by- products (degradation products, metabolites, and
impurities that are of toxicological concern). b Pyrolysis refers to
chemical change brought about by the action of heat (burning).

Page 18 GAO- 03- 485 Pesticides on Tobacco

The tiered approach to analyzing residues on tobacco reflects the fact
that, typically, pesticide residues on tobacco decline over time, as the
tobacco is stored, cured, manufactured into cigarettes, and burned during
smoking. 18 EPA uses the tiered approach for tobacco, in part, because the
agency has concluded that the potential for harm to human health from
pesticide residues on tobacco at or below the 0.1- ppm level is extremely

low and unlikely to result in a risk of concern to smokers. 19 According
to EPA officials in the Health Effects Division, since August 1999, EPA*s
policy for assessing the health risks from using pesticides on tobacco has
been to evaluate the risks of short- term exposure to residues on tobacco
and to quantify the estimated health risks using a consistent method and
set of assumptions. 20 This policy is applied to all newly registered
pesticides, as well as to currently registered pesticides as they are
periodically reviewed to ensure they meet current human health and
environmental safety standards in accordance with the requirements of the
1988 amendments to FIFRA. EPA officials attribute the more structured
approach to advances in the science of risk assessment and the 1996
enactment of FQPA, which has spurred the agency to more systematically
quantify the exposure to pesticide residues in food and drinking water and
from residential uses.

EPA selected the margin of exposure method to quantify the health risks
associated with exposure to pesticide residues in smoke. As discussed
earlier, a margin of exposure shows how far the actual (or estimated)
human exposure to a substance is from levels that have been shown to

18 EPA does not use the tiered approach to analyze pesticide residues on
food and feed crops. Instead, it generally requires data on how the plants
metabolize the pesticide and studies of residues that remain at the time
of harvest (field trials), and, in some cases, as the food is processed.

19 EPA*s conclusion is based on its analysis, which compared its threshold
(0. 1 ppm) with residue levels at which inhalation exposure to other
pesticide residues produces no harmful effects. Specifically, the 0.1- ppm
threshold represents an estimated actual exposure to pesticide residues
equivalent to half of the lowest residue level that EPA officials know of
for inhalation exposure that does not produce adverse effects. 20 Prior to
1999, the determination of the methodology and assumptions for assessing
the

risk to smokers of pesticide residues on tobacco was left to the
discretion of individual toxicologists, who made independent
determinations for each chemical based on what they considered to be
reasonable assumptions. Quantitative risk assessments generally were not
performed. This new policy was formalized as guidance in August 2000 and
is available electronically to EPA*s risk assessors as part of the Health
Effects Division*s Risk Science Assessment Review Committee Library.

Page 19 GAO- 03- 485 Pesticides on Tobacco

cause no harm in animal studies. To estimate exposure, EPA typically uses
(1) the residue levels identified in tobacco field trials or pyrolysis
studies and (2) standard assumptions for key variables that affect
exposure. Specifically, EPA assumes that people smoke 15 cigarettes a day
21 and that they weigh about 150 pounds, if male, and 130 pounds, if
female.

Moreover, EPA assumes 100 percent of the pesticide residue on the tobacco
is inhaled and absorbed. In practice, some residues will be trapped in
cigarette butts, and the amount of smoke inhaled varies widely among
people. EPA officials said the assumptions are conservative* that is, they
are protective of public health* because they tend to overstate, rather
than understate, the extent to which smokers are exposed to the
potentially toxic effects of the pesticides.

Also according to EPA officials, the agency does not include exposure to
the residues in tobacco smoke in its aggregate health risk assessments of
individual pesticides, which are required by FQPA, because the added
exposure from residues in smoke is minimal. In addition, EPA has chosen
not to assess the risk of either intermediate- or long- term exposure to
pesticide residues in smoke because of the severity and quantity of health
effects associated with the use of tobacco products themselves.

Specifically, exposure to tobacco products* particularly cigarettes* is
the single major preventable cause of cancer and heart and lung disease in
the United States.

Finally, although experts and public health officials are concerned about
the potential for harm, particularly to children, from exposure to
pesticides, little is known directly about the chronic effects of
pesticide use in general in the United States* for example, in agriculture
and in schools. 22 Moreover, studies linking adverse human health effects
to exposure to pesticide residues on tobacco are rare, according to public
health officials and experts we spoke to. And while a number of federally
sponsored studies of the effects of exposure to pesticides are underway,
it will be years, if not decades, before conclusive results are known.
Officials

and experts we spoke with about possible harm from pesticide residues on
21 Pierce, J. P., et al., 1989. Tobacco Use in 1986 * Methods and Basic
Tabulations from Adult Use of Tobacco Survey. U. S. Department of Health
and Human Services Publication Number OM90- 2004. Office on Smoking and
Health, Rockville, Maryland.

22 U. S. General Accounting Office, Pesticides: Improvements Needed to
Ensure the Safety of Farmworkers and Their Children, GAO/ RCED- 00- 40
(Washington, D. C.: Mar. 14, 2000) and Pesticides: Use, Effects, and
Alternatives to Pesticides in Schools, GAO/ RCED- 00- 17 (Washington, D.
C.: Nov. 29, 1999).

Page 20 GAO- 03- 485 Pesticides on Tobacco

tobacco generally agreed that such residues could incrementally add to the
risk, and some also believed the known harm from using tobacco products
dwarfs any potential effect from exposure to pesticide residues in the
smoke.

EPA*s health risk assessments have identified a number of potential
adverse health effects associated with the pesticides used on tobacco and
other crops that, in some cases, have led the agency to impose special

limitations on the uses of these pesticides. The risks that required
mitigation stemmed from (1) potential exposure of workers who apply
pesticides or harvest crops and (2) potential exposure of the general
population to pesticide residues in food or drinking water or from
pesticides used in the home or in public. None of the risks requiring
mitigation were associated with exposure to residues on tobacco or in
tobacco smoke.

Our review of studies and other documentation related to EPA*s completed
reregistration reviews of 13 of the 37 pesticides commonly used on tobacco
identified the health risks associated with them and the related

mitigation measures the agency required. 23 The following cases illustrate
some of the health risks that have required mitigation.

EPA has classified 1,3- D, a widely used fumigant that controls soil-
borne pests and diseases, as a probable carcinogen* that is, evidence from
human and animal studies suggests that 1,3- D, once ingested or inhaled,
is likely to cause cancer. In its risk assessment, EPA determined that
1,3- D

could make its way to groundwater and pose a risk of cancer for residents
who obtained their drinking water from wells near treated fields. To
mitigate the potential cancer risks and as a condition for reregistration,
EPA required that wells used for drinking water be located 100 or more
feet from treated fields and prohibited the use of 1,3- D altogether in 11
states with porous soil. 24 In addition, vapors from 1, 3- D* which is
injected as a liquid into soil, where it quickly evaporates* can move into
the air. Consequently, EPA also required (1) a 300- foot buffer between
occupied

23 The 13 pesticides are 1,3- D, acephate, chlorpyrifos, diazinon,
disulfoton, endosulfan, ethoprop, ethephon, maleic hydrazide, metalaxyl,
methidathion, pebulate, and pendimethalin. 24 The 11 states where 1, 3- D
cannot be used are Maine, Massachusetts, Minnesota, Montana,

New Hampshire, New York, North Dakota, South Dakota, Utah, Vermont, and
Wisconsin. EPA Concludes That Risks

Associated with Pesticide Use Can Be Significant, but Those Associated
with Pesticide Residues on Tobacco Appear to Be Minimal

Some of EPA*s Risk Assessments Result in Special Mitigation Measures

Page 21 GAO- 03- 485 Pesticides on Tobacco

buildings and fields treated with the pesticide and (2) workers who apply
the pesticide to wear respirators and protective clothing, among other
things. Further, because of 1,3- D*s volatility and potential to harm
humans, EPA classified it as a *restricted use* pesticide, which means it
can only be applied by, or under the supervision of, individuals trained
to handle

particularly toxic or harmful pesticides. Currently, 1,3- D is registered
for use on soils in which all food and feed crops may be planted.
Moreover, according to the 1997 survey, an estimated 13 million pounds of
1,3- D were applied to tobacco annually during the survey period* almost
twice the amount of chloropicrin, the second most commonly used pesticide
on tobacco. 25 Despite the health risks posed by injecting 1,3- D into
soil, EPA identified no risks associated with residues on tobacco leaves
or in tobacco smoke because 1,3- D metabolizes to nontoxic by- products
and is subsequently absorbed by the plant. 26 Similarly, EPA determined
that residues on tobacco of chlorpyrifos*

another pesticide frequently used on tobacco and food crops and one of the
most widely used organophosphate insecticides in the United States* were
below the agency*s threshold of concern. But the agency determined that
chlorpyrifos presented potential health risks unrelated to its use on
tobacco that required strict mitigation measures. Specifically, the agency
identified health risks to children from exposure to chlorpyrifos. Before
2000, chlorpyrifos was one of the insecticides used most often in
residential and commercial settings* for example, on carpets and in
schools, daycare centers, hotels, and restaurants* and on food crops. EPA
identified significant risks to children from these many uses and required
stringent measures to address them. Between 1997 and 2000, EPA cancelled
nearly all indoor and outdoor residential uses and prohibited the use of
chlorpyrifos in schools and public parks. In addition, manufacturers
agreed to eliminate the use of chlorpyrifos on tomatoes and restrict its
use on apples. 27 EPA also identified concerns for some workers who mix,
load,

25 The 1997 survey estimated that, of the 37 pesticides used on tobacco in
the 1990s, 12 were applied to tobacco in amounts less than 100, 000 pounds
and 8 were shown as not used during this survey (1994- 98). See appendix
II for more detail on the estimated amounts of pesticides used on tobacco.
26 Although 1,3- D is also widely used on soils where food crops are
planted, EPA does not require that food use tolerances be established for
this pesticide because no residues remain on plants grown in treated soil.
27 Chlorpyrifos is currently registered to control foliage- borne and
soil- borne pests on food and feed crops; at golf courses; and on
nonstructural wood, such as utility poles and fence posts, as well as to
kill adult mosquitoes. Structural treatments for termites are also
registered uses but are being phased out by the end of 2005.

Page 22 GAO- 03- 485 Pesticides on Tobacco

and apply chlorpyrifos in agricultural and other nonresidential settings.
As a result, EPA required that workers wear a respirator and a double
layer of clothing, including chemical- resistant gloves, shoes, and
headgear. Workers must also use water- soluble packages to mix powdered
forms of chlorpyrifos and remain in an enclosed cockpit when aerially
spraying a field. EPA also set a time interval between applications of the
pesticide

and when workers can reenter treated areas, ranging from 24 hours for most
crops to 5 days for others. EPA did not, however, identify risks
associated with chlorpyrifos used on tobacco because residue levels on
green tobacco were below 0.1 ppm.

EPA also identified a range of potential harmful effects from other
exposures to the other pesticides we reviewed. For 11 pesticides,
including 1,3- D and chlorpyrifos, EPA identified a range of concerns,
largely for exposures of workers* particularly those engaged in spraying
the pesticides* that required at least some mitigation. Most often the

mitigation measures included the use of enclosed mixing systems and
tractor cabs, additional protective respirators and clothing, reductions
in the rate and frequency of application, and increases in the time
between application and reentry to the treated areas. In some cases, such
as for acephate, disulfoton, and ethoprop* all of which are
organophosphate pesticides* certain uses were cancelled, including use on
golf courses and lawns and indoor and outdoor residential uses. Three of
these 11

pesticides* disulfoton, endosulfan, and ethoprop* also raised concerns
about dietary or drinking water exposure for which EPA required such
mitigation as canceling use on some foods, reducing the rate and frequency
of application on others, and requiring buffer zones between treated
fields and water bodies. EPA placed a number of additional restrictions on
the use of endosulfan, a highly toxic and persistent organochlorine
pesticide, including restricting use on cotton and tobacco to certain
states; eliminating or reducing aerial spray applications on crops

such as strawberries, nuts, and tobacco; and requiring buffer zones
between treated areas and bodies of water. 28 In addition, EPA required
that all products containing endosulfan be labeled as restricted use
pesticides, which can only be used by, or under the supervision of,
specially trained 28 EPA determined that endosulfan can be used on tobacco
in only 6 of the 16 states where

it is grown* Indiana, Kentucky, Ohio, Pennsylvania, Tennessee, and West
Virginia* which account for about 40 percent of domestic production.

Page 23 GAO- 03- 485 Pesticides on Tobacco

applicators. 29 EPA also noted that it may require further restrictions on
acephate once the agency completes its assessment of the cumulative
exposure to organophosphate pesticides because this organophosphate
pesticide degrades in plants to another organophosphate pesticide. 30 EPA
found that 2 of the 13 pesticides we reviewed presented no concerns that
needed changes in existing conditions on how to use and apply the
pesticides. 31 The pesticides we reviewed, including ones no longer
approved for use in the United States, are used in many other tobacco-
producing countries,

according to experts. Researchers and advocacy groups have raised concerns
about adverse health effects on tobacco workers in other countries from
exposure to pesticides, citing such factors as the absence of cautionary
labels on some pesticides and the limited use of protective clothing by
agricultural workers. For example, researchers found elevated rates of
depression and suicide rates that were twice the national average among
tobacco producers in Brazil, a leading tobacco exporter. And although many
factors, such as poverty and stress, may play a role in suicide, one group
of researchers noted tobacco producers in Brazil routinely used
organophosphate pesticides, which have been shown to cause depression.
Moreover, these researchers reported that suicides are more likely to
occur during planting and harvesting seasons, when organophosphate
pesticides are used intensively. To some extent, such harmful exposure may
occur because pesticide regulations in other countries may be less
stringent than those in the United States or because other countries*
enforcement of regulations may be more limited, according to advocacy
groups.

Regarding pesticide residues on domestic tobacco, overall, EPA officials
did not find associated health risks that required mitigation. Further,
the data we reviewed on 13 pesticides were consistent with statements from
EPA officials that the residues on tobacco were below the agency*s

29 Of the 13 pesticides we reviewed, all of the currently approved
products containing 1,3- D, disulfoton, and ethoprop are restricted use
products, and some of the products containing chlorpyrifos are restricted
use products. 30 In June 2002, EPA issued a preliminary cumulative risk
assessment for organophosphates for public comment. The agency, in
consultation with scientific advisors, will revise the assessment on the
basis of comments and data that were submitted. No date has been set to
issue the final cumulative risk assessment for organophosphate pesticides.
31 Maleic hydrazide and metalaxyl.

Page 24 GAO- 03- 485 Pesticides on Tobacco

identified level of concern in 11 cases. EPA did not evaluate the
remaining 2 pesticides* diazinon and pendimethalin* for use on tobacco. In
the case of diazinon, evaluating residue data was not relevant because the
pesticide was no longer approved for use on tobacco at the time EPA
conducted its evaluation. In the case of pendimethalin, at the time we
conducted this

work, EPA had not yet reviewed the relevant data received from the
manufacturer. EPA approved the reregistration for this pesticide, but its
use on tobacco is subject to the agency*s evaluation of this data.

Of the 11 pesticides that EPA evaluated for use on tobacco, 3 left
residues on green or cured tobacco that were less than 0.1 ppm* and one
left no residues at all. Specifically, the maximum residues of ethoprop on
green tobacco were 0.01 ppm, the residues of chlorpyrifos were 0.09 ppm,
the

residues of pebulate on both green and cured tobacco were less than 0.02
ppm, and the plant metabolism study for 1,3- D showed no residues
remaining on the plant. Manufacturers provided pyrolysis studies in two of
the four cases in which the residue levels on green tobacco were 0.1 ppm
or less. The pyrolysis study for ethoprop identified residues in the smoke

that were below the agency*s level of concern. The pyrolysis study for a
by- product of chlorpyrifos that was initially of concern to the agency
identified the by- product in the smoke. However, EPA subsequently
concluded that the by- product, which accounted for more than 10 percent
of the residue in the smoke, was not of toxicological concern because,
unlike its parent compound, it does not act toxically on the nervous
system.

Of the seven pesticides that progressed through EPA*s tiered risk
assessment approach because residues on cured tobacco were greater than
0.1 ppm, pyrolysis studies were conducted on five. No residues were found
in the smoke of four of these five pesticides; the residues of the fifth
were not of sufficient magnitude to require further study or evaluation.
One of the remaining two pesticides with residue levels greater than

0.1 ppm was evaluated using a study of the health effects on rats exposed
to residues in smoke, 32 and one was approved subject to EPA*s review of
requested additional residue data, including a pyrolysis study, to confirm
EPA*s assessment that residues on tobacco do not pose a risk to human
health.

32 EPA issued the reregistration decision for this pesticide, metalaxyl,
in 1994. EPA officials said they would no longer substitute a rat
inhalation study for a pyrolysis study. Researchers found no difference
between rats exposed to smoke containing metalaxyl

residues and those that were not.

Page 25 GAO- 03- 485 Pesticides on Tobacco

The reregistration decisions for 7 of the 13 pesticides we reviewed were
issued after EPA implemented guidance in 1999 requiring quantification of
the risks of short- term exposure to pesticide residues in tobacco smoke.
33 However, none of the human health risk assessments or other

documentation we reviewed contained this information* that is, the margin
of exposure estimate* because the health risk assessments supporting these
decisions were completed before the policy was implemented. For pesticides
with many uses and much data, several years may elapse between the initial
scientific assessment of the tobacco use and the issuance of the
reregistration decision.

Not including the 13 pesticides mentioned above, we reviewed five
additional health risk assessments EPA prepared after it developed the
policy requiring the quantification of the risks of short- term exposure
to pesticide residues in tobacco smoke that did include estimates of
margin of exposure. EPA generally does not have concerns about adverse
health effects when a margin of exposure is greater than 100* that is,
when the pesticide causes no adverse effects at levels 100 or more times
greater than the expected actual exposure to the pesticide. Consequently,
a margin of exposure greater than 100 is considered to reflect risk that
is below EPA*s level of concern. As table 4 shows, EPA*s recent health
risk assessments of five pesticides approved for use on tobacco* four of
which were newly registered and one reregistered* generally indicated that
the margins of exposure were substantially greater than 100. Although one

margin of exposure was below 100, EPA officials told us that because they
used very conservative assumptions to estimate exposure, resulting in an
extreme overstatement of actual exposure, EPA was not concerned about the
potential for adverse health effects. For these five pesticides, EPA

concluded that no mitigation related to the use on tobacco was required.
Overall, EPA officials said that potential risks from exposure to residues
on tobacco had never been high enough to require mitigation.

33 The other 6 registration decisions were issued prior to the 1999
guidance.

Page 26 GAO- 03- 485 Pesticides on Tobacco

Table 4: Margins of Exposure for Five Pesticides Approved for Use on
Tobacco Pesticide Margin of exposure

(male) Margin of exposure

(female)

Actigard 518,518 444,444 Carbaryl 104 89 Dimethomorph 1,400 1,200
Pymetrozine 3,333 2,857 Thiamethoxam 3,500 3,000 Source: GAO and EPA.
Note: GAO*s analysis of EPA risk assessment documents. EPA requires that
pesticide manufacturers provide most of the studies it

considers in assessing the health risks of pesticides, and the agency*s
evaluations of these studies are critical to the assessment process. EPA
officials were generally able to provide us with copies of the studies and
evaluations we requested, but documentation of the agency*s evaluation of
the quality of the residue studies and other data upon which it relied to
evaluate the potential for adverse health effects was inconsistent.
Specifically, for eight of the pesticides, EPA officials were unable to
provide their evaluations of the validity and reliability of residue data
used in their assessments of potential health risks. In addition, for
chlorpyrifos, EPA officials were unable to provide the residue studies and
agency evaluations of them from the early 1980s. As a result, we examined
subsequent EPA evaluations that referred to the results of these early
studies and the agency*s conclusion that the residues were below the level
of concern. According to EPA officials, they were unable to locate the
documents, in part, because not all records from this time have yet been
converted to electronic format, and the paper copies could not be located

among the substantial backlog of paper documents. EPA officials noted that
each pesticide registration could consist of 100 or more studies from
pesticide manufacturers, each of which requires one or more agency
evaluations. The officials reported that, as resources permit, contract
and

agency staff are converting documents to electronic format to make them
more readily available for review. While EPA is required to regulate
residues of pesticides approved for use on human food and animal feed
crops, no such requirement applies to pesticides approved for use on
tobacco. However, primarily as a matter of trade equity, USDA does (1)
regulate residues of selected pesticides that are prohibited in the United
States but that may be used on imported tobacco and (2) test certain types
of imported and domestic tobacco to EPA Evaluations of Studies Are Not
Always Available

Federal Regulation of Pesticide Residues on Tobacco Is Limited

Page 27 GAO- 03- 485 Pesticides on Tobacco

ensure they do not exceed residue limits. USDA has not reevaluated the
pesticides it regulates since 1989, although changes in the pesticides
used on tobacco have occurred since then. Through its testing programs,
USDA has found that a small fraction of imported and domestic tobacco
exceeds the residue limits.

As discussed previously, EPA regulates pesticides in the United States by
granting registrations, which permit the distribution, sale, and use of
the pesticides according to directions identified on the label. EPA also
regulates the residues of pesticides that are approved for use on human
food and animal feed crops by setting tolerances* maximum concentrations
of residues that may remain on crops. FDA and USDA test

food and feed crops to ensure that residue levels do not exceed the
tolerances EPA has set. Because tobacco is not used as food or feed,
however, EPA does not set tolerances for residues of pesticides approved
for use on tobacco, 34 and FDA and USDA do not test tobacco for maximum
concentrations of residues of approved pesticides. Consequently, residues

of pesticides approved for use on tobacco in the United States are not
federally regulated.

Instead, federal regulation of pesticide residues on tobacco focuses
exclusively on pesticides not approved for use on tobacco. The Dairy and
Tobacco Adjustment Act of 1983, as amended, requires USDA to

(1) establish maximum allowable concentrations for residues of selected
pesticides that are not approved for use on tobacco in the United States
but that are likely used on tobacco in some other countries and (2) test
imported and domestic flue- cured and burley tobacco to ensure the residue
levels do not exceed the maximum levels allowed. 35 In selecting which
pesticide residues to regulate, USDA is to consider pesticides

whose use on tobacco has been cancelled, suspended, revoked, or otherwise
prohibited under FIFRA. The regulation helps ensure that domestic tobacco
producers are not placed at an unfair disadvantage in the market because
they are not allowed to use certain pesticides that may

34 In addition, EPA does not require validation by an independent
laboratory of the analytic method used to measure pesticide residues on
nonfood crops as it does for methods used to measure pesticide residues on
food and feed. The purpose of the external validation is to support
enforcement of tolerances by ensuring that competent analysts can apply
the method used.

35 7 U. S. C. 511r. Federal Regulation

Focuses on Pesticides Not Approved for Use on Tobacco

Page 28 GAO- 03- 485 Pesticides on Tobacco

be used in other countries; it also helps protect the public from exposure
to the residues of highly toxic pesticides not approved for use on tobacco
in the United States.

While the focus of U. S. regulation of pesticide residues on tobacco is on
those pesticides not approved for use on tobacco, some other countries
have set limits on residues of pesticides that are used on tobacco.
Further, as in the United States, some countries limit the concentration
of residues as measured on tobacco leaf. However, at least one country*
Germany* limits the pesticide residues as measured in cigarettes and other
tobacco products. Appendix III provides information on the limits
established by

Germany, Italy, and Spain. USDA has implemented the Dairy and Tobacco
Adjustment Act, in part, by setting 15 residue limits (maximum allowable
concentrations) covering 20 pesticides currently not approved for use on
tobacco in the United States that the agency believed were used in other
countries. Most of the

pesticides USDA regulates, such as DDT and toxaphene, are organochlorine
pesticides. As discussed earlier, organochlorine pesticides persist in the
environment and accumulate in the bodies of humans and animals, and many
are highly toxic* a number of them have been banned for these reasons.
Eleven of the 15 residue limits apply to individual pesticides and 4 apply
to 2 or more pesticides in combination. 36 For example, aldrin and
dieldrin are summed because dieldrin is the primary degradation product of
aldrin. Table 5 lists the residue limits included in USDA*s testing
program, with the 12 organochlorine pesticides highlighted. As indicated
in the table, methoxychlor is the only organochlorine pesticide included
in USDA*s testing program that is currently approved for other uses in the
United States, such as on food crops.

36 USDA*s method of measuring residues is consistent with EPA*s and FDA*s.
USDA Has Not Reevaluated the Pesticides

it Regulates Since 1989

Page 29 GAO- 03- 485 Pesticides on Tobacco

Table 5: USDA*s Residue Limits for Pesticides on Tobacco Pesticide
(organochlorine pesticides in bold)

Residue limit (parts per

million) Approved for

nontobacco use( s) 1. Chlordane 3.0 No 2. Dibromochloropropane (DBCP) 1.0
No 3. Dicamba 5.0 Yes

4. Endrin 0.1 No 5. Ethylene dibromide (EDB) 0.1 No 6. Formothion 0.5 No

7. Hexachlorobenzene (HCB) 0.1 No

8. Methoxychlor 0.1 Yes

9. Toxaphene 0.3 No 10. 2,4- D 5.0 Yes 11. 2,4,5- T 0.1 Yes

12. Sum of aldrin and dieldrin 0.1 No 13. Sum of cypermethrin and
permethrin 3.0 Yes

14. Sum of DDT, TDE, and DDE 0.4 No

15. Sum of heptachlor and heptachlor epoxide 0.1 No Source: 7 CFR 29. 427,
USDA, and EPA. Note: GAO*s analysis of EPA data. USDA*s Agricultural
Marketing Service (AMS) initially established

maximum allowable concentrations of pesticides in August 1986 after
determining the countries from which the United States imports tobacco,
the pesticides that might reasonably be expected to be used on tobacco in
those countries, and the pesticides not approved for use in the United

States. In 1989, AMS revised the number of pesticides to its current total
of 20 residues.

Although in 1986 USDA stated its intent to periodically reevaluate the
pesticides it regulates, the department has not done so since 1989.
According to officials at USDA, reevaluating the regulated pesticide
residues has not been a priority of the department. However, since USDA
selected the pesticides it would test in 1989, tobacco uses have been
cancelled for more than 30 pesticides that had been approved for use on
tobacco. 37 For example, by 2000, EPA had cancelled all tobacco uses of
lindane* a highly persistent, organochlorine pesticide that may cause

37 Not including the cancelled pesticides, about 100 pesticides had
approval for tobacco uses in the 1990s.

Page 30 GAO- 03- 485 Pesticides on Tobacco

cancer and harm the environment. USDA does not currently regulate
pesticide residues of lindane because it was still approved for tobacco
when USDA last reevaluated the regulated pesticides. Other pesticides,
such as trichlorfon and diazinon, are also candidates for regulation* that
is, pesticides no longer approved for use on tobacco in the United States
but likely to be used in some other countries. As appendix III shows, some

countries that set limits for pesticides used on tobacco have established
them for trichlorfon and diazinon* one of the leading causes of acute
insecticide poisoning for humans. However, because USDA has not revised
the regulated pesticide residues it tests for, the department*s testing
program may not include some pesticides with characteristics similar to
those of pesticides currently included in the testing program and that may
still be used in other countries. Tobacco and pesticide experts with whom
we spoke agreed that periodic reevaluations of the regulated pesticides
would be appropriate. Furthermore, two of these experts* a toxicologist
who has measured residues on tobacco for many years and a former
government official who now represents tobacco producers* told

us that many of the pesticides USDA currently regulates, particularly the
organochlorine pesticides, warrant continued inclusion in the testing
program because they are persistent in the environment, accumulate in the
body, and continue to be used on crops overseas.

Also as required by the Dairy and Tobacco Adjustment Act, USDA tests
certain imported and domestic tobacco to ensure that residues do not
exceed the maximum allowable concentrations the agency established. USDA
is required to test samples of two types of tobacco* flue- cured and
burley* that are commonly imported from other countries and also produced
in the United States to determine whether they conform to the pesticide
residue limits. These two types of tobacco are the major components of
cigarettes, and imports of them have continued to increase over time. For
example, USDA reported that imports of flue- cured tobacco represented
about 12 percent of the flue- cured tobacco used in the United States in
1980 and about 36 percent in 2001. 38 USDA is not required to test other
types of imported tobacco, such as oriental tobacco, which is added to
cigarettes for purposes of flavor and aroma but which is not grown in the
United States.

38 USDA Economic Research Service U. S. Tobacco Import Update (TBS- 2002-
02), Feb. 2003. USDA Tests Imported and

Domestic Tobacco for Regulated Pesticides

Page 31 GAO- 03- 485 Pesticides on Tobacco

Tobacco is imported into the United States in large, sealed shipping
containers that hold approximately 40,000 pounds of tobacco in 90 to 96
boxes weighing about 440 pounds each. In 1986, AMS began testing imported
flue- cured and burley tobacco, which represented about 60 percent of the
tobacco imported into the United States in 2001. Random samples of
imported flue- cured and burley tobacco are tested for residues

of the 20 regulated pesticides. AMS inspectors use a computer program to
randomly select one box of tobacco from each shipping container. The
domestic testing program began in 1989 and is administered by the USDA
Farm Service Agency (FSA) 39 under a cooperative agreement with AMS.
Similar to the AMS program for tobacco imports, FSA tests randomly
selected samples of domestic flue- cured and burley tobacco for the 20
regulated pesticides not approved for use in the United States. FSA tests
the portion of domestically grown flue- cured and burley tobacco that
becomes loan stock (surplus tobacco) under USDA*s tobacco price support
program. 40 The proportion of domestic tobacco that becomes loan stock
varies each year, depending on tobacco quality and demand from
manufacturers, and has declined in recent years. Additional information

about the domestic loan stock program is provided in appendix IV. For 1999
through 2001, USDA*s testing programs found less than 1 percent of
domestically produced or imported flue- cured and burley tobacco with
residue levels above the allowable levels. 41 According to agency
officials,

those results are consistent with results obtained since testing began in
1986. More specifically, for 1999 through 2001, the FSA domestic testing
program found a small fraction of a percentage of domestically produced
tobacco in excess of the limits. FSA found 4 samples of flue- cured
tobacco and 24 samples of burley tobacco* representing more than 12,000
pounds of tobacco* that exceeded the maximum allowable concentrations of 2
of the regulated pesticides* methoxychlor and permethrin. AMS found
residues of DDT/ TDE/ DDE, cypermethrin, and ethylene dibromide in excess
of the limits on less than 1 percent of the imported tobacco entering the
United States during this time.

39 FSA was formed in 1994 from programs in several agencies, including
tobacco programs from the former Agricultural Stabilization and
Conservation Service. 40 The price support program is administered by
stabilization cooperatives* owned by

tobacco growers* under agreement with USDA*s Commodity Credit Corporation
and auction warehouses. 41 Imported tobacco percentage calculated by
weight of imported flue- cured and burley tobacco; domestic percentage
calculated by weight of flue- cured and burley loan stock.

Page 32 GAO- 03- 485 Pesticides on Tobacco

If imported tobacco exceeds any of the limits, the importer is notified of
the violation and may choose to appeal the result or reexport the tobacco
to another country. When an importer appeals, AMS inspectors randomly
select three additional samples for testing, and the residue levels for
the four samples are averaged. If the average result is below the limits,
the tobacco is cleared for entry into the United States. However, if the
average exceeds the limits, the container of tobacco is denied entry and
is typically reexported. Under the Dairy and Tobacco Adjustment Act,
domestically

produced flue- cured or burley tobacco not meeting the residue
requirements must be destroyed. According to USDA officials, because of
restrictions on the disposition of products contaminated by pesticides,
boxes of domestic tobacco are typically disposed of in an approved
landfill with a permit from EPA. To ensure that pesticides can be used
without posing an unreasonable risk

to human health, EPA conducts risk assessments of exposures to the
pesticides it evaluates for use in the United States, including exposure
to pesticide residues on tobacco. EPA*s decision to limit its quantitative
assessment of the risks associated with pesticides on tobacco to the
effects of short- term exposure, and not include the long- term exposure
of smokers, recognizes that the pesticides are used on a crop that itself
poses very significant health risks to humans through use in various
consumer products* primarily cigarettes. Overall, EPA*s health risk
assessments show that the pesticides used on tobacco and other crops are
probably a greater hazard for those who handle them than for those who
inhale

tobacco smoke. Nonetheless, while the risks of some exposures, such as
acute poisoning, are clear, less is known with certainty about the effects
of long- term exposure to small amounts of pesticides, such as residues in
food and water, on tobacco, or in the environment.

While historically EPA has required pesticide manufacturers to provide
data on the residues remaining on tobacco, its assessments of the health
effects associated with exposure to the residues were not identified in
risk assessment documents and generally were not quantified. Mirroring the
improvements in risk assessment methods in recent years, EPA has adopted a
more formal and consistent approach to evaluating the health risks
associated with pesticides used on tobacco and has started to

document, in its risk assessment documents, its conclusions on the
potential for short- term risks from pesticide residues that may remain in
tobacco smoke. As a result, interested parties are better informed about
the potential risks, and EPA is appropriately more accountable for its
assessments. Conclusions

Page 33 GAO- 03- 485 Pesticides on Tobacco

When used as intended* most commonly in cigarettes* tobacco is generally
inhaled into the body. However, because it is not a food, tobacco is
regulated as a nonfood crop with regard to pesticide residues. That is, no
residue limits are established or monitored for pesticides approved for
use on tobacco, as is done for foods. While the regulation of pesticide
residues on tobacco is limited because it does not include pesticides
approved for use on this crop, USDA tests tobacco for residues of 20
pesticides not approved for domestic use on tobacco, primarily for
purposes of trade equity. Because many of the tested pesticides are known
to harm humans and the environment, the USDA testing program helps
minimize the public*s exposure to some highly toxic pesticides. The
universe of pesticides not approved for use on tobacco has grown since
USDA selected the pesticides it tests, but USDA has not reevaluated the
program*s coverage in 14 years. The USDA testing program would be improved
by assessing the current universe of pesticides not approved for use on
tobacco and determining whether an update to its program is warranted.

To better protect the public from exposures to residues of pesticides not
approved for use on tobacco in the United States and ensure that domestic
tobacco producers are not placed at an unfair disadvantage relative to
producers in other countries, we recommend that the Secretary of the
Department of Agriculture direct the Administrators of the Agricultural
Marketing Service and the Farm Service Agency to periodically review and
update the pesticides for which they set residue limits and test imported
and domestic tobacco.

We provided copies of our draft report to EPA and USDA for review and
comment. In commenting on the draft, EPA officials said we accurately
characterized the agency*s risk assessment process for pesticides used on

tobacco, and USDA officials agreed with our recommendation to periodically
review and update the pesticides for which the department sets residue
limits and tests tobacco. USDA officials said they plan to annually review
and update the testing program for tobacco. We conducted our review from
May 2002 through March 2003 in

accordance with generally accepted government auditing standards. Our
scope and methodology are discussed in appendix I. Recommendation for

Executive Action Agency Comments

Page 34 GAO- 03- 485 Pesticides on Tobacco

As agreed with your office, unless you publicly announce the contents of
this report earlier, we plan no further distribution until 30 days from
the report date. At that time we will send copies of this report to the
Administrator, EPA; the Secretary of Agriculture; and other interested
parties. We will make copies available to others upon request. In
addition, the report will be available at no charge on GAO*s Web site at
http:// www. gao. gov.

If you or your staff have any questions, please call me at (202) 512-
3841. Key contributors to this report are listed in appendix V. Sincerely
yours,

John B. Stephenson Director, Natural Resources and Environment

Appendix I: Objectives, Scope, and Methodology

Page 35 GAO- 03- 485 Pesticides on Tobacco

This report provides information on (1) the pesticides commonly used on
tobacco and the potential health risks associated with them; (2) how the
Environmental Protection Agency (EPA) assesses and mitigates health risks
associated with pesticides used on tobacco; and (3) how, and the extent to
which, EPA, U. S. Department of Agriculture (USDA), and other federal
agencies regulate and monitor pesticide residues on tobacco. In addition,
this report provides information on the regulatory residue limits adopted
by three countries that are significant importers of tobacco grown in the
United States.

To identify the chemicals commonly used on tobacco, we reviewed pesticide-
use databases developed by the National Center for Food and Agricultural
Policy (NCFAP), a nonprofit research organization, under a cooperative
agreement with USDA. These databases summarized national use of 235
pesticides on 87 food and nonfood crops for the period 1990 though 1998.
These databases, compiled from more than 130 federal and state surveys and
reports, include pesticide use on cropland in the coterminous 48 states
and do not include new pesticides approved by EPA since 1997. We also
reviewed data from national surveys conducted in the 1990s by the U. S.
Geological Survey and USDA and from several tobaccoproducing states*
Kentucky, North Carolina, Tennessee, and Virginia. In total, 53 pesticides
were identified as being used on tobacco in one or more of the surveys. Of
these, 37 were identified in one or more of the surveys that included
national data, and we refer to this latter group as the pesticides that
were commonly used on tobacco in the United States during the 1990s. To
identify the adverse health effects associated with these 37 pesticides,
we collected and reviewed relevant human health risk assessments prepared
by EPA*s Office of Pesticide Programs. Where such

assessments were not available, we reviewed documents from academic
experts who maintain a database on pesticides and other toxic chemicals
for USDA, and other programs within EPA. In addition we interviewed,

and reviewed reports prepared or recommended by, experts on the human
health effects of exposure to pesticides and other toxins at the National
Cancer Institute, the National Institute of Environmental Health Sciences,
the National Center for Environmental Health, Johns Hopkins Bloomberg

School of Public Health, and the Institute for Cancer Prevention (formerly
the American Health Foundation). To determine how EPA assesses and
mitigates potential health risks from pesticide residues on tobacco, we
reviewed agency policies and procedures on identifying the levels of
pesticide residues on tobacco and the related health risks, and we
interviewed EPA officials in the Office of Pesticide Programs who perform
these tasks. In addition, we examined in Appendix I: Objectives, Scope,
and Methodology

Appendix I: Objectives, Scope, and Methodology

Page 36 GAO- 03- 485 Pesticides on Tobacco

detail how EPA implemented its policies and procedures for 13 of the 37
pesticides commonly used on tobacco. That is, we reviewed pesticide
residue studies submitted to EPA and EPA*s evaluations of pesticide
residues and their potential health effects conducted as part of the
pesticide registration process under the Federal Insecticide, Fungicide,
and Rodenticide Act for 1,3- D, acephate, chlorpyrifos, diazinon,
disulfoton,

endosulfan, ethoprop, ethephon, maleic hydrazide, metalaxyl, methidathion,
pebulate, and pendimethalin. We focused primarily on those pesticides for
which EPA had completed registrations between 1994 and 2002. We did not
independently evaluate the validity or scientific merit of the studies
that EPA relied upon to reach its conclusions.

To determine the extent to which EPA, USDA, and other federal agencies
regulate and monitor pesticide residues on tobacco, we met with cognizant
officials and reviewed authorizing legislation, regulations, and
documentation on how programs related to pesticide residues on tobacco are
implemented. In addition, we analyzed USDA data on tobacco

production, imports, and residue- testing results. We also interviewed
academic and tobacco industry experts and reviewed residue data collected
by North Carolina State University.

To provide information on other countries that have adopted regulatory
limits on pesticide residues, we reviewed articles by academic experts on
the international regulation of pesticides on tobacco. We provide
information on three major importers of U. S. tobacco* Germany, Italy, and
Spain* as examples of regulatory approaches in other countries, focusing
on the residue limits they set. We did not examine how, or the extent to
which, these countries monitor or enforce their pesticide residue limits.
We updated and clarified the information on the three countries* residue
limits provided in the articles with information from the Cooperation
Centre for Scientific Research Relative to Tobacco (CORESTA), an
international tobacco research organization, and officials responsible for
oversight of pesticides and tobacco in Germany and Spain.

To identify countries that import U. S. tobacco, we extracted data from
the United States International Trade Commission*s interactive tariff and
trade database on the countries that received U. S. flue- cured and burley
tobacco from 1996 through 2001. We conducted our review from May 2002
through March 2003 in

accordance with generally accepted government auditing standards.

Appendix II: Pesticide Use on Tobacco and Other Crops

Page 37 GAO- 03- 485 Pesticides on Tobacco

Most of the pesticides used on tobacco are widely used on food and other
crops. As shown in table 6, tobacco use of most pesticides represents a
small portion of the total use. However, for some pesticides*
dimethomorph, fenamiphos, flumetralin, maleic hydrazide, mefenoxam, and
sulfentrazone* most of the use in 1994 through 1998 was on tobacco. Table
6: Pesticide Use on Tobacco and All Crops, 1990- 98

Use in pounds* 1992 survey (1990- 93) Use in pounds* 1997 survey (1994-
98) Pesticide Tobacco All crops Tobacco All crops

1,3- dichloropropene (1,3- D) 11,537,540 40,083,611 13,279,285 34,717,237
Chloropicrin 577,082 11,086,567 6,761,644 13,882,188 Maleic hydrazide
1,789,208 2,073,238 1,790,089 2,143,154 Acephate 1,570,457 3,389,865
871,899 2,462,354 Methyl bromide 5,356,748 44,196,554 685,026 32,803,943
Pendimethalin 321,931 20,281,766 473,718 27,284,718 Chlorpyrifos 685,554
14,764,535 406,822 13,463,879 Fenamiphos 257,142 614,937 379,841 726,675
Mancozeb a 8,062,374 356,811 9,585,777 Flumetralin a a 352,742 352,742
Metalaxyl 371,645 855,400 271,368 659,997 Clomazone a 1,801,776 217,617
2,531,160 Ethoprop 438,274 1,449,743 182,321 1,010,807 Endosulfan a
1,796,726 172,766 1,601,195 Mefenoxam a a 139,199 210,101 Pebulate 412,000
673,046 131,665 343,322 Ethephon 113,238 2,701,284 102,130 5,407,986
Napropamide 191,840 500,695 92,622 448,400 Sulfentrazone a a 69,073 69,073
Imidacloprid a a 67,896 272,207 Aldicarb 159,044 4,022,468 59,719
4,277,552 Dimethomorph a a 36,818 51,536 Methomyl 57,137 2,754,907 29,773
1,997,489 Malathion 7,549 3,377,678 15,437 5,809,943 Disulfoton 52,578
1,806,527 13,495 1,196,066 Sethoxydim a 1,350,566 9,579 1,717,271 Spinosad
a a 2,815 117,315 Carbaryl 16,487 4,570,414 2,057 4,857,542 Fonofos 12,798
3,233,797 16 417,372 Benefin 56,963 478,205 a 161,983

Bacillus thuringiensis b b b b Carbofuran 149,965 5,101,406 a 3,398,067
Diazinon 53,670 1,265,739 a 918,087 Diphenamid 81,624 105,009 a a Appendix
II: Pesticide Use on Tobacco and Other Crops

Appendix II: Pesticide Use on Tobacco and Other Crops

Page 38 GAO- 03- 485 Pesticides on Tobacco

Use in pounds* 1992 survey (1990- 93) Use in pounds* 1997 survey (1994-
98) Pesticide Tobacco All crops Tobacco All crops

Isopropalin 129,287 129,287 a a Methidathion 68 372,953 a 314,091
Trichlorfon 722 13,974 a a Grand total 24,400,552 182,915,047 26,974,241
175,211,229

Source: National Center for Food and Agricultural Policy. a Not identified
as being used in survey.

b Neither survey estimated the use of Bacillus thuringiensis in pounds.

Appendix III: Germany, Italy, and Spain Have Adopted Regulatory Limits for
Pesticide Residues on Tobacco Page 39 GAO- 03- 485 Pesticides on Tobacco

Several countries that are major importers of U. S. tobacco have adopted
regulations for specific pesticide residues on various forms of tobacco.
For example, Germany*s residue limits (maximum residue levels) apply to
finished products, such as cigarettes, whereas limits in Italy and Spain
generally apply to tobacco leaf. Although they have somewhat different
regulatory approaches to pesticides on tobacco, Germany, Italy, and Spain
differ from the United States in that they regulate residues of pesticides

approved for use on tobacco in addition to regulating some residues of
pesticides not approved for use on tobacco.

According to 2003 data from CORESTA* the Cooperation Centre for Scientific
Research Relative to Tobacco* Germany, Italy, and Spain have residue
limits on tobacco for 79, 100, and 58 pesticides, respectively. 1 Of the
37 pesticides commonly used on tobacco in the United States during the
1990s, Germany has limits for 20, Italy for 24, and Spain for 21 (see
table 7). None of these countries have adopted limits for 7 of the
pesticides commonly used on U. S. tobacco during the 1990s. 2 Table 7:
Residue Limits Adopted by Germany, Italy, and Spain for Pesticides

Commonly Used on Tobacco in the United States during the 1990s Residue
limits in ppm (country) Pesticide Germany a Italy b Spain c Acephate d 1.5
d

Aldicarb 10.0 0.6 (green) 3.0 (cured) 5.0 Benefin d 0.01 0.02 Carbaryl 3.0
3.0 0.1 Carbofuran 20.0 0.1 10.0 Chlorpyrifos d 0.2 0.05 Diazinon 1.0 d
0.02 Dichloropropene d d 0.05 Diphenamid 1.5 0.1 5.0 Disulfoton 1.0 0.4 d
Endosulfan 20.0 1.0 d

1 CORESTA is an international research association whose members are
companies and research institutes having research and development
activities related to tobacco. It has 190 members from 52 countries.

2 The seven pesticides are Bacillus thuringiensis, chloropicrin,
clomazone, dimethomorph, mefanoxam, spinosad, and sulfentrazone. Appendix
III: Germany, Italy, and Spain Have

Adopted Regulatory Limits for Pesticide Residues on Tobacco

Appendix III: Germany, Italy, and Spain Have Adopted Regulatory Limits for
Pesticide Residues on Tobacco Page 40 GAO- 03- 485 Pesticides on Tobacco

Residue limits in ppm (country) Pesticide Germany a Italy b Spain c
Ethephon

d 16.0 (green) 80.0 (cured) d

Ethoprop 3.0 0.02 0.02 Fenamiphos 15.0 0.1 0.02 Flumetralin

20.0 2.0 (green) 10.0 (cured) 5.0

Fonophos 1.0 0.05 d Imidacloprid

d 10.0 (green) 50.0 (cured) 5.0

Isopropalin 0.5 0.02 d Malathion 3.0 0.5 0.5 Maleic hydrazide 80.0 80.0 a
80.0 Mancozeb

50.0 e 2.0 (green) 10.0 f (cured) 0.05 g Metalaxyl d 1.0 3.0

Methidathion 1.0 d d Methomyl 2.0 d d Methyl bromide d d 20.0 Napropamide
0.1 0.1 0.05 Pebulate 0.5 d 0.05 Pendimethalin d 0.05 0.05 Sethoxydim d
0.5 d Trichlorfon 1.0 0.1 0.1

Number of limits for commonly used pesticides 20 24 21

Source: CORESTA and European Court of Auditors. Note: GAO*s analysis of
CORESTA and European Court of Auditors data. a Residue limit on finished
products.

b Residue limit on green tobacco if not otherwise specified. c Residue
limit on dried tobacco. d Country has not adopted limits for this
pesticide. e Residue limit established for the entire class of
dithiocarbamates except metam. f Residue limit established for the entire
class of dithiocarbamates except thiram.

g Residue limit established for the entire class of dithiocarbamates
except metam and thiram.

In addition to residue limits for approved pesticides, Germany and Italy
collectively have residue limits on tobacco that apply to 15 of the 20
pesticides not approved for use in the United States that USDA monitors in

its tobacco testing programs. The 15 pesticides are aldrin, dieldrin,
chlordane, cypermethrin, DDT, DDE, endrin, ethylene dibromide, formothion,
heptachlor, heptachlor epoxide, hexachlorobenzene,

Appendix III: Germany, Italy, and Spain Have Adopted Regulatory Limits for
Pesticide Residues on Tobacco Page 41 GAO- 03- 485 Pesticides on Tobacco

methoxychlor, permethrin, and TDE. Further, where no specific pesticide
limits are set for tobacco products in Germany, residues of pesticides not
approved for use on tobacco in Germany may be present in amounts that are
not likely to pose a risk to human health. In Italy and Spain, residues of
pesticides not approved for use on tobacco in those countries must not
exceed the limit of detection, generally between 0.01 ppm and 0.05 ppm. We
did not examine how, or the extent to which, these countries monitor or
enforce their pesticide residue limits.

From 1971 to 2000, researchers at the North Carolina State University
(NCSU) collected limited data on the residues of various pesticides on
some domestically grown tobacco. NCSU data for the 1990s included six
pesticides for which Germany, Italy, or Spain have residue limits. The
domestic tobacco tested by NCSU identified residue levels that were (1)
consistently below the lowest limit for endosulfan, flumetralin, and
metalaxyl; (2) generally above the limit for maleic hydrazide; 3 and (3)
more varied for fenamiphos and the dithiocarbamates* a class of fungicides
that includes mancozeb. For example, in 1995 residue levels on flue- cured
tobacco were below the lowest limit for fenamiphos* 0.02 ppm adopted

by Spain* but exceeded this limit in 1992 and 1994. Also, in 1991 and 1997
residue levels of dithiocarbamates were generally lower on burley tobacco
than limits in Germany and Italy* 50 ppm and 10 ppm, respectively* but
exceeded Spain*s limit of 0.05 ppm.

3 According to tobacco experts with whom we spoke, tobacco with high
levels of maleic hydrazide may be blended with tobacco from other sources
to reduce overall maleic hydrazide levels.

Appendix IV: USDA Tests Domestic Tobacco in the Loan Stock Program Page 42
GAO- 03- 485 Pesticides on Tobacco

The USDA Farm Service Agency (FSA) 1 tests the portion of domestically
grown flue- cured and burley tobacco that becomes loan stock (surplus
tobacco) under USDA*s tobacco price support program for the 20 regulated
pesticides. 2 To receive price supports, tobacco must be sold in USDA-
approved auction warehouses and inspected by USDA graders. At the auction
warehouse, each individual lot of tobacco is sold to the highest bidder.
If the highest bid is below the government*s loan (support) price, or no
bid is received, the stabilization cooperative makes loans to growers
whose tobacco does not bring the minimum price at auction with funds
borrowed from USDA*s Commodity Credit Corporation. The growers* tobacco,
which is consigned to the cooperative as loan stock, is pledged as
collateral to the credit corporation for the money borrowed. The
cooperative receives, processes, stores, and later sells the loan stock
tobacco when demand increases, with the proceeds used to repay the credit
corporation loan, plus interest. An alternative to traditional auction
marketing* growers contracting to sell their tobacco directly to

manufacturers* also reduces the amount of tobacco going to auction and
thus potentially to loan stock. For the most recently completed marketing
season* growing year 2001* 20 percent of domestic tobacco was sold at
auction, and 2.4 percent became loan stock.

After auction, the tobacco is processed in distinct *runs* of
approximately 100,000 pounds, when the tobacco is stemmed, redried, finely
chopped, and placed into boxes holding approximately 440 pounds. The
tobacco cooperative randomly selects one box from each run and draws a
one- pound sample of tobacco for pesticide testing at USDA*s laboratory.
If the sample exceeds any of the residue limits, the box of tobacco from

which it came is destroyed. The adjacent boxes, processed before and after
the original box, are also sampled. The testing continues with adjacent
boxes of tobacco until the samples are found to be below the residue
limits. Because the samples are drawn by the tobacco cooperatives, FSA
resamples 5 percent of the tested inventory (or 25 samples, whichever is
less) for oversight purposes each year. 1 FSA was formed in 1994 from
programs in several agencies, including tobacco programs

from the former Agricultural Stabilization and Conservation Service. 2 The
price support program is administered by stabilization cooperatives* owned
by tobacco growers* under agreement with USDA*s Commodity Credit
Corporation and auction warehouses. Appendix IV: USDA Tests Domestic
Tobacco

in the Loan Stock Program

Appendix IV: USDA Tests Domestic Tobacco in the Loan Stock Program Page 43
GAO- 03- 485 Pesticides on Tobacco

Historically a substantial portion of domestic tobacco was sold at
auctions in conjunction with the tobacco price support program, but in
recent years most domestic tobacco has been sold under contract directly
to cigarette manufacturers* approximately 80 percent in 2001. Officials
from USDA and tobacco associations told us the market has changed because
manufacturers asserted that auction markets were not providing quality
tobacco with the characteristics they required. The recent, dramatic shift
in the way tobacco is marketed* with a 60 to 80 percent reduction in the
amount of tobacco at auction* has decreased the amount of domestic

tobacco that potentially becomes loan stock and thus is tested. Although
the amount of domestically produced tobacco that becomes loan stock has
varied greatly, an average of 13 percent became loan stock over the past
decade. In 2001, only about 2 percent of domestically produced tobacco has
become loan stock, reducing the amount of domestic tobacco subjected to
pesticide testing. The officials with whom we spoke said that this change
is not likely to be reversed.

Appendix V: GAO Contacts and Staff Acknowledgments Page 44 GAO- 03- 485
Pesticides on Tobacco

John B. Stephenson, (202) 512- 3841 Christine Fishkin, (202) 512- 6895

In addition to those named above, Nancy Crothers, Laura Gatz, Terrance
Horner, Richard Johnson, Ilga Semeiks, Tina Smith, and Cheryl Williams
made key contributions to this report. Appendix V: GAO Contacts and Staff

Acknowledgments GAO Contacts Acknowledgments

(360204)

The General Accounting Office, the audit, evaluation and investigative arm
of Congress, exists to support Congress in meeting its constitutional
responsibilities and to help improve the performance and accountability of
the federal government for the American people. GAO examines the use of
public funds; evaluates federal programs and policies; and provides
analyses, recommendations, and other assistance to help Congress make
informed oversight, policy, and funding decisions. GAO*s commitment to
good government is reflected in its core values of accountability,
integrity, and reliability.

The fastest and easiest way to obtain copies of GAO documents at no cost
is through the Internet. GAO*s Web site (www. gao. gov) contains abstracts
and fulltext files of current reports and testimony and an expanding
archive of older products. The Web site features a search engine to help
you locate documents using key words and phrases. You can print these
documents in their entirety, including charts and other graphics.

Each day, GAO issues a list of newly released reports, testimony, and
correspondence. GAO posts this list, known as *Today*s Reports,* on its
Web site daily. The list contains links to the full- text document files.
To have GAO e- mail

this list to you every afternoon, go to www. gao. gov and select
*Subscribe to daily E- mail alert for newly released products* under the
GAO Reports heading.

The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to: U. S. General Accounting Office 441 G Street NW, Room LM
Washington, D. C. 20548 To order by Phone: Voice: (202) 512- 6000

TDD: (202) 512- 2537 Fax: (202) 512- 6061

Contact: Web site: www. gao. gov/ fraudnet/ fraudnet. htm E- mail:
fraudnet@ gao. gov Automated answering system: (800) 424- 5454 or (202)
512- 7470 Jeff Nelligan, managing director, NelliganJ@ gao. gov (202) 512-
4800

U. S. General Accounting Office, 441 G Street NW, Room 7149 Washington, D.
C. 20548 GAO*s Mission Obtaining Copies of

GAO Reports and Testimony

Order by Mail or Phone To Report Fraud, Waste, and Abuse in Federal
Programs Public Affairs
*** End of document. ***