This is the accessible text file for GAO report number GAO-08-944 
entitled 'Concentrated Animal Feeding Operations: EPA Needs More 
Information and a Clearly Defined Strategy to Protect Air and Water 
Quality from Pollutants of Concern' which was released on September 24, 
2008.

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as part 
of a longer term project to improve GAO products' accessibility. Every 
attempt has been made to maintain the structural and data integrity of 
the original printed product. Accessibility features, such as text 
descriptions of tables, consecutively numbered footnotes placed at the 
end of the file, and the text of agency comment letters, are provided 
but may not exactly duplicate the presentation or format of the printed 
version. The portable document format (PDF) file is an exact electronic 
replica of the printed version. We welcome your feedback. Please E-mail 
your comments regarding the contents or accessibility features of this 
document to Webmaster@gao.gov. 

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. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately. 

Report to Congressional Requesters: 

United States Government Accountability Office: 
GAO: 

September 2008: 

Concentrated Animal feeding operations: 

EPA Needs More Information and a Clearly Defined Strategy to Protect 
Air and Water Quality from Pollutants of Concern: 

GAO-08-944: 

GAO Highlights: 

Highlights of GAO-08-944, a report to congressional requesters. 

Why GAO Did This Study: 

Concentrated Animal Feeding Operations (CAFO) are large livestock and 
poultry operations that raise animals in a confined situation. CAFOs 
can improve the efficiency of animal production but large amounts of 
manure produce can, if not properly managed, degrade air and water 
quality. The Environmental Protection Agency (EPA) is responsible for 
regulating CAFOs and requires CAFOs that discharge certain pollutants 
to obtain a permit. 

This report discusses the (1) trends in CAFOs over the past 30 years, 
(2) amounts of waste they generate, (3) findings of key research on 
CAFOs’ health and environmental impacts, (4) EPA’s progress in 
developing CAFO air emissions protocols, and (5) effect of recent court 
decisions on EPA’s regulation of CAFO water pollutants. GAO analyzed 
U.S. Department of Agriculture’s (USDA) data from 1982 through 2002, 
for large farms as a proxy for CAFOs; reviewed studies, EPA documents, 
laws, and regulations; and obtained the views of federal and state 
officials. 

What GAO Found: 

Because no federal agency collects consistent, reliable data on CAFOs, 
GAO could not determine the trends in these operations over the past 30 
years. However, using USDA data for large farms that raise animals as a 
proxy for CAFOs, it appears that the number of these operations 
increased by about 230 percent, going from about 3,600 in 1982 to 
almost 12,000 in 2002. Also, during this 20-year period the number of 
animals per farm had increased, although it varied by animal type. 
Moreover, GAO found that EPA does not have comprehensive, accurate 
information on the number of permitted CAFOs nationwide. As a result, 
EPA does not have the information it needs to effectively regulate 
these CAFOs. EPA is currently working with the states to establish a 
new national data system. 

The amount of manure generated by large farms that raise animals 
depends on the type and number of animals raised, but large operations 
can produce more than 1.6 million tons of manure a year. Some large 
farms that raise animals can generate more raw waste than the 
populations of some U.S. cities produce annually. In addition, 
according to some agricultural experts, the clustering of large 
operations in certain geographic areas may result in large amounts of 
manure that cannot be effectively used as fertilizer on adjacent 
cropland and could increase the potential of pollutants reaching nearby 
waters and degrading water quality. 

Since 2002, at least 68 government-sponsored or peer-reviewed studies 
have been completed that examined air and water quality issues 
associated with animal feeding operations and 15 have directly linked 
air and water pollutants from animal waste to specific health or 
environmental impacts. EPA has not yet assessed the extent to which 
these pollutants may be impairing human health and the environment 
because it lacks key data on the amount of pollutants that are being 
emitted from animal feeding operations. 

As a first step in developing air emissions protocols for animal 
feeding operations, in 2007, a 2-year nationwide air emissions 
monitoring study, largely funded by industry, was initiated. However, 
as currently structured, the study may not provide the scientific and 
statistically valid data it was intended to provide and that EPA needs 
to develop air emissions protocols. Furthermore, EPA has not 
established a strategy or timetable for developing a more sophisticated 
process-based model that considers the interaction and implications of 
all emission sources at an animal feeding operation. 

Two recent federal court decisions have affected EPA’s ability to 
regulate water pollutants discharged by CAFOs. The 2005 Waterkeeper 
case required EPA to abandon the approach that it had proposed in 2003 
for regulating CAFO water discharges. Similarly, the 2006 Rapanos case 
has complicated EPA’s enforcement of CAFO discharges because EPA 
believes that it must now gather significantly more evidence to 
establish which waters are subject to the Clean Water Act’s permitting 
requirements. 

What GAO Recommends: 

To more effectively regulate CAFOs, GAO recommends that EPA complete 
its inventory of permitted CAFOs, reassess the current nationwide air 
emissions monitoring study, and establish a strategy and timetable for 
developing a process-based model for measuring CAFO air emissions. EPA 
partially agreed with GAO’s recommendations. 

To view the full product, including the scope and methodology, click on 
[hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-08-944]. For more 
information, contact Anu Mittal (202) 512-3841, mittala@gao.gov. 

[End of section] 

Contents: 

Letter: 

Results in Brief: 

Background: 

The Number of Large Farms Raising Animals Has Increased, but Specific 
Data on CAFOs Are Not Available: 

Large Farms That Raise Animals Can Produce Thousands of Tons of Manure 
Each Year, and Regional Clustering of Farms Can Exacerbate Manure 
Management Problems: 

Studies Have Identified Impacts of Pollutants from Animal Waste, but 
EPA Has Not Assessed the Extent of Such Impacts: 

It Is Unclear If EPA's Efforts to Develop Air Emissions Protocols for 
Animal Feeding Operations Will Be Effective and How EPA Intends to 
Regulate These Emissions in the Future: 

Two Federal Court Decisions Have Affected EPA's and Some States' 
Ability to Regulate Water Pollutants Discharged by CAFOs: 

Conclusions: 

Recommendations for Executive Action: 

Agency Comments and Our Evaluation: 

Appendix I: Objectives, Scope, and Methodology: 

Appendix II: EPA's Definition of Concentrated Animal Feeding 
Operations: 

Appendix III: Nationwide Trends in the Number of All Animal Farms and 
the Number of Animals Raised on Large Farms, 1982-2002: 

Appendix IV: Government-Sponsored or Peer-Reviewed Studies Completed 
Since 2002 on the Impacts of Pollutants from Animal Waste: 

Appendix V: Comments from the Environmental Protection Agency: 

Appendix VI: GAO Contact and Staff Acknowledgments: 

Tables: 

Table 1: Nationwide Trends in the Number of Large Farms Raising Animals 
for All Animal Types, 1982 through 2002: 

Table 2: Median Number of Animals Raised on Large Farms, by Animal 
Type, 1982 through 2002: 

Table 3: Nationwide Trends in the Number of Animals Raised on Large 
Farms as a Proportion of the Number of Animals Raised on All Farms, by 
Animal Type, 1982 and 2002: 

Table 4: Estimated Typical Manure Production for Three Different Sizes 
of Large Farms That Raise Animals, 2002: 

Table 5: Studies Completed Since 2002 Linking Water Pollutants from 
Animal Feeding Operations with Impacts on Human Health or the 
Environment: 

Table 6: Studies Completed Since 2002 Directly Linking Air Pollutants 
from Animal Feeding Operations to Impacts on Human Health: 

Table 7: Studies Completed Since 2002 Finding No Links between 
Pollutants from Animal Feeding Operations and Impacts on Human Health 
or the Environment: 

Table 8: Studies Completed Since 2002 That Found an Indirect Link 
between Pollutants from Animal Feeding Operations and Human Health or 
Environmental Impacts: 

Table 9: Number of Participants in the Air Compliance Agreement, 
Funding Provided by Animal Type, and Source of the Funding for the 
National Air Emissions Monitoring Study: 

Table 10: States That Reported Having Regulations for Air Emissions 
from Animal Feeding Operations, 2008: 

Table 11: State Officials' Views of the Impact of the Waterkeeper 
Decision on Their CAFO Programs: 

Table 12: EPA Designation of Large, Medium, and Small CAFOs for Various 
Size Thresholds by Animal Type: 

Table 13: Nationwide Trends in the Number of All Farms That Raise 
Animals for All Animal Types, 1982 through 2002: 

Table 14: Nationwide Trends in the Number of Animals Raised on Large 
Farms per Day for All Animal Types, 1982 through 2002: 

Figures: 

Figure 1: Geographic Concentration of Hogs in Five Contiguous North 
Carolina Counties, 2002: 

Figure 2: EPA Timeline for Development of Air Emission Protocols for 
Animal Feeding Operations: 

Abbreviations: 

CAFO: Concentrated Animal Feeding Operation: 

CERCLA: Comprehensive Environmental Response, Compensation and 
Liability Act of 1980: 

EPA: Environmental Protection Agency: 

EPCRA: Emergency Planning and Community Right-to-Know Act of 1986: 

NAS: National Academy of Sciences: 

NPDES: National Pollutant Discharge Elimination System: 

USDA: U.S. Department of Agriculture: 

[End of section] 

United States Government Accountability Office:
Washington, DC 20548: 

September 4, 2008: 

Congressional Requesters: 

Over the last 40 years, diversified, independent, family-owned-and- 
operated farms that produce a variety of crops and a few animals are 
becoming a smaller share of the agricultural sector and are being 
replaced by fewer, much larger farms. For animal production, this 
change has meant a movement to significantly larger operations that can 
raise, for example, as many as 2 million chickens or 800,000 hogs at 
one facility at one time. 

These large-scale livestock and poultry operations are generally 
referred to as animal feeding operations. An animal feeding operation 
is one that (1) raises animals in a confined situation for a total of 
45 days or more during a 12-month period and (2) brings feed to the 
animals rather than having the animals graze or seek feed in pastures 
and fields or on rangeland. Concentrated animal feeding operations 
(CAFO) are a subset of animal feeding operations and usually operate on 
a much larger scale. Generally, a CAFO is an animal feeding operation 
that raises enough animals to meet or exceed certain minimum 
thresholds, depending upon the type of livestock being raised. For 
example, as defined in Clean Water Act regulations, an animal feeding 
operation would be considered a CAFO if it raised 1,000 or more beef 
cattle, 2,500 hogs weighing more than 55 pounds, or 125,000 broiler 
chickens.[Footnote 1] In addition, an animal feeding operation of any 
size can be designated a CAFO if it meets certain conditions, such as 
being a significant contributor of pollutants to federally regulated 
waters.[Footnote 2] 

While CAFOs may have improved the efficiency of the animal production 
industry, they have also raised environmental and health concerns 
because of the large amounts of manure they produce. Generally, to 
minimize potential environmental problems, these operations retain the 
manure that they produce in storage facilities onsite and periodically 
dispose of it by spreading it on nearby or adjacent cropland as 
fertilizer. If the manure is properly contained and managed, it can 
benefit crop production; if improperly contained and managed, it can 
degrade air and water quality, thereby potentially impairing human 
health and damaging the environment. Specifically, these operations can 
potentially degrade air quality because large amounts of manure may 
emit unsafe quantities of ammonia, hydrogen sulfide, and particulate 
matter,[Footnote 3] and they can potentially degrade water quality 
because pollutants in manure such as nitrogen, phosphorus, bacteria, 
and organic matter could enter nearby water bodies. 

Several federal laws provide the Environmental Protection Agency (EPA) 
with the authority to regulate water and air pollutants from CAFOs. The 
Clean Water Act specifically addresses CAFOs by requiring EPA to 
consider CAFOs like any other industry if they discharge pollutants 
into federally regulated waters. As a result, CAFOs that have such 
discharges must obtain a permit that establishes design standards and 
management practices for retaining and disposing of manure in such a 
way as to limit the amounts and types of pollutants from manure that 
are released into federally regulated waters. EPA, or the states that 
have been authorized by EPA to administer the Clean Water Act, are 
responsible for issuing these permits. In contrast, three other acts 
provide EPA with certain authorities related to air emissions from 
these operations, although they do not specifically cite CAFOs as 
regulated entities. Under the Clean Air Act, any animal feeding 
operation, regardless of size, that exceeds established air emission 
thresholds for certain pollutants can be regulated. For example, 
pollutants such as particulate matter that are emitted by animal 
feeding operations are regulated under the Clean Air Act and other 
pollutants such as hydrogen sulfide or ammonia may be regulated under 
the act in certain circumstances. Similarly, the Comprehensive 
Environmental Response, Compensation, and Liability Act of 1980 
(CERCLA) and the Emergency Planning and Community Right-to-Know Act of 
1986 (EPCRA) do not specifically mention CAFOs, but they do require 
owners or operators of these facilities to report to federal or state 
and local authorities when a "reportable quantity" of certain hazardous 
substances, such as hydrogen sulfide or ammonia,[Footnote 4] is 
released into the environment. Together, CERCLA's and EPCRA's reporting 
requirements provide government authorities, emergency management 
agencies, and citizens the ability to know about the source and 
magnitude of hazardous releases into the environment. 

In light of the growing concerns regarding the potential human health 
and environmental impacts of CAFOs, you asked us to determine the (1) 
trends in CAFOs over the past 30 years; (2) amount of waste they 
generate; (3) findings of recent key academic, industry, and government 
research on the impacts of CAFOs on human health and the environment, 
and the extent to which EPA has assessed the nature and severity of 
such impacts; (4) progress that EPA and the states have made in 
regulating and controlling the emissions of, and in developing 
protocols to measure, air pollutants from CAFOs that could affect air 
quality; and (5) extent to which recent court decisions have affected 
EPA and the states' ability to regulate CAFO discharges that impair 
water quality. 

In conducting our work, we reviewed laws and regulations and federal 
and state agencies' documents; met with officials from EPA and the U.S. 
Department of Agriculture (USDA), industry, citizen and environmental 
groups, and academia. We also spoke with state officials and visited 
CAFOs in eight states. These states were Arkansas, California, 
Colorado, Iowa, Maryland, Minnesota, North Carolina, and Texas. We 
chose these states because they were geographically dispersed and 
contained numerous CAFOs representing various animal types. In 
addition, to determine trends in CAFOs over the past 30 years, from 
1974 through 2002, we obtained the most recent data available from USDA 
on large farms that raise animals to use as a proxy for CAFO data. 
However, because of limitations in USDA's data for 1974 through 1982, 
we could not determine from these data which farms prior to 1982 would 
meet EPA's minimum size thresholds for CAFOs. Consequently, our 
analysis of trends in CAFOs focuses on the 20-year period between 1982 
and 2002. We also obtained and reviewed the data that EPA compiled over 
the last 5 years from each of its regions on the number of CAFOs that 
were issued a permit. To identify the amount of waste CAFOs generate, 
we estimated the amounts of manure generated by various size farms that 
raise animals. To provide a perspective of the amount of waste that 
large farms that raise animals can generate, we selected certain cities 
based on their population and estimated the amount of sanitary waste 
generated by the human population and compared these amounts with the 
amount of waste generated by three different sizes of large farms 
[Footnote 5]. To report on key research on the impacts of CAFOs on 
human health and the environment, we reviewed EPA's 2003 Rule 
regulating discharges from CAFOs under the Clean Water Act and the 
National Academy of Sciences study on air emissions from animal feeding 
operation[Footnote 6]s. We also conducted library and Internet searches 
to identify key studies completed since 2002 on air and water 
pollutants from waste generated by animal feeding operations. We 
compared the findings from these studies with EPA assessments to date 
and interviewed EPA officials regarding these assessments. To assess 
the progress that EPA and the states have made in regulating and 
controlling the air emissions of, and in developing protocols to 
measure, air pollutants from CAFOs, we reviewed relevant documents and 
interviewed EPA officials, as well as officials responsible for an 
ongoing national air emissions monitoring study. In addition, we 
contacted state officials in all 50 states to determine which states 
had developed air emission regulations applicable to CAFOs. Finally, to 
determine the extent to which recent court decisions have affected EPA 
and the states' ability to regulate CAFO discharges that impair water 
quality, we reviewed the results of recent federal and state court 
decisions. We also interviewed EPA and state officials on how the court 
decisions have affected their ability to regulate CAFOs. A more 
detailed description of our scope and methodology is presented in 
appendix I. 

We conducted this performance audit between July 2007 and August 2008, 
in accordance with generally accepted government auditing standards. 
These standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe that 
the evidence obtained provides a reasonable basis for our findings and 
conclusions based on our audit objectives. 

Results in Brief: 

No federal agency collects accurate and consistent data on the number, 
size, and location of CAFOs. However, according to USDA officials, the 
data USDA collects for large farms that raise animals can serve as a 
proxy in estimating trends in CAFOs nationwide from 1982 through 2002. 
Using these data, we found that the number of large farms that raise 
animals has increased 234 percent, from about 3,600 in 1982 to almost 
12,000 in 2002. We found that the number of animals raised on these 
large farms had also increased, but the rate of increase varied greatly 
by animal type. For example, the average number of hogs raised on large 
farms increased by 37 percent, from about 3,400 in 1982 to nearly 4,600 
in 2002. In contrast, during the same time period, the average number 
of broiler chickens raised on large farms only increased by about 3 
percent, from approximately 155,000 to nearly 160,000. Furthermore, 
almost half of the livestock and poultry raised in the United States in 
2002, about 43 percent, were raised on large farms. Over the last 5 
years, EPA has been compiling data from its regions in an effort to 
develop information on the number of permitted CAFOs nationwide. 
However, we determined that these data are inconsistent and inaccurate 
and do not provide necessary information on the characteristics of 
these CAFOs. Without a systematic and coordinated process for 
collecting and maintaining accurate and complete information on the 
number, size, and location of permitted CAFOs, EPA does not have the 
information that it needs to effectively regulate these operations. EPA 
has indicated that it is working with the states to develop and 
implement a new national system to collect and maintain these data. 

The amount of manure that a large farm raising animals can generate 
depends on the types and numbers of animals being raised at a specific 
operation; such farms can produce from over 2,800 tons to more than 1.6 
million tons of manure annually. For example, a layer farm that meets 
EPA's minimum large CAFO threshold of 82,000 laying hens could produce 
more than 2,800 tons of manure a year, while a farm with 10,000 beef 
cattle (cattle fattened with feed) could produce about 117,000 tons of 
manure a year. In fact, some large farms can produce more raw waste 
than the human population of a large U.S. city. For example, a very 
large hog farm, with as many as 800,000 hogs, generates more than 1.6 
million tons of manure annually--more than one and a half times the 
sanitary waste produced by the about 1.5 million residents of 
Philadelphia, Pennsylvania in 1 year. Furthermore, while manure is a 
valuable resource often used as fertilizer, agricultural experts and 
government officials have raised concerns about the large amounts of 
manure produced by animal feeding operations that are increasingly 
clustered within specific geographic areas within a state. For example, 
five contiguous North Carolina counties had an estimated hog population 
of over 7.5 million hogs in 2002 and the hog operations in these 
counties could have produced as much as 15.5 million tons of manure 
that year. According to agricultural experts and government officials 
that we spoke to, such clustering of operations raises concerns that 
the amount of manure produced could result in the overapplication of 
manure to croplands in these areas and the release of excessive levels 
of some pollutants that could potentially damage water quality. 

At least 68 government-sponsored or peer-reviewed studies have been 
completed on air and water quality issues associated with animal waste 
since 2002 and 15 of these studies have directly linked pollutants from 
animal waste to specific health or environmental impacts. Of the 
remaining 53 studies, 7 found no impacts, 12 made indirect linkages 
between these pollutants and health and environmental impacts, and 34 
of the studies focused on measuring the amount of water or air 
pollutants emitted by animal feeding operations. However, EPA has not 
yet assessed the extent to which air and water pollution from CAFOs may 
be impairing human health and the environment because it lacks key data 
on the amount of pollutants that CAFOs are discharging. Of the 15 
studies we found directly linking pollutants from animal waste to human 
health or environmental impacts, 8 focused on water pollutants and 7 on 
air pollutants. Most of the water studies found that nutrients or 
hormones released from animal feeding operations were causing 
environmental harm, such as reproductive disorders in fish and degraded 
water quality. One water study found that animal feeding operations 
were causing pathogens such as E. coli to contaminate drinking water, 
which were then causing gastrointestinal illnesses in humans. 
Similarly, all seven air studies linked air emissions from animal 
feeding operations to adverse human health effects. Specifically, six 
found exposure to these emissions caused respiratory inflammation and 
one found an increased incidence of headaches, eye irritation, and 
nausea in people working at or living near these operations. According 
to EPA officials, although the agency has long recognized the potential 
impacts that water pollutants from CAFOs can have on human health and 
the environment, it has not yet assessed these impacts because it lacks 
information on the extent to which water pollutants are actually being 
discharged by CAFOs. According to other officials at EPA, the agency 
does not have the resources needed to conduct a study that would 
provide this information. Likewise, EPA has not yet assessed the air 
quality impacts from animal feeding operation emissions because, 
according to agency officials, it lacks key data on the extent to which 
these operations are emitting pollutants. To gather this information, 
EPA entered into a series of agreements with animal feeding operations 
to implement a national air emissions monitoring study that is 
currently ongoing and is being funded by the industry and will measure 
and quantify air emissions from animal feeding operations. 

The ongoing national air emissions monitoring study is considered a 
first step in EPA's efforts to develop protocols for measuring and 
quantifying air contaminants from animal feeding operations; however, 
it is not clear if the study will provide EPA the data that it needs to 
develop these protocols. EPA believes that this 2-year study, initiated 
in 2007, will provide a scientific basis for estimating air emissions 
from animal feeding operations so that the agency can develop protocols 
that these operations can use to more quickly determine if they exceed 
regulatory thresholds. However, concerns have been raised that the 
animal feeding operations being monitored in the study do not represent 
a valid sample of all animal feeding operations and that the data 
collected during the early phases of the study may be incomplete. As a 
result, it is uncertain whether the study will ultimately provide data 
of sufficient quantity and quality that will enable the agency to 
develop its planned protocols. In addition, it is uncertain if and when 
EPA will develop a process-based model that considers the interaction 
and implications of all sources of emissions at an animal feeding 
operation. Furthermore, other EPA actions make it unclear at this time 
how the agency intends to regulate air emissions from animal feeding 
operations once the data collection effort is complete. For example, 
EPA has not yet decided if it will aggregate the emissions occurring on 
an animal feeding operation or if the emissions from barns and manure 
storage areas will be considered separately when determining if an 
operation has exceeded air emissions thresholds. Moreover, in December 
2007, EPA proposed a rule to exempt releases to the air of hazardous 
substances, such as ammonia and hydrogen sulfide, from manure at farms, 
including animal feeding operations, which meet or exceed their 
reportable quantity from both CERCLA and EPCRA notification 
requirements. EPA stated that, in all instances, the source and nature 
of the release make emergency responses unnecessary, impractical, and 
unlikely for these operations, and hence it found notifications to be 
unnecessary. It is unclear to us how EPA made this determination when 
it has not yet completed its data collection effort and does not yet 
know the extent to which animal feeding operations are emitting these 
pollutants. In the absence of federal guidance on how to regulate air 
emissions from animal feeding operations, officials in six states told 
us that they are regulating some emissions covered under the Clean Air 
Act, CERCLA, and EPCRA. For example, Minnesota has established state 
emissions thresholds for hydrogen sulfide that apply to CAFOs and the 
state requires CAFO operators to develop an air emissions plan 
specifying how they will control these emissions. 

Two recent federal court decisions have affected EPA's and some states' 
ability to regulate CAFOs for pollutants that may impair water quality. 
Specifically: 

* In 2005, in Waterkeeper Alliance Inc. v. EPA (Waterkeeper), the U.S. 
Court of Appeals for the Second Circuit set aside key provisions of a 
CAFO rule EPA had issued in 2003. This rule would have provided EPA 
with comprehensive information on the universe of CAFOs and their 
operations and would have subjected large numbers of previously 
unregulated CAFOs to monitoring and reporting requirements, as well as 
periodic inspections. However, the court concluded that EPA did not 
have the authority under the Clean Water Act to require CAFOs that were 
not discharging pollutants into federally regulated waters to apply for 
permits. As a result, CAFO operators currently determine for themselves 
whether they need to apply for a federal permit, and EPA must rely on 
other means of acquiring information about CAFOs that are illegally 
discharging pollutants, such as through citizens' reports. EPA has 
developed proposed revisions to its 2003 rule in response to the 
court's ruling. The resulting rule is currently awaiting the Office of 
Management and Budget's approval, but EPA is not certain when that 
review will be completed and the final rule issued. The Waterkeeper 
decision has had mixed impacts on states' regulation of CAFOs. Some 
states have not been affected by the Waterkeeper decision because they 
have used their own authorities to adopt regulations more stringent 
than federal regulations. As a result, these states, such as Minnesota, 
have continued to require all CAFOs to obtain state permits. In 
contrast, officials in those states, such as Colorado, that base their 
regulations on the Clean Water Act and federal regulations told us that 
their programs will remain in limbo until EPA issues its final revised 
rule. 

* The Supreme Court's 2006 decision--Rapanos v. United States 
(Rapanos)--has also complicated EPA's enforcement of CAFO regulations. 
The Court's decision has raised questions that have not yet been 
resolved about which "waters" are considered federal waters and, 
therefore, fall under the jurisdiction of the Clean Water Act. 
According to EPA enforcement officials, the agency may be less likely 
to seek enforcement against a CAFO that it believes is discharging 
pollutants into a water body because it is now more difficult to prove 
that the water body is federally regulated. Congress is considering 
legislation that seeks to clearly define the scope of the Clean Water 
Act and resolve the questions raised by the Rapanos decision. 

To more effectively regulate CAFOs, we are recommending that the 
Administrator of EPA direct the agency to complete its efforts to 
develop a comprehensive national inventory of permitted CAFOs that 
incorporates appropriate internal controls to ensure the quality of the 
data collected. To ensure that the national air emissions monitoring 
study will provide the scientific and statistically valid data that EPA 
needs for developing its air emissions protocols, we are recommending 
that EPA reassess the current data collection efforts, including its 
internal controls. We are also recommending that EPA establish a 
strategy and timetable for developing a process-based model that will 
provide more sophisticated air emissions estimating methodologies for 
animal feeding operations. In commenting on a draft of this report, EPA 
partially agreed with our recommendations. 

Background: 

The livestock and poultry industry is vital to our nation's economy, 
supplying meat, milk, eggs, and other animal products; however, the 
past several decades have seen substantial changes in America's animal 
production industries. As a result of domestic and export market 
forces, technological changes, and industry adaptations, food animal 
production that was integrated with crop production has given way to 
fewer, larger farms that raise animals in confined situations. These 
large-scale animal production facilities are generally referred to as 
animal feeding operations. CAFOs are a subset of animal feeding 
operations and generally operate on a larger scale. While CAFOs may 
have improved the efficiency of the animal production industry, their 
increased size and the large amounts of manure they generate have 
resulted in concerns about the management of animal waste and the 
potential impacts this waste can have on environmental quality and 
public health. 

Animal manure can be, and frequently is, used beneficially on farms to 
fertilize crops and to restore nutrients to soil. However, if 
improperly managed, manure and wastewater from animal feeding 
operations can adversely impact water quality through surface runoff 
and erosion, direct discharges to surface water, spills and other dry- 
weather discharges, and leaching into the soil and groundwater. Excess 
nutrients in water can result in or contribute to low levels of oxygen 
in the water and toxic algae blooms, which can be harmful to aquatic 
life. Improperly managed manure can also result in emissions to the air 
of particles and gases, such as ammonia, hydrogen sulfide, and volatile 
organic compounds, which may also result in a number of potentially 
harmful environmental and human health effects. 

Most agricultural activities are considered to be nonpoint sources of 
pollution because the pollution that occurs from these activities is in 
conjunction with soil erosion caused by water and surface runoff of 
rainfall or snowmelt from diffuse areas such as farms and rangeland. 
However, section 502(14) of the Clean Water Act specifically defines 
point sources of pollution to include CAFOs, which means that under the 
act, CAFOs that discharge into federally regulated waters are required 
to obtain a federal permit called a National Pollutant Discharge 
Elimination System (NPDES) permit. These permits generally allow a 
point source to discharge specified pollutants into federally regulated 
waters under specific limits and conditions. These permits are issued 
by EPA or a state agency authorized by EPA to implement the NPDES 
program for that state. Currently, 45 states are authorized to 
administer the NPDES permit program, and their programs must be at 
least as stringent as the federal program.[Footnote 7] In 1976, in 
accordance with the Clean Water Act's designation of CAFOs as point 
sources, EPA defined which poultry and livestock facilities constituted 
a CAFO and established permitting regulations for CAFOs. According to 
EPA regulations issued in 1976, to be considered a CAFO a facility must 
first be considered an animal feeding operation. Animal feeding 
operations are agricultural operations where the following conditions 
are met: 

* animals are fed or maintained in a confined situation for a total of 
45 days or more in any 12-month period, and: 

* crops, vegetation, forage growth, or post harvest residues are not 
sustained during normal growing seasons over any portion of the lot. 

If an animal feeding operation met EPA's criteria and either met or 
exceeded minimum size thresholds based on the type of animals being 
raised, EPA considered the operation to be a CAFO. For example, an 
animal feeding operation would be considered a CAFO if it raised 1,000 
or more beef cattle, 2,500 pigs weighing more than 55 pounds, or 
125,000 chickens. In addition, EPA could designate an animal feeding 
operation of any size as a CAFO under certain circumstances. For 
example, if an animal feeding operation was a significant contributor 
of pollutants to federally regulated water, EPA could designate the 
operation as a CAFO. Appendix II lists the full text of EPA's current 
CAFO definition, including the size thresholds established for small, 
medium, and large CAFOs. 

Under EPA's 1976 CAFO regulations, certain animal feeding operations 
did not require permits. These included (1) those animal feeding 
operations that only discharged during a 25-year, 24-hour storm event-
-which is the amount of rainfall during a 24-hour period that occurs on 
average once every 25 years or more and (2) chicken operations that use 
dry manure-handling systems--systems that do not use water to handle 
their waste. In addition, EPA generally did not regulate animal waste 
that was applied to cropland or pastureland. 

In January 2003, we reported that although EPA believed that many 
animal feeding operations degrade water quality, it had placed little 
emphasis on its permit program and that exemptions in its regulations 
allowed as many as 60 percent of the largest operations to avoid 
obtaining permits.[Footnote 8] In its response to our 2003 report, EPA 
acknowledged that the CAFO program was hampered by outdated regulations 
and incomplete attention by EPA and the states. EPA pointed out that it 
had revised its permitting regulations for CAFOs to eliminate the 
exemptions that allowed most animal feeding operations to avoid 
regulation. The revisions, issued in February 2003 and known as the 
2003 CAFO rule, resulted, in part, from the settlement of a 1989 
lawsuit by the Natural Resources Defense Council and Public Citizen, in 
which these groups alleged that EPA had failed to comply with the Clean 
Water Act. EPA's 2003 CAFO rule included the following key provisions: 

* Duty to apply. All CAFOs were required to apply for an NPDES permit 
unless the permitting authority determined that the CAFO had no 
potential to discharge to federally regulated waters. 

* Expanded CAFO definitions to include all poultry operations and stand-
alone operations raising immature animals. The previous rule had 
applied only to poultry operations that used a liquid manure-handling 
system. The 2003 rule expanded the CAFO definition to all types of 
poultry operations, and EPA officials estimated that this revision 
could result in almost 2,200 additional poultry operations requiring a 
permit. 

* More stringent design standard for new facilities in the swine, 
poultry, and veal categories. Under the previous rule, facilities were 
to be designed, constructed, and operated to contain runoff from a 25- 
year, 24-hour rainfall event; this continues to be the rule for 
existing facilities. For new facilities, the 2003 rule established a no-
discharge standard that can be met if the facilities are designed, 
constructed, and operated to contain the runoff from a 100-year, 24- 
hour storm event. 

* Best management practices. Operations would be required to implement 
best management practices for applying manure to cropland and for 
animal production areas. The rule required, among other things, 
specified setbacks from streams, vegetated buffers, depth markers in 
lagoons, and other impoundments for production areas to prevent or 
reduce pollution from the operation. 

* Nutrient management plans. CAFO operations would be required to 
develop a plan for managing the nutrient content of animal manure as 
well as the wastewater resulting from CAFO operations, such as water 
used to flush manure from barns. 

* Compliance schedule. The 2003 rule required newly defined CAFOs to 
apply for permits by April 2006 and existing CAFOs to develop and 
implement nutrient management plans by December 31, 2006.[Footnote 9] 

According to EPA officials, the 2003 rule was expected to ultimately 
lead to better water quality because the revised regulations would 
extend coverage to more animal feeding operations that could 
potentially discharge and contaminate water bodies and subject these 
operations to periodic inspections. 

Three laws provide EPA with certain authorities related to air 
emissions from animal feeding operations: the Clean Air Act,[Footnote 
10] the Comprehensive Environmental Response, Compensation, and 
Liability Act of 1980 (CERCLA), and the Emergency Planning and 
Community Right-to-Know Act of 1986 (EPCRA).[Footnote 11] Although 
these laws provide EPA with authority related to air emissions from 
various sources, they do not expressly identify animal feeding 
operations as a regulated entity. Specifically: 

* The Clean Air Act authorizes EPA to regulate stationary and mobile 
sources of air pollution and emphasizes controlling sources that emit 
more than threshold quantities of regulated pollutants. Livestock 
producers and other agricultural sources whose emissions meet or exceed 
specific statutory or regulatory thresholds are therefore subject to 
Clean Air Act requirements. Although EPA has authorized states and 
local governments to carry out certain portions of the act, EPA retains 
concurrent enforcement authority. 

* Taken together, CERCLA and EPCRA require owners or operators of a 
facility to report to federal or state authorities the release of 
hazardous substances that meet or exceed their reportable quantities so 
as to alert federal, state, and local agencies, as well as the public, 
to the release of these substances. Section 103 of CERCLA requires that 
the person in charge of a facility notify the National Response Center 
of any non-permitted release of "hazardous substances" in a reportable 
quantity as soon as he or she has knowledge of that release. Section 
304 of EPCRA requires that the owner or operator of a facility at which 
a hazardous chemical is produced, used, or stored give immediate notice 
of a release of any "extremely hazardous substance" to the community 
emergency coordinator. Among the reportable substances that could be 
released by livestock facilities are hydrogen sulfide and ammonia. The 
reportable quantity for each of these hazardous substances is 100 
pounds in a 24-hour period. Under these acts, EPA can assess civil 
penalties for failure to report releases of hazardous substances or 
extremely hazardous substances that equal or exceed their reportable 
quantities--up to $32,500 per day or $32,500 per violation for first 
time offenders. 

EPA is also working with USDA to address the impacts of animal feeding 
operations on air and water quality and public health. In 1998, EPA 
entered into a memorandum of understanding with USDA that calls for the 
agencies to coordinate on air quality issues relating to agriculture 
and share information. In addition, in 1999, the two agencies issued a 
unified national strategy aimed at having the owners and operators of 
animal feeding operations take actions to minimize water pollution from 
confinement facilities and land application of manure and in 2001 
adopted an agreement to develop a process for working together 
constructively. To help minimize water pollution from animal feeding 
operations and meet EPA's regulatory requirements, USDA, through its 
Natural Resources Conservation Service, provides financial and 
technical assistance to CAFO operators in developing and implementing 
nutrient management plans. 

The Number of Large Farms Raising Animals Has Increased, but Specific 
Data on CAFOs Are Not Available: 

Because no federal agency collects accurate and consistent data on the 
number, size, and location of CAFOs nationwide, it is difficult to 
determine precise trends in CAFOs over the last 30 years. According to 
USDA officials, the data USDA collects for large farms raising animals 
can be used as a proxy for estimating trends in CAFOs nationwide. Using 
these data, we determined that between 1982 and 2002, the number of 
large farms raising animals has increased sharply, from about 3,600 to 
almost 12,000. Moreover, EPA has compiled some data from its regions on 
the number of CAFOs that have been issued permits; however, these data 
are inconsistent and inaccurate. As a result, EPA does not have a 
systematic way of identifying and inspecting all of the CAFOs 
nationwide that have been issued permits. 

Since 1982 the Number of Large Farms Raising Animals Has Increased as 
Has the Average Number of Animals on Farms: 

We found that the number of large farms raising animals for all animal 
types increased by 234 percent between 1982 and 2002. Table 1 shows the 
changes in the number of large farms by animal type for 1982 through 
2002. 

Table 1: Nationwide Trends in the Number of Large Farms Raising Animals 
for All Animal Types, 1982 through 2002: 

Type of animal farm: Beef cattle[A]; 
1982: 966; 
1987: 1,014; 
1992: 1,004; 
1997: 958; 
2002: 982; 
Percentage change, 1982-2002: 2. 

Type of animal farm: Dairy cow; 
1982: 541; 
1987: 712; 
1992: 1,009; 
1997: 1,445; 
2002: 1,939; 
Percentage change, 1982-2002: 258. 

Type of animal farm: Hog[B]; 
1982: 916; 
1987: 1,257; 
1992: 2,061; 
1997: 4,170; 
2002: 5,571; 
Percentage change, 1982-2002: 508. 

Type of animal farm: Layer; 
1982: 720; 
1987: 808; 
1992: 788; 
1997: 788; 
2002: 706; 
Percentage change, 1982-2002: (2). 

Type of animal farm: Broiler; 
1982: 173; 
1987: 357; 
1992: 737; 
1997: 1,331; 
2002: 2,227; 
Percentage change, 1982-2002: 1,187. 

Type of animal farm: Turkey; 
1982: 278; 
1987: 437; 
1992: 504; 
1997: 577; 
2002: 570; 
Percentage change, 1982-2002: 105. 

Type of animal farm: Total of all animal types[C]; 
1982: 3,594; 
1987: 4,585; 
1992: 6,103; 
1997: 9,269; 
2002: 11,995; 
Percentage change, 1982-2002: 234. 

Source: GAO analysis of USDA data. 

Notes: The phrase "all animal types" refers to the following animals: 
beef cattle, dairy cows, hogs, layers, broilers, and turkeys. 

The criteria for a large farm varied by animal type, consistent with 
EPA's CAFO thresholds, and represent the average number of animals on a 
farm per day. 

[A] Beef cattle includes only cattle on feed, not grazing on pasture, 
and sold weighing 500 pounds or more. 

[B] Hogs include swine of all sizes from birth to market size. 

[C] The number of large farms for all animal types is the total of 
large farms for each animal type and may include some farms multiple 
times if they were considered large for more than one animal type. 

[End of table] 

As table 1 shows, large broiler and hog farms experienced the largest 
increase, with large farms raising broilers increasing by 1,187 percent 
and large farms raising hogs increasing by 508 percent. Large farms 
raising layers and large farms raising beef cattle remained relatively 
stable over these 20 years, while layer farms were the only farms that 
experienced an overall decrease in number over the period, declining by 
2 percent. In contrast, while the number of large farms raising animals 
has increased, the number of all farm raising animals has decreased. 
Appendix III presents trends in the number of all farms raising 
animals, from 1982 to 2002. 

Just as the number of large farms for almost all animal types increased 
between 1982 and 2002, so did the size of these farms as illustrated by 
the median number of animals raised on each farm.[Footnote 12] Table 2 
shows the trends in the median number of animals raised on large farms 
for all animal types from 1982 through 2002. 

Table 2: Median Number of Animals Raised on Large Farms, by Animal 
Type, 1982 through 2002: 

Animal type: Beef cattle[A]; 
1982: 2,820; 
1987: 2,950; 
1992: 2,919; 
1997: 3,308; 
2002: 3,424; 
Percentage change, 1982-2002: 21. 

Animal type: Dairy cows; 
1982: 910; 
1987: 988; 
1992: 1,020; 
1997: 1,100; 
2002: 1,200; 
Percentage change, 1982-2002: 32. 

Animal type: Hogs[B]; 
1982: 3,350; 
1987: 3,500; 
1992: 3,778; 
1997: 4,334; 
2002: 4,588; 
Percentage change, 1982-2002: 37. 

Animal type: Layers; 
1982: 131,530; 
1987: 146,383; 
1992: 155,319; 
1997: 168,000; 
2002: 180,000; 
Percentage change, 1982-2002: 37. 

Animal type: Broilers; 
1982: 154,830; 
1987: 168,593; 
1992: 159,840; 
1997: 161,820; 
2002: 159,840; 
Percentage change, 1982-2002: 3. 

Animal type: Turkeys; 
1982: 80,000; 
1987: 79,500; 
1992: 81,000; 
1997: 79,697; 
2002: 80,491; 
Percentage change, 1982-2002: 1. 

Source: GAO's analysis of USDA data. 

Note: We used the median number of animals raised on large farms to 
represent the average concentration of animals raised on large farms 
per day. 

The criteria for a large farm varied by animal type, consistent with 
EPA's CAFO thresholds, and represent the average number of animals on a 
farm per day. 

The median is the point above and below which half of the cases exist. 
For large animal farms, half of the farms of a particular animal type 
have more animals than the median farm and half have fewer animals. For 
example, in the table above, half of large layer farms in 2002 have 
more than 180,000 layers and half have less than 180,000 layers. 

[A] Beef cattle includes only cattle on feed, not grazing on pasture, 
and sold weighing 500 pounds or more. 

[B] Hogs include swine of all sizes from birth to market size. 

[End of table] 

The layer and hog sectors had the largest increases in the median 
number of animals raised per farm, both growing by 37 percent between 
1982 and 2002. Specifically, for layers, large farms increased the 
number of birds they raised from 131,530 in 1982 to 180,000 in 2002 and 
for hogs, large farms increased the number of animals they raised from 
3,350 in 1982 to 4,588 in 2002. In contrast, large farms that raised 
either broilers or turkeys only increased slightly in size with an 
overall increase of 3 and 1 percent, respectively, from 1982 to 2002. 

The increases in the number of large farms for almost all animal types, 
as well as the increases in the median number of animals raised on 
these farms, are also reflected in the percentage of animals raised on 
large farms as compared with animals raised on all farms. Specifically, 
the number of animals raised on large farms increased from over 257 
million in 1982 to over 890 million in 2002--an increase of 246 
percent. In contrast, the number of animals raised on all farms 
increased from over 1,145 million in 1982 to 2,072 million in 2002--an 
increase of 81 percent. This is particularly noteworthy because the 
number of animals raised on large farms only accounted for 22 percent 
of animals raised on all farms in 1982; yet, the number of animals 
raised on large farms accounted for 43 percent of animals raised on all 
farms in 2002. Table 3 shows the trends in the number of animals raised 
on large farms and the number of animals raised on all farms from 1982 
to 2002. 

Table 3: Nationwide Trends in the Number of Animals Raised on Large 
Farms as a Proportion of the Number of Animals Raised on All Farms, by 
Animal Type, 1982 and 2002: 

Animal type: Beef cattle[A];
Number of animals raised on all animal farms: 1982: 11,064,096; 
Number of animals raised on all animal farms: 2002: 11,264,122; 
Percent change, 1982-2002: 2; 
Number of animals raised on large farms: 1982: 6,601,928; 
Number of animals raised on large farms: 2002: 8,677,892; 
Percent change, 1982-2002: 31; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 1982: 60; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 2002: 77. 

Animal type: Dairy cows; 
Number of animals raised on all animal farms: 1982: 10,849,880; 
Number of animals raised on all animal farms: 2002: 9,103,959; 
Percent change, 1982-2002: (16); 
Number of animals raised on large farms: 1982: 632,583; 
Number of animals raised on large farms: 2002: 3,183,086; 
Percent change, 1982-2002: 403; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 1982: 6; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 2002: 35. 

Animal type: Hogs[B]; 
Number of animals raised on all animal farms: 1982: 45,944,318; 
Number of animals raised on all animal farms: 2002: 66,318,763; 
Percent change, 1982-2002: 44; 
Number of animals raised on large farms: 1982: 4,176,477; 
Number of animals raised on large farms: 2002: 47,789,951; 
Percent change, 1982-2002: 1,044; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 1982: 9; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 2002: 72. 

Animal type: Layers; 
Number of animals raised on all animal farms: 1982: 386,638,856; 
Number of animals raised on all animal farms: 2002: 420,742,205; 
Percent change, 1982-2002: 9; 
Number of animals raised on large farms: 1982: 160,005,126; 
Number of animals raised on large farms: 2002: 304,500,225; 
Percent change, 1982-2002: 90; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 1982: 41; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 2002: 72. 

Animal type: Broilers; 
Number of animals raised on all animal farms: 1982: 612,092,410; 
umber of animals raised on all animal farms: 2002: 1,440,501,856; 
Percent change, 1982-2002: 135; 
Number of animals raised on large farms: 1982: 52,140,827; 
Number of animals raised on large farms: 2002: 457,461,691; 
Percent change, 1982-2002: 777; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 1982: 9; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 2002: 32. 

Animal type: Turkeys; 
Number of animals raised on all animal farms: 1982: 78,550,564; 
Number of animals raised on all animal farms: 2002: 124,152,525; 
Percent change, 1982-2002: 58; 
Number of animals raised on large farms: 1982: 33,443,754; 
Number of animals raised on large farms: 2002: 68,417,853; 
Percent change, 1982-2002: 105; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 1982: 43; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 2002: 55. 

Animal type: Total of all animal types[C]; 
Number of animals raised on all animal farms: 1982: 1,145,140,124; 
Number of animals raised on all animal farms: 2002: 2,072,083,430; 
Percent change, 1982-2002: 81; 
Number of animals raised on large farms: 1982: 257,000,695; 
Number of animals raised on large farms: 2002: 890,030,698; 
Percent change, 1982-2002: 246; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 1982: 22; 
The number of animals raised on large farms as a percentage of the 
number of animals raised on all animal farms: 2002: 43. 

Source: GAO analysis of USDA data. 

Note: The phrase "all animal types" refers to the following animals: 
beef cattle, dairy cows, hogs, layers, broilers, and turkeys. 

A farm was included in all farms, for a particular animal type, only if 
it had one or more animals of that type. For example, if a farm had 
broilers only, it would not be counted in all farms for other animal 
types. If a farm raised no animals of any type, then it would also not 
be included in all farms. 

Reported percentages have been rounded to the nearest whole number but 
calculations involving percentages used non-rounded percentages. 

[A] Beef cattle includes only cattle on feed, not grazing on pasture, 
and sold weighing 500 pounds or more. 

[B] Hogs include swine of all sizes from birth to market size. 

[C] The number of large farms for all animal types is the total number 
of large farms for each animal type and may include some farms multiple 
times if they were considered large for more than one animal type. 

[End of table] 

As table 3 shows, most of the beef cattle, hogs, and layers raised in 
the United States in 2002 were raised on large farms. Specifically, 77 
percent of beef cattle and 72 percent of both hogs and layers were 
raised on large farms. 

EPA Does Not Have a Systematic Means of Identifying Permitted CAFOs 
Because It Lacks Accurate Data: 

EPA does not have its own data collection process to determine the 
number, size, and location of CAFOs that have been issued permits 
nationwide. Since 2003, the agency has compiled quarterly estimates 
from its regions on the number of permits that have been issued to 
CAFOs. These data are developed by EPA's regional offices or originates 
with the state permitting authority. However, we determined that these 
data are inconsistent and inaccurate and do not provide EPA with the 
reliable data that it needs to identify and inspect permitted CAFOs 
nationwide. For example, according to EPA some uncertainty in the data 
exists because some states may be using general permits to cover more 
than one operation. In addition, EPA has not established adequate 
internal controls to ensure that the data are correctly reported. For 
example, officials from 17 states told us that data reported by EPA for 
their states were inaccurate. In one case, when we asked a state 
official for the number of CAFOs in his state, the official realized 
that the CAFO numbers reported by EPA's regional office were incorrect 
because of a clerical error, which resulted in some CAFO statistics for 
the state being doubled. After the state official discovered this error 
the state's data were corrected and resubmitted to EPA. Without a 
systematic and coordinated process for collecting and maintaining 
accurate and complete information on the number, size, and location of 
permitted CAFOs nationwide, EPA does not have the information it needs 
to effectively regulate these operations. 

In commenting on a draft of this report, EPA stated that the 
information from permit files is available to EPA upon request; 
however, the information is currently not readily compiled in a 
national database. EPA is currently working with the states to develop 
and implement a new national data system to collect and record 
operation-specific information. As part of this effort, the agency 
plans to develop national requirements for data that should be 
collected and entered into the database by the states. According to 
EPA, it may require the states to provide data that identifies 
operations that have been issued or applied for a CAFO permit as well 
as operations that should have applied for a permit based on an 
inspection or enforcement action. 

Large Farms That Raise Animals Can Produce Thousands of Tons of Manure 
Each Year, and Regional Clustering of Farms Can Exacerbate Manure 
Management Problems: 

The amount of manure a large farm that raises animals can generate 
primarily depends on the types and numbers of animals raised on that 
farm, and the amount of manure produced can range from over 2,800 tons 
to more than 1.6 million tons a year. To further put this in 
perspective, the amount of manure produced by large farms that raise 
animals can exceed the amount of waste produced by some large U.S. 
cities. In addition, multiple large farms that raise animals may be 
located in a relatively small area, such as two or more adjacent 
counties, which raises additional concerns about the potential impacts 
of the manure produced, stored, and disposed of by these farms. 

Table 4 shows the estimated number of animals and the typical amounts 
of manure produced each year, by type of animal, for three different 
sizes of large farms: (1) large farms that meet EPA's thresholds for 
each animal type, (2) large farms that raise the median number of 
animals according to our analysis of USDA farm census data, and (3) 
large farms that fell into the 75th percentile based on our analysis. 
As table 4 shows, a dairy farm that meets the minimum threshold of 700 
dairy cows could produce almost 17,800 tons of manure a year; a median- 
sized dairy farm with 1,200 dairy cows could produce about 30,500 tons 
of manure a year; and a larger dairy farm with 1,900 dairy cows could 
produce almost 48,300 tons of manure a year. 

Table 4: Estimated Typical Manure Production for Three Different Sizes 
of Large Farms That Raise Animals, 2002: 

Animal type: Beef cattle[D]; 
EPA's minimum thresholds for large CAFOs[A]: 1,000; 
Estimated tons of manure produced annually by large CAFOs meeting EPA's 
minimum threshold: 11,690; 
Median number of animals raised on large farms[B]: 3,424; 
Estimated tons of manure produced annually by large farms that raised 
median number of animals: 40,025; 
Number of animals raised on large farms in the 75th percentile[C]: 
10,000; 
Estimated tons of manure produced annually by large farms in the 75th 
percentile: 116,895. 

Animal type: Dairy cows; 
EPA's minimum thresholds for large CAFOs[A]: 700; 
Estimated tons of manure produced annually by large CAFOs meeting EPA's 
minimum threshold: 17,793; 
Median number of animals raised on large farms[B]: 1,200; 
Estimated tons of manure produced annually by large farms that raised 
median number of animals: 30,502; 
Number of animals raised on large farms in the 75th percentile[C]: 
1,900; 
Estimated tons of manure produced annually by large farms in the 75th 
percentile: 48,295. 

Animal type: Hogs[E]; 
EPA's minimum thresholds for large CAFOs[A]: 2,500; 
Estimated tons of manure produced annually by large CAFOs meeting EPA's 
minimum threshold: 5,100; 
Median number of animals raised on large farms[B]: 4,588; 
Estimated tons of manure produced annually by large farms that raised 
median number of animals: 9,360; 
Number of animals raised on large farms in the 75th percentile[C]: 
7,700; 
Estimated tons of manure produced annually by large farms in the 75th 
percentile: 15,708. 

Animal type: Layers; 
EPA's minimum thresholds for large CAFOs[A]: 82,000; 
Estimated tons of manure produced annually by large CAFOs meeting EPA's 
minimum threshold: 2,843; 
Median number of animals raised on large farms[B]: 180,000; 
Estimated tons of manure produced annually by large farms that raised 
median number of animals: 6,242; 
Number of animals raised on large farms in the 75th percentile[C]: 
400,000; 
Estimated tons of manure produced annually by large farms in the 75th 
percentile: 13,870. 

Animal type: Broilers; 
EPA's minimum thresholds for large CAFOs[A]: 125,000; 
Estimated tons of manure produced annually by large CAFOs meeting EPA's 
minimum threshold: 4,125; 
Median number of animals raised on large farms[B]: 159,840; 
Estimated tons of manure produced annually by large farms that raised 
median number of animals: 5,275; 
Number of animals raised on large farms in the 75th percentile[C]: 
195,383; 
Estimated tons of manure produced annually by large farms in the 75th 
percentile: 6,448. 

Animal type: Turkeys; 
EPA's minimum thresholds for large CAFOs[A]: 55,000; 
Estimated tons of manure produced annually by large CAFOs meeting EPA's 
minimum threshold: 3,633; 
Median number of animals raised on large farms[B]: 80,491; 
Estimated tons of manure produced annually by large farms that raised 
median number of animals: 5,317; 
Number of animals raised on large farms in the 75th percentile[C]: 
124,500; 
Estimated tons of manure produced annually by large farms in the 75th 
percentile: 7,719. 

Source: GAO analyses based on EPA CAFO definitions, USDA data, and 
standards for manure production cited by the American Society of 
Agricultural and Biological Engineers, "Manure Production and 
Characteristics," March 2005. 

Note: The amounts of manure reported are estimates. The actual amount 
of manure produced by an animal will vary based on, among other things, 
feeding programs, feeds used, climatic conditions, production 
techniques, and animal genetics. 

EPA reports its minimum thresholds for large CAFOs in terms of 
inventory data for all the animal types included in table 4. To be able 
to compare the annual manure estimates for EPA's thresholds, the 
median, and 75th percentile animal counts, we used USDA data on animal 
sales, inventories, and production cycles, and adjusted these to 
determine typical inventory during a year. 

The criteria for a large farm varied by animal type, consistent with 
EPA's CAFO thresholds, and represent the average number of animals on a 
farm per day. 

[A] This category captures the minimum inventory threshold that an 
animal feeding operation must meet to be designated as a large CAFO by 
EPA and the Clean Water Act. 

[B] This column represents the median-sized animal farm in 2002, for 
each animal type. The median is the point above and below which half of 
the cases exist. For large farms that raise animals, half of the farms 
of a particular animal type have more animals than the median farm and 
half have fewer animals. For example, in table 4, half of large layer 
farms have more than 180,000 layers and half have less than 180,000 
layers. 

[C] This column represents the farms ranked in the 75th percentile for 
the amount of animals raised per farm in 2002, for each animal type. 
The 75th percentile is the point where 25 percent of the cases are 
larger and 75 percent are smaller. For large farms that raise animals, 
the 75th percentile indicates the larger of the large farms. The 75th 
percentile gives a more complete picture of how big a large farm can 
be. For example, for beef cattle the 75th percentile farm is about 3 
times larger than the median-size farm and 25 percent of the beef 
cattle farms are larger than 10,000 cattle. 

[D] Beef cattle includes only cattle on feed, not grazing on pasture, 
and sold weighing 500 pounds or more. The beef cattle manure estimates 
are for cattle fed from about 700 pounds to about 1,200 pounds. 

[E] Hogs include swine of all sizes from birth to market size. The hog 
manure estimates are for hogs fed from about 27 pounds to about 260 
pounds. 

[End of table] 

Additionally, individual large farms that raise animals can generate as 
much waste as certain U.S. cities.[Footnote 13] For example, a dairy 
farm meeting EPA's large CAFO threshold of 700 dairy cows can create 
about 17,800 tons of manure annually, which is more than the about 
16,000 tons of sanitary waste per year generated by the almost 24,000 
residents of Lake Tahoe, California. Likewise, a median-sized beef 
cattle operation with 3,423 head of beef cattle can produce more than 
40,000 tons of manure annually, which is more than the almost 38,900 
tons of sanitary waste per year generated by the nearly 57,000 
residents of Galveston, Texas. Similarly, some larger farms can produce 
more waste than some large U.S. cities. For example, a large farm with 
800,000 hogs could produce over 1.6 million tons of manure per year, 
which is one and a half times more than the annual sanitary waste 
produced by the city of Philadelphia, Pennsylvania--about 1 million 
tons--with a population of almost 1.5 million.[Footnote 14] Moreover, a 
beef cattle farm with 140,000 head of cattle could produce over 1.6 
million tons of manure annually, more than the almost 1.4 million tons 
of sanitary waste generated by the more than 2 million residents of 
Houston, Texas.[Footnote 15] 

Although manure is considered a valuable commodity, especially in 
states with large amounts of farmland, like Iowa, where it is used as 
fertilizer for field crops, in some parts of the country, large farms 
that raise animals are clustered in a few contiguous counties. This 
collocation of large farms that raise animals has resulted in a 
separation of animal production from crop production because many of 
these operations purchase feed rather than grow it on adjacent 
cropland. As a result, there is much less cropland on which the manure 
can be applied as fertilizer. This clustering of large farms that raise 
animals has occurred because of structural changes in the farming 
sector. According to agricultural experts and USDA officials, the 
overall decrease in the number of farms and increase in the average 
number of animals raised on a farm may have occurred because these 
operations wanted to achieve economies of size. To achieve these 
economies, operators often need significant amounts of capital, which 
they obtain through production contracts with large processing 
companies. 

A USDA report identified this concern as early as 2000 when it found 
that between 1982 and 1997 as livestock production became more 
spatially concentrated that when manure was applied to cropland, crops 
were not fully using the nutrients in manure and this could result in 
ground and surface water pollution from the excess nutrients.[Footnote 
16] According to the report, the number of counties where farms 
produced more manure nutrients, primarily nitrogen and phosphorus, than 
could be applied to the land without accumulating nutrients in the soil 
increased. Specifically, the numbers of counties with excess manure 
nitrogen increased by 103 percent, from 36 counties in 1982 to 73 
counties in 1997. Similarly, the number of counties with excess manure 
phosphorous increased by 57 percent, from 102 counties in 1982 to 160 
counties in 1997. As a result, the potential for runoff and leaching of 
these nutrients from the soil was high, and water quality could be 
impaired, according to USDA. Agricultural experts and government 
officials who we spoke to during our review echoed the findings of 
USDA's report and provided several examples of more recent clustering 
trends that have resulted in degraded water quality, including the 
following: 

* As a result of adopting the poultry industry's approach of developing 
close ties between producers and processors,[Footnote 17] North 
Carolina experienced a rapid growth in the number of hog CAFOs, 
primarily in five contiguous counties. Based on our analysis of 2002 
USDA data, we estimated that the hog population of the five North 
Carolina counties was more than 7.5 million hogs in 2002 and that hog 
operations in these counties produced as much as 15.5 million tons of 
manure that year. Figure 1 shows the geographic concentration of hog 
farms in North Carolina in 2002. 

Figure 1: Geographic Concentration of Hogs in Five Contiguous North 
Carolina Counties, 2002: 

[See PDF for image] 

This figure is a map of the state of North Carolina depicting the 
geographic concentration of hogs in five contiguous counties. The 
following information is included: 

Total number of hogs in five counties: 7,551,920 producing about 15.5 
million tons of manure per year[A]. 

Bladen County hog population: 1,150,667; 
Duplin County hog population: 2,575,444; 
Greene County hog population: 520,539; 
Sampson County hog population: 2,587,421; 
Wayne County hog population: 717,849. 

Source: GAO analysis of USDA data. 

Note: Hog populations are the number of hogs on a typical day per 
county in 2002. The number of hogs was estimated by dividing hogs-to- 
market sales by two production cycles and adjusting for inventory on 
hand at the end of the year. 

[A] This is the amount of manure that would be produced if all of these 
hogs were in the feeder-finish production phase where they start at 
about 27 pounds and are marketed at about 260 pounds. The amount of 
manure would be less if a large percentage of these hogs were nursery 
pigs (up to about 27 pounds). Although we were unable to determine what 
percentage of hogs in these counties was not in the feeder-finish 
production cycle, we adjusted our estimates based on 1997 USDA data 
that showed that 25 percent of swine sold were not in the feeder-finish 
production cycle. 

[End of figure] 

According to North Carolina agricultural experts, excessive manure 
production has contributed to the contamination of some of the surface 
and well water in these counties and the surrounding areas. According 
to these experts, this contamination may have occurred because the hog 
farms are attempting to dispose of excess manure but have little 
available cropland that can effectively use it. According to state 
officials, partly out of concern for the potential contamination of 
waterways and surface water from manure, in 1997, North Carolina placed 
a moratorium on new swine farms and open manure lagoons, which was 
subsequently continued through 2007. While the moratorium included 
exceptions that could allow a new swine farm to begin operations in 
this area, according to state officials, the requirements for these 
exceptions are so stringent that they effectively have prevented the 
construction of new swine operations or the expansion of existing 
operations. 

* Similarly, a California water official told us that the geographic 
clustering of large farms that raise animals is causing concern in his 
state as well. Our analysis of USDA data shows that in 2002 two 
counties in the San Joaquin Valley in California had 535,443 dairy cows 
that produced about 13.6 million tons of manure that year. According to 
the official, because of the limited flow of water through the Valley, 
once pollutants reach the water, they do not dissipate, resulting in a 
long-term accumulation of these pollutants. 

* Regional clustering is also occurring in Arkansas. Two counties in 
northwest Arkansas, located on the Arkansas-Oklahoma border, raised 
14,264,828 broiler chickens that produced over 471,000 tons of manure 
that year. According to EPA Region 6 officials, the Arkansas-Oklahoma 
border is an area of concern due to the number of poultry operations 
(primarily broilers, but also turkeys and layers) within this area. 
Furthermore, region 6 officials identified numerous water bodies in 
northwest Arkansas and northeast Oklahoma that have been impaired by 
manure from animal feeding operations and identified these locations as 
"areas of general ground water concern." 

While USDA officials acknowledge that regional clustering of large 
animal feeding operations has occurred, they told us that they believe 
the nutrient management plans that they have helped livestock and 
poultry producers develop and implement have reduced the likelihood 
that pollutants from manure are entering ground and surface water. They 
also believe that as a result of new technologies such as calibrated 
manure spreaders, improved animal feeds, and systems that convert 
manure into electricity, large animal feeding operations are able to 
more effectively use the manure being generated. However, USDA could 
not provide information on the extent to which these techniques are 
being utilized or their effectiveness in reducing water pollution from 
animal waste. 

Studies Have Identified Impacts of Pollutants from Animal Waste, but 
EPA Has Not Assessed the Extent of Such Impacts: 

Since 2002, at least 68 government-sponsored or peer-reviewed studies 
have been completed on air and water pollutants from animal feeding 
operations. Of these 68 studies, 15 have directly linked pollutants 
from animal waste generated by these operations to specific health or 
environmental impacts, 7 have found no impacts, and 12 have made 
indirect linkages between these pollutants and health and environmental 
impacts. In addition, 34 of the studies have focused on measuring the 
amount of certain pollutants emitted by animal feeding operations that 
are known to cause human health or environmental impacts at certain 
concentrations. Appendix IV presents information, including the 
sponsor, the pollutants, and impacts, identified for each of the 68 
studies we reviewed.[Footnote 18] Although EPA is aware of the 
potential impacts of air and water pollutants from animal feeding 
operations, it lacks data on the number of animal feeding operations 
and the amount of discharges actually occurring. Without such data, 
according to EPA officials, the agency is unable to assess the extent 
to which these pollutants are harming human health and the environment. 

Some Recent Studies Directly Link Pollutants from Animal Waste to 
Health and Environmental Impacts: 

Of the 15 studies completed since 2002 that we reviewed that directly 
link pollutants from animal waste to human health or environmental 
impacts, 8 focused on water pollutants and 7 on air pollutants. 
Academic experts and industry and EPA officials told us that only a few 
studies directly link CAFOs with health or environmental impacts 
because the same pollutants that CAFOs discharge also often come from 
other sources including smaller livestock operations; row crops using 
commercial fertilizers; and wastes from humans, municipalities, or 
wildlife, making it difficult to distinguish the actual source of 
pollution. Table 5 shows the eight government-sponsored or peer- 
reviewed studies completed since 2002 that found direct links between 
water pollutants from animal waste and impacts on human health or the 
environment. 

Table 5: Studies Completed Since 2002 Linking Water Pollutants from 
Animal Feeding Operations with Impacts on Human Health or the 
Environment: 

Study title: Effects of the Feedlot Contaminant 17-Trenbolone on 
Reproductive Endocrinology of the Fathead Minnow; 
Sponsor[A]: EPA; 
Pollutant(s) studied: Hormones; 
Impact identified: Adverse effects to reproductive system of aquatic 
life. 

Study title: Endocrine-Disrupting Effects of Cattle Feedlot Effluent on 
an Aquatic Sentinel Species, the Fathead Minnow; 
Sponsor[A]: University of Florida, St. Mary's College of Maryland, 
University of Nebraska, EPA, Tufts University; 
Pollutant(s) studied: Hormones; 
Impact identified: Adverse effects to reproductive system of aquatic 
life. 

Study title: Effects of the Androgenic Growth Promoter 17ß-Trenbolone 
on Fecundity and Reproductive Endocrinology of the Fathead Minnow; 
Sponsor[A]: EPA, University of Minnesota; 
Pollutant(s) studied: Hormones; 
Impact identified: Adverse effects to reproductive system of aquatic 
life. 

Study title: In Vitro and in Vivo Effects of 17ß-Trenbolone: A Feedlot 
Effluent Contaminant; 
Sponsor[A]: EPA; 
Pollutant(s) studied: Hormones; 
Impact identified: Reproductive malformations in laboratory rats and 
human cells. 

Study title: Characterization of Waterborne Outbreak-associated 
Campylobacter jejuni, Walkerton, Ontario; 
Sponsor[A]: Health Canada; 
Pollutant(s) studied: Bacteria; 
Impact identified: Gastrointestinal illness and death in humans. 

Study title: Impact of Animal Waste Application on Runoff Water Quality 
in Field Experimental Plots; 
Sponsor[A]: Jackson State University, National Institutes of Health-
Center for Environmental Health, Louisiana State University; 
Pollutant(s) studied: Nutrients, bacteria; 
Impact identified: Water degradation. 

Study title: Nutrient Loading Patterns on an Agriculturally Impacted 
Stream System in Huntingdon County Pennsylvania over Three Summers; 
Sponsor[A]: Juniata College; 
Pollutant(s) studied: Nutrients; 
Impact identified: Water degradation; unable to sustain aquatic life. 

Study title: Concentrated Animal Feeding Operations, Row Crops, and 
Their Relationship to Nitrate in Eastern Iowa Rivers; 
Sponsor[A]: University of Iowa; 
Pollutant(s) studied: Nutrients; 
Impact identified: Water degradation. 

Source: GAO's analysis of identified studies. 

[A] Sponsor refers to the organization under whose auspices the 
research was conducted or with whom the primary researchers were 
affiliated. 

[End of table] 

As table 5 shows, EPA sponsored four of the water quality studies that 
identified reproductive alterations in aquatic species caused by 
hormones in discharges from animal feeding operations. Two of these 
studies found that hormones from these discharges caused a significant 
decline in the fertility of female fish in nearby water bodies. 
Similarly, three other studies found water bodies impaired by higher 
nitrogen and phosphorus levels from manure runoff from animal feeding 
operations. For example, the study by Juniata College found that the 
runoff resulted in nutrient concentrations in the water that were too 
high to sustain fish populations. Only one of the eight water pollutant 
studies linked pollutants from animal feeding operations to human 
health effects. This study, conducted by Health Canada, directly linked 
water discharges from a cattle farm to bacteria found in nearby waters. 
These bacteria, which included Campylobacter and E. coli, caused 
gastrointestinal illnesses in more than 2,300 residents and 7 deaths in 
a nearby community. 

Table 6 shows the seven government-sponsored or peer-reviewed studies 
completed since 2002 that we reviewed that directly link air pollutants 
from animal feeding operations with human health effects. 

Table 6: Studies Completed Since 2002 Directly Linking Air Pollutants 
from Animal Feeding Operations to Impacts on Human Health: 

Study title: Feedlot Dust Stimulation of Interleukin-6 and 8 Requires 
Protein Kinase C-Epsilon Human Bronchial Epithelial Cells; 
Sponsor[A]: Nebraska Medical Center, Department of Veterans Affairs 
Medical Center, Texas A&M; 
Pollutant(s) studied: Dust; 
Impact identified: Respiratory inflammation. 

Study title: Farm Residence and Exposures and the Risk of Allergic 
Diseases In New Zealand Children; 
Sponsor[A]: University of Otago, New Zealand; 
Pollutant(s) studied: Dust; 
Impact identified: Greater prevalence of allergies in children living 
on farms. 

Study title: Exhaled Nitric Oxide and Bronchial Responsiveness in 
Healthy Subjects Exposed to Organic Dust; 
Sponsor[A]: National Institute of Environmental Medicine, Sweden; 
Pollutant(s) studied: Dust; 
Impact identified: Respiratory inflammation (occupational). 

Study title: Hog Barn Dust Extract Augments Lymphocyte Adhesion to 
Human Airway Epithelial Cells; 
Sponsor[A]: Department of Veterans Affairs Medical Center, University 
of Nebraska Medical Center; 
Pollutant(s) studied: Dust; 
Impact identified: Respiratory inflammation (occupational). 

Study title: Hog Barn Dust Extract Stimulates IL-8 And IL-6 Release in 
Human Bronchial Epithelial Cells Via PKC Activation; 
Sponsor[A]: Department of Veterans Affairs Medical Center, University 
of Nebraska Medical Center; 
Pollutant(s) studied: Dust; 
Impact identified: Respiratory inflammation (occupational). 

Study title: Experimental Human Exposure to Inhaled Grain Dust and 
Ammonia: Towards a Model of Concentrated Animal Feeding Operations; 
Sponsor[A]: University of Iowa; 
Pollutant(s) studied: Dust, ammonia; 
Impact identified: Tightening of airway in asthmatics (occupational). 

Study title: Symptomatic Effects of Exposure to Diluted Air Sampled 
from a Swine Confinement Atmosphere on Healthy Human Subjects; 
Sponsor[A]: Duke University; 
Pollutant(s) studied: Hydrogen sulfide, ammonia, endotoxin, dust, odor; 
Impact identified: Headaches, eye irritation, nausea. 

Source: GAO's analysis of identified studies. 

[A] Sponsor refers to the organization under whose auspices the 
research was conducted or with whom the primary researchers were 
affiliated. 

[End of table] 

As table 6 shows, six of these studies identified airway inflammation 
or wheezing in people working at or living on an animal feeding 
operation. For example, the studies conducted by the Department of 
Veterans Affairs show that the dust of hog confinement facilities 
induces airway inflammation in workers. The seventh study, completed by 
Duke University in a laboratory setting, exposed healthy volunteers to 
air emissions consistent with those that would occur downwind from 
animal feeding operations. These volunteers reported headaches, eye 
irritation, and nausea following this exposure. According to experts 
who we spoke with, the effects of air emissions from animal feeding 
operations on workers are well known, but the impacts of these 
emissions on nearby communities are still uncertain, and more research 
is needed to identify these impacts. Additionally, experts said it is 
difficult to determine which specific contaminant or mixture of 
contaminants causes particular health symptoms. For example, while 
hydrogen sulfide causes respiratory and other health problems, other 
contaminants emitted from animal feeding operations, such as ammonia, 
can also cause similar symptoms. 

Some Studies Found No Links between Pollutants from Animal Feeding 
Operations and Harm to Human Health or the Environment: 

We found seven government-sponsored or peer-reviewed studies that have 
been completed since 2002 that found no impact on human health or the 
environment from pollutants released by animal feeding operations. 
These seven studies are shown in table 7. 

Table 7: Studies Completed Since 2002 Finding No Links between 
Pollutants from Animal Feeding Operations and Impacts on Human Health 
or the Environment: 

Study title: Prevalence of Escherichia coli O157:H7 Bacterial 
Infections Associated with the Use of Animal Wastes in Louisiana for 
the Period 1996-2004; 
Sponsor[A]: Grambling State University, Louisiana State University, 
Jackson State University; 
Pollutant(s) studied: Escherichia coli; 
Finding(s): No clear indication that any cases of E. coli infection are 
related to animal waste. 

Study title: Prevalence of Selected Bacterial Infections Associated 
with the Use of Animal Waste in Louisiana; 
Sponsor[A]: Jackson State University, Louisiana State University; 
Pollutant(s) studied: Escherichia coli; 
Finding(s): No clear indication that any cases of E. coli infection are 
related to animal waste. 

Study title: Impacts of Swine Manure Pits on Groundwater Quality; 
Sponsor[A]: Illinois State Geological Survey, University of Illinois, 
Illinois Department of Agriculture; 
Pollutant(s) studied: Chloride, ammonium, phosphate, potassium, 
nitrate, bacteria; 
Finding(s): Manure seepage from swine facilities has had limited 
impacts on groundwater. 

Study title: Ground-Water Quality and Effects of Poultry Confined 
Animal Feeding Operations on Shallow Ground Water, Upper Shoal Creek 
Basin, Southwest Missouri, 2000; 
Sponsor[A]: U.S. Geological Survey; 
Pollutant(s) studied: Nutrients, bacteria; 
Finding(s): The results do not indicate that poultry CAFOs are 
affecting the shallow ground water with respect to nutrients and fecal 
bacteria. 

Study title: Environmental Exposure to Endotoxin and Its Relation to 
Asthma in School-Age Children; 
Sponsor[A]: Institute of Social and Preventive Medicine (Switzerland), 
Children's Hospital (Austria), Philipps University (Germany), Ruhr 
University (Germany), University Children's Hospital (Switzerland), 
University of Munich (Germany; 
Pollutant(s) studied: Dust; 
Finding(s): Decreased risk of hay fever, asthma, and wheeze in children 
exposed to high levels of endotoxin in dust. 

Study title: Ecological Associations between Asthma Prevalence and 
Potential Exposure to Farming; 
Sponsor[A]: University of North Carolina; 
Pollutant(s) studied: Farm air; 
Finding(s): Farm exposures may be protective against childhood asthma. 

Study title: Atmospheric Pollutants and Trace Gases: Atmospheric 
Ammonia, Volatile Fatty Acids, and Other Odorants near Beef Feedlots; 
Sponsor[A]: Research Centre, Agriculture and Agri-Food Canada; 
Pollutant(s) studied: Ammonia, odor, organic compounds, dust; 
Finding(s): Odorants from feedlots were effectively dispersed. Emitted 
ammonia was deposited to the soil downwind. 

Source: GAO's analysis of identified studies. 

[A] Sponsor refers to the organization under whose auspices the 
research was conducted or with whom the primary researchers were 
affiliated. 

[End of table] 

As table 7 shows, the results of a U.S. Geological Survey study did not 
indicate that poultry animal feeding operations were causing an 
increase of nutrient concentrations and fecal bacteria in groundwater. 
Similarly, another study by Agriculture and Agri-Food Canada found that 
odorants, including ammonia and dust emitted by animal feeding 
operations, never exceeded the established irritation threshold. 
According to EPA and academic experts we spoke with, the concentrations 
of air pollutants and water pollutants emitted by animal feeding 
operations can vary, which may account for the differences in the 
findings of these studies. These variations may be the result of 
numerous factors, including the type of animals being raised, feed 
being used, and manure management system being employed, as well as the 
climate and time of day when the emissions occur. 

Some Recent Studies Indirectly Link Pollutants from Animal Feeding 
Operations with Human Health and Environmental Impacts: 

We also identified 12 government-sponsored or peer-reviewed studies 
completed since 2002 that indirectly link pollutants from animal 
feeding operations to human health or environmental impacts. While 
these studies found that animal feeding operations were the likely 
cause of human health or environmental impacts occurring in areas near 
the operations, they could not conclusively link waste from animal 
feeding operations to the impacts, often because other sources of 
pollutants could also be contributing. For example, 5 of these 12 
studies found an increased incidence of asthma or respiratory problems 
in people living or attending school near animal feeding operations, 
compared with a control group. These studies hypothesized that the 
pollutants emitted from animal feeding operations were likely the cause 
of the increased incidence of asthma, but some of these studies 
acknowledged that pollutants from other sources could also be 
contributing to the increased incidence. Table 8 lists the 12 studies 
that have been completed since 2002 that made indirect links between 
emissions from animal feeding operations and human health and 
environmental impacts. 

Table 8: Studies Completed Since 2002 That Found an Indirect Link 
between Pollutants from Animal Feeding Operations and Human Health or 
Environmental Impacts: 

Study title: Associations between Indicators of Livestock Farming 
Intensity and Incidence of Human Shiga Toxin-Producing Escherichia coli 
Infection; 
Sponsor[A]: University of Guelph; Université de Montréal; Centre for 
Infectious Disease Prevention and Control - Health Canada; 
Impact(s): The strongest associations with human Escherichia coli 
infection were the ratio of beef cattle to human population and the 
application of manure to the surface of agricultural land by a solid 
spreader and by a liquid spreader. 

Study title: The Potential Impact of Flooding on Confined Animal 
Feeding Operations in Eastern North Carolina; 
Sponsor[A]: University of North Carolina; 
Impact(s): Flood events have a significant potential to degrade 
environmental health because of dispersion of wastes from industrial 
animal operations in areas with vulnerable populations. 

Study title: Odor from Industrial Hog Farming Operations and Mucosal 
Immune Function in Neighbors; 
Sponsor[A]: University of North Carolina, Duke University; 
Impact(s): This study suggests that malodor from industrial swine 
operations can affect the secretory immune system, although the reduced 
levels reported are still within normal range. 

Study title: Environmental Stressors, Perceived Control, and Health: 
The Case of Residents Near Large-Scale Hog Farms in Eastern North 
Carolina; 
Sponsor[A]: University of North Carolina Wilmington; 
Impact(s): Residents living near large-scale hog farms in eastern North 
Carolina report symptoms related to respiratory, sinus, and nausea 
problems. 

Study title: Asthma Prevalence and Morbidity Among Rural Iowa 
Schoolchildren; 
Sponsor[A]: University of Iowa, EPA; Impact(s): Among children who 
wheeze, farm and nonfarm children were equally likely to have been 
given a diagnosis of asthma and had comparable morbidity. Asthma in 
rural schoolchildren was comparable to schoolchildren in large cities. 

Study title: Occupational Asthma in Newly Employed Workers in Intensive 
Swine Confinement Facilities; 
Sponsor[A]: Institute of Agricultural Rural and Environmental Health, 
University of Saskatchewan, Laval University; 
Impact(s): Newly employed workers in intensive swine confinement 
facilities reported development of acute onset of wheezing and cough 
suggestive of asthma. 

Study title: Asthma and Farm Exposures in a Cohort of Rural Iowa 
Children; 
Sponsor[A]: University of Iowa, EPA, Colorado State University, Kaiser 
Permanente; 
Impact(s): There was a high prevalence of asthma health outcome among 
farm children living on farms that raise swine and raise swine and add 
antibiotics. 

Study title: Asthma Symptoms among Adolescents Who Attend Public 
Schools That Are Located Near Confined Swine Feeding Operations; 
Sponsor[A]: University of North Carolina, RTI International; 
Impact(s): Estimated exposure to airborne pollution from confined swine 
feeding operations is associated with adolescents' wheezing symptoms. 

Study title: Airway Responses of Healthy Farmers and Nonfarmers to 
Exposure in a Swine Confinement Building; 
Sponsor[A]: National Institute of Environmental Medicine (Sweden), 
National Institute for Working Life (Sweden); 
Impact(s): Altered lung function and bronchial responsiveness was found 
in nonfarming subjects. Only minor alterations were found in the 
farmers. 

Study title: Environmental Exposure to Confined Animal Feeding 
Operations and Respiratory Health of Neighboring Residents; 
Sponsor[A]: Institute for Occupational and Environmental Medicine 
(Germany), National Research Centre for Environment and Health 
(Germany), Boston University, Municipal Health Service Amsterdam; 
Impact(s): Respiratory disease was found among residents living near 
confined animal feeding operations. 

Study title: School Proximity to Concentrated Animal Feeding Operations 
and Prevalence of Asthma in Students; 
Sponsor[A]: University of Iowa Carver College of Medicine, University 
of Iceland; 
Impact(s): Children in the study school, located one-half mile from a 
CAFO, had a significantly increased prevalence of physician-diagnosed 
asthma. 

Study title: Lung Function and Farm Size Predict Healthy Worker Effect 
in Swine Farmers; 
Sponsor[A]: University of Saskatchewan (Canada); 
Impact(s): Some swine workers are less affected by swine air and 
continue in the profession. Other workers are more affected. 

Source: GAO's analysis of identified studies. 

[A] Sponsor refers to the organization under whose auspices the 
research was conducted or with whom the primary researchers were 
affiliated. 

[End of table] 

Many Recent Studies Have Measured the Level of Pollutants Emitted by 
Animal Feeding Operations: 

Thirty-four government-sponsored or peer-reviewed studies completed 
since 2002 have focused on measuring the amounts of water or air 
pollutants emitted by animal feeding operations that are known to cause 
harm to humans or the environment. Specifically: 

* Nineteen of the 34 studies focused on water pollutants. Four studies 
found increased levels of phosphorus or nitrogen in surface water and 
groundwater near animal feeding operations. According to EPA, excessive 
amounts of these nutrients can deplete oxygen in water, which could 
result in fish deaths, reduced aquatic diversity, and illness in 
infants. The other 15 studies measured water pollutants such as 
pathogens, hormones, and antibiotics. 

* Fifteen of the 34 studies focused on measuring air emissions from 
animal feeding operations. Seven of the 15 studies found high levels of 
ammonia surrounding animal feeding operations. EPA considers ammonia a 
hazardous substance that may harm human health or the environment, and 
that must be reported when emissions exceed its reportable quantity. 
The other eight studies measured the levels of other air pollutants, 
such as hydrogen sulfide, particulate matter, and carbon dioxide. 

Appendix IV provides additional details about each of the 34 studies. 

EPA Has Not Yet Assessed the Extent of the Human Health and 
Environmental Impacts of Pollutants from Animal Feeding Operations: 

While EPA recognizes the potential impacts that water and air 
pollutants from animal feeding operations can have on human health and 
the environment, it lacks the data necessary to assess how widespread 
these impacts are and has limited plans to collect the data it needs. 

Water quality. EPA has long recognized the impacts of pollution from 
CAFOs on water quality. For example, almost a decade ago, in its 1998 
study on feedlot point sources, EPA documented environmental impacts 
that may be attributed to these operations.[Footnote 19] This report 
identified pollutants from animal feeding operations and listed about 
300 spills and runoff events that were attributable to animal feeding 
operations from 1985 through 1997. More recently when developing the 
2003 CAFO rule, EPA documented the potential water quality impacts from 
CAFOs. It reported that contaminants in manure will have an impact on 
water quality if significant amounts reach surface water or 
groundwaters. Moreover, as discussed above, numerous studies completed 
since 2002 have provided additional information on the direct and 
indirect impacts of discharges from animal feeding operations on human 
health and the environment, and many more studies have been completed 
that have measured the amounts of pollutants being discharged. 

EPA officials we spoke with acknowledged that the potential human 
health and environmental impacts of some CAFO water pollutants, such as 
nitrogen, phosphorus, and pathogens, are well known. They told us that 
the agency has recently focused its research efforts on obtaining more 
information on emerging pollutants, such as hormones and antibiotics, 
and on how the concentrations of nutrients and pathogens differ among 
the various types of animal feeding operations. However, these 
officials also stated that EPA does not have data on the number and 
location of CAFOs nationwide and the amount of discharges from these 
operations. Without this information and data on how pollutant 
concentrations vary by type of operation, it is difficult to estimate 
the actual discharges occurring and to assess the extent to which CAFOs 
may be contributing to water pollution. According to agency officials, 
because of a lack of resources, the agency currently has no plans for a 
national study to collect information on CAFO water discharges. 
However, the agency has recently taken the following three steps that 
may help gather additional data on CAFO pollutants that affect water 
quality: 

* EPA has begun research to determine (1) how the concentration of 
pathogens and nutrients vary in manure on the basis of certain 
characteristics, such as animal type and animal feed, and (2) how 
manure management techniques can reduce the amount of pathogens and 
nutrients in runoff. 

* EPA has set a long-term research goal, as part of its Multi-Year Plan 
for Endocrine Disruptors (FY2007-2013), to characterize the magnitude 
and extent of the impact of hormones released by CAFOs and to determine 
the impact of management strategies on the fate and effects of 
hormones. At the time of our review, according to an EPA official, the 
agency had only limited preliminary findings because it has just 
recently begun this work. 

* EPA and the U.S. Geological Survey have discussed a joint project to 
identify (1) the location of CAFOs nationwide and (2) those watersheds 
where many CAFOs might be located. According to EPA officials, this 
project is still in the discussion phase. 

Air quality. More recently, EPA has recognized concerns about the 
possible health impacts from air emissions produced by animal feeding 
operations. Prompted in part by public concern, EPA and USDA 
commissioned a 2003 study by the National Academy of Sciences (NAS) to 
evaluate the scientific information needed to support the regulation of 
air emissions from animal feeding operations.[Footnote 20] The NAS 
report identified several air pollutants from animal feeding operations 
and their potential impacts. For example, the study identified ammonia 
and hydrogen sulfide as two air pollutants emitted from animal feeding 
operations that can impair human health. According to the study, 
ammonia can cause eye, nose, and throat irritation at certain 
concentrations, and hydrogen sulfide can cause respiratory distress. 
While such effects are known to occur, the study noted that additional 
research is warranted to determine if air emissions from animal feeding 
operations are occurring in high enough concentrations to cause these 
effects. The NAS report also concluded that in order to determine the 
human health and environmental effects of air emissions from animal 
feeding operations, EPA and USDA would first need to obtain accurate 
estimates of emissions and their concentrations from animal feeding 
operations with varying characteristics, such as animal type, animal 
feed, manure management techniques, and climate. 

Since the NAS report was issued, EPA has conducted one hypothetical 
assessment of the impacts of air emissions from animal feeding 
operations. In 2004, EPA updated a preliminary analysis to estimate the 
levels of emissions of ammonia and hydrogen sulfide that occur downwind 
from a manure lagoon and that could pose a risk to human health. EPA 
found that ammonia would not reach levels associated with respiratory 
irritation if emitted at the reportable quantity of 100 pounds per day. 
[Footnote 21] On the other hand, the agency found that hydrogen sulfide 
could cause respiratory irritation and central nervous system effects 
about one mile downwind if emitted at the reportable quantity of 100 
pounds per day.[Footnote 22] EPA officials who conducted this analysis 
told us that there have been no documented cases of hydrogen sulfide 
emissions from animal feeding operations exceeding the reportable 
quantity. However, other officials noted that the agency does not know 
exactly what type of species and what size of operations are likely to 
have emissions above the reportable quantity, and, as noted in the NAS 
report, accurate measurements of the air pollutants being emitted by 
animal feeding operations are currently not known. 

In 2007, a national air emissions monitoring study to collect data on 
air emissions from animal feeding operations was undertaken as part of 
a series of consent agreements EPA entered into with individual animal 
feeding operations. This study, funded by industry and approved by EPA, 
is intended to help the agency determine how to measure and quantify 
air emissions from animal feeding operations. The data collected will 
in turn be used to estimate air emissions from animal feeding 
operations with varying characteristics, and, according to EPA 
officials, it is only the first step in a long-term effort to 
accurately quantify air emissions from animal feeding operations. 
According to agency officials, until EPA can determine the actual level 
of emissions occurring, it will be unable to assess the extent to which 
these emissions are affecting human health and the environment. 
Progress in conducting the national air emissions monitoring study is 
discussed in greater detail in the following section. 

It Is Unclear If EPA's Efforts to Develop Air Emissions Protocols for 
Animal Feeding Operations Will Be Effective and How EPA Intends to 
Regulate These Emissions in the Future: 

The National Air Emissions Monitoring Study--a 2-year effort to collect 
data on air emissions from animal feeding operations--is intended to 
provide a scientific basis for estimating air emissions from these 
operations. The results of this study were intended to help EPA develop 
protocols that will allow it to determine which operations do not 
comply with applicable federal laws. As currently structured, however, 
the study may not provide the quantity and quality of data needed for 
developing appropriate methods for estimating emissions. Furthermore, 
it is uncertain if and when EPA will develop a process-based model that 
considers the interaction and implications of all sources of emissions 
at an animal feeding operation. Also, other more recent decisions 
suggest that the agency has not yet determined how it intends to 
regulate air emissions from animal feeding operations. In the absence 
of federal guidance on how to regulate air emissions from animal 
feeding operations, a few states have developed their own regulations. 

A National Air Emissions Monitoring Study Has Begun, but the Study May 
Not Provide the Data EPA Needs to Develop Air Emissions Protocols: 

According to EPA, although it has the authority to require animal 
feeding operations to monitor their emissions and come into compliance 
with the Clean Air Act on a case-by-case basis, this approach has 
proven to be time and labor intensive. As an alternative to the case- 
by-case approach, in January 2005, EPA offered animal feeding 
operations an opportunity to sign a voluntary consent agreement and 
final order, known as the Air Compliance Agreement. To participate in 
the agreement, animal feeding operations were required to take the 
following actions: 

* Pay a civil penalty ranging from $200 to $1,000 per animal feeding 
operation, depending on the number of animals at the operation and the 
number of operations that each participant signed up.[Footnote 23] 

* Pay up to $2,500 per farm to help fund a nationwide emissions 
monitoring study and make their facilities available as a monitoring 
site for emissions testing. 

* Once emission protocols are published, apply for all applicable air 
permits and comply with permit conditions, if deemed necessary. 

* Once emission protocols are published, report any releases of ammonia 
and hydrogen sulfide above the threshold levels established by CERCLA 
and EPCRA.[Footnote 24],[Footnote 25] 

In return for meeting these requirements, EPA agreed not to sue 
participating animal feeding operations for certain past violations or 
violations occurring during the emissions monitoring study.[Footnote 
26] 

Almost 13,900 animal feeding operations were approved for participation 
in the agreement, representing the egg, broiler chicken, dairy, and 
swine industries. Some turkey operations volunteered but were not 
approved because there were too few operations to fund a monitoring 
site, and the beef cattle industry chose not to participate. EPA 
collected a total of $2.8 million in civil penalties from participating 
animal feeding operations and deposited these funds into the U.S. 
Treasury. An additional $14.8 million was collected by a nonprofit, 
industry-established organization to fund the national air emissions 
monitoring study. Industry groups representing the participating 
operations provided the funding for the study as was called for under 
the agreement. Table 9 shows the level of participation by type of 
operation and the amount of funding provided by different industry 
groups for the national air emissions monitoring study. 

Table 9: Number of Participants in the Air Compliance Agreement, 
Funding Provided by Animal Type, and Source of the Funding for the 
National Air Emissions Monitoring Study (Dollars in millions): 

Animal type: Swine; 
Air Compliance Agreement: Number of participants: 1,878; 
Air Compliance Agreement: Number of animal feeding operations: 4,865; 
National Air Emissions Monitoring Study: Funding provided: $6.0; 
National Air Emissions Monitoring Study: Funding source: National Pork 
Board. 

Animal type: Dairy; 
Air Compliance Agreement: Number of participants: 474; 
Air Compliance Agreement: Number of animal feeding operations: 573; 
National Air Emissions Monitoring Study: Funding provided: $5.0; 
National Air Emissions Monitoring Study: Funding source: National Milk 
Producers Council. 

Animal type: Layers; 
Air Compliance Agreement: Number of participants: 218; 
Air Compliance Agreement: Number of animal feeding operations: 2,693; 
National Air Emissions Monitoring Study: Funding provided: $2.8; 
National Air Emissions Monitoring Study: Funding source: United Egg 
Producers. 

Animal type: Broilers; 
Air Compliance Agreement: Number of participants: 41; 
Air Compliance Agreement: Number of animal feeding operations: 5,752; 
National Air Emissions Monitoring Study: Funding provided: $1.0; 
National Air Emissions Monitoring Study: Funding source: National 
Chicken Council. 

(Dollars in millions): Animal type: Total; 
Air Compliance Agreement: Number of participants: 2,611; 
Air Compliance Agreement: Number of animal feeding operations: 13,883; 
National Air Emissions Monitoring Study: Funding provided: $14.8; 
National Air Emissions Monitoring Study: Funding source: [Empty]. 

Source: EPA. 

[End of table] 

The purpose of the National Air Emissions Monitoring Study is to 
collect data that will provide a scientific basis for measuring and 
estimating air emissions from animal feeding operations and will help 
EPA to determine operations' compliance status. To provide a framework 
for the monitoring study and develop a sampling plan that was 
representative of animal feeding operations in the United States, in 
2003 EPA convened a panel of industry experts, university and 
government scientists, and other stakeholders knowledgeable in the 
field. In 2004, the nonprofit organization founded by the various 
livestock sectors selected an independent science adviser to oversee 
the data collection at 20 of the 13,883 animal feeding operations that 
were selected to participate in the study. Their selection was 
submitted to and approved by EPA. Data collection began in May 2007. 
Once 2 years of data has been collected, EPA will use these data to 
develop air emissions protocols. Figure 6 shows EPA's expected timeline 
for the development of air emissions protocols. 

Figure 2: EPA Timeline for Development of Air Emission Protocols for 
Animal Feeding Operations: 

[See PDF for image] 

This figure provides the following information on the timeline: 

* Protocol development: January 2004 through Mid-2004;
* Staffing, budgeting: Early 2004 through January 2006; 
* Producer signup: Mid-2005 through January 2006; 
* EPA reviews Consent Agreements: Mid-2005 through January 2007; 
* Site selections: Mid-2005 through January 2007; 
* EPA reviews plan: Mid-2006 through January 2007; 
* National Air Emissions Monitoring Study: 
- Equipment Acquisition/Purdue Training sessions: January 2007 through 
January 2008; 
- Site setup: Early 2007 through early 2008; 
- Data collection: January 2008 through January 2010; 
* EPA develops Emission Estimated Methodologies: January 2010 and 
beyond. 

Source: EPA. 

[End of figure] 

However, the National Air Emissions Monitoring Study may not provide 
the data that EPA needs to develop comprehensive protocols for 
quantifying air emissions from animal feeding operations for a variety 
of reasons. First, the monitoring study does not include the 16 
combinations of animal types and geographic regional pairings 
recommended by EPA's expert panel. The panel recommended this approach 
so that the study sample would be representative of the vast majority 
of participating animal feeding operations, accounting for differences 
in climatic conditions, manure-handling methods, and density of 
operations. However, EPA approved only 12 of the 16 combinations 
recommended by the expert panel, excluding southeastern broiler, 
eastern layer, midwestern turkey, and southern dairy operations. 
Second, site selection for the study has been a concern since the plan 
to select monitoring sites for the monitoring study was announced in 
2005. At that time, many agricultural experts, environmental groups, 
and industry and state officials disagreed with the site selection 
methodology. In commenting on EPA's Federal Register notice of the 
Animal Feeding Operation Consent Agreement and Final Order, these 
experts and officials stated that the study did not include a 
sufficient number of monitoring sites to establish a statistically 
valid sample. Without such a sample, we believe that EPA will not be 
able to accurately estimate emissions for all types of operations. More 
recently, in June 2008, the state of Utah reached an agreement with EPA 
to separately study animal feeding operations in the state because of 
the state's continuing concerns that the National Air Emissions 
Monitoring Study will not collect information on emissions from 
operations in Rocky Mountain states and therefore may not be meaningful 
for those operations that raise animals in arid areas. Finally, 
agricultural experts have raised concerns that the National Air 
Emissions Monitoring Study does not include other sources that can 
contribute significantly to emissions from animal feeding operations. 
For example, these experts have noted that the monitoring study will 
not capture data on ammonia emissions from feedlots and manure applied 
to fields. According to these experts, feedlots and manure on fields, 
as well as other excluded sources, account for approximately half of 
the total ammonia emissions from animal feeding operations. 

Furthermore, USDA's Agricultural Air Quality Task Force has also 
recently raised concerns about the quantity and quality of the data 
being collected during the early phases of the study and how EPA will 
eventually use the information.[Footnote 27] In particular, the task 
force expressed concern that the technologies used to collect emissions 
data were not functioning reliably. For example, according to data 
provided by EPA, almost one-third of the preliminary data from one site 
were incomplete during a 2-month data collection period. The task force 
was also concerned about EPA's plans to extrapolate the data across a 
variety of CAFO operating configurations. At its May 2008 task force 
meeting, the members requested that the Secretary of Agriculture ask 
EPA to review the first 6 months of the study's data to determine if 
the study needs to be revised in order to yield more useful 
information. 

EPA acknowledged that emissions data should be collected for every type 
of animal feeding operation and practice, but EPA officials stated that 
such an extensive study is impractical. According to EPA officials, the 
industry identified those monitoring sites that they believed best 
represented the type of operations and manure management practices that 
are in their various animal sectors. EPA reviewed and approved these 
site selections. According to EPA, it believes that the selected sites 
provide a reasonable representation of the various animal sectors. EPA 
has also indicated that it plans to use other relevant information to 
supplement the study data and has identified some potential additional 
data sources. For example, a study conducted at two broiler facilities 
in Kentucky has been accepted as meeting the emissions study's 
requirements. However, according to agricultural experts, until EPA 
identifies all the supplemental data that it plans to use, it is not 
clear if these data, together with the emissions study data, will 
enable EPA to develop comprehensive air emissions protocols. 

Furthermore, EPA has also indicated that completing the National Air 
Emissions Monitoring Study is only the first step in a multiyear effort 
to develop a process-based model for predicting overall emissions for 
animal feeding operations. A process-based model would capture 
emissions data from all sources and use these data to assess the 
interaction of all sources and the impact that different manure 
management techniques have on air emissions for the entire operation. 
For example, technologies are available to decrease emissions from 
manure lagoons by, among other things, covering the lagoon to capture 
the ammonia. However, if an operation spreads the lagoon liquid as 
fertilizer for crops, ammonia emissions could increase on the field. 
According to NAS, a process-based model is needed to provide 
scientifically sound estimates of air emissions from animal feeding 
operations that can be used to develop management and regulatory 
programs. Although EPA plans to develop a process-based model after 
2011, it has not yet established a timetable for completing this model 
and, therefore, it is uncertain when EPA will have more sophisticated 
approaches that will more accurately estimate emissions from animal 
feeding operations. 

Recent EPA Decisions Suggest That the Agency Has Not Yet Determined How 
It Plans to Regulate Air Emissions from Animal Feeding Operations: 

Two recent decisions by EPA suggest that the agency has not yet 
determined how it intends to regulate air emissions from animal feeding 
operations. EPA's first decision in this context was made in December 
2007. At that time EPA proposed to exempt releases to the air of 
hazardous substances from manure at farms that meet or exceed the 
reportable quantities from both CERCLA and EPCRA notification 
requirements. According to EPA, this decision was in response to 
language that was contained in congressional committee reports related 
to EPA's appropriations legislation for 2005 and 2006. EPA was directed 
to promptly and expeditiously provide clarification on the application 
of these laws to poultry, livestock, and dairy operations. In addition, 
the agency received a petition from the National Chicken Council, the 
National Turkey Federation, and the U.S. Poultry and Egg Association 
seeking an exemption from the CERLA and EPCRA reporting requirements 
for ammonia emissions from poultry operations. The petition argued that 
ammonia emissions from poultry operations pose little or no risk to 
public health, and emergency response is inappropriate. In proposing 
the rule, EPA noted that the agency would not respond to releases from 
animal wastes under CERCLA or EPCRA nor would it expect state and local 
governments to respond to such releases because the source and nature 
of these releases are such that emergency response is unnecessary, 
impractical, and unlikely. It also noted that it had received 26 
comment letters from state and local response agencies supporting the 
exemption for ammonia from poultry operations. However, during the 
public comment period ending on March 27, 2008, a national association 
representing state and local emergency responders with EPCRA 
responsibilities questioned whether EPA had the authority to exempt 
these operations until the agency had data from its monitoring study to 
demonstrate actual levels of emissions from animal feeding operations. 
This national association further commented that EPA should withdraw 
the proposal because it denied responders and the public the 
information necessary to protect themselves from dangerous releases. 
[Footnote 28] The timing of this proposed exemption, before the 
National Air Emissions Monitoring Study has been completed, we believe 
calls into question the basis for EPA's decision. 

The second decision that EPA has recently made that calls into question 
how the agency intends to regulate air emissions from animal feeding 
operations involves the timing of key regulatory decisions. EPA has 
stated that it will not make key regulatory decisions on how federal 
air regulations apply to animal feeding operations until after 2011, 
when the monitoring study is completed. According to EPA, the agency 
will issue guidance defining the scope of the term "source" as it 
relates to animal agriculture and farm activities. As a result, EPA has 
not decided if it will aggregate the emissions occurring on an animal 
feeding operation as one source or if the emissions from the barns, 
lagoons, feed storage, and fields will each be considered as a separate 
source when determining if an operation has exceeded air emissions' 
reportable quantities. Depending on the approach EPA takes, how 
emissions are calculated could differ significantly. For example, 
according to preliminary data EPA has received from an egg-laying 
operation in Indiana, individual chicken barns may exceed the CERCLA 
reportable quantities for ammonia. Moreover, if emissions from all of 
the barns on the operation are aggregated, they might be more than 500 
times the CERCLA reportable quantities. In addition, EPA does not 
intend to issue guidance to address emissions, and sources of 
emissions, that cannot reasonably pass through a stack, chimney, or 
other functionally equivalent opening, i.e., fugitive emissions, until 
after the conclusion of the monitoring study. 

EPA has already been asked to clarify what it considers a source on an 
animal feeding operation but has declined to do so. In a 2004 ruling on 
an appeal of a civil suit against a swine operation, the U.S. Court of 
Appeals for the 10th Circuit overturned a 2002 federal district court 
ruling that a farm's individual barns, lagoons, and land application 
areas could be considered separate "sources" for purposes of CERCLA 
reporting requirements.[Footnote 29] The Court of Appeals ruled that 
the whole farm site was the proper entity to be assessed for purposes 
of CERCLA reporting. The Court invited EPA to file a friend-of-the- 
court brief in order to clarify the government's position on this 
issue, but EPA declined to do so within the court-specified time frame. 
[Footnote 30] Another court reached similar conclusions in 2003. 
[Footnote 31] Despite these court rulings, EPA has indicated that it 
will not decide on what it considers a source until the National Air 
Emissions Monitoring Study is completed. 

Lacking Federal Guidance, Some States Have Begun to Regulate Air 
Emissions from Animal Feeding Operations: 

In the absence of federal guidance on how to regulate air emissions 
from animal feeding operations, officials in 6 states, out of the 47 
states that responded to our survey, are regulating some emissions 
covered under the Clean Air Act, CERCLA and EPCRA. As table 10 shows, 
state officials in California, Idaho, Minnesota, Missouri, Nebraska, 
and North Dakota reported that they have developed state air 
regulations for certain pollutants that are emitted by CAFOs. 

Table 10: States That Reported Having Regulations for Air Emissions 
from Animal Feeding Operations, 2008: 

California: 
Hydrogen sulfide: [Check]; 
Ammonia: [Check]; 
Particulate matter: [Check]; 
Volatile organic compounds: [Check]. 

Idaho: 
Hydrogen sulfide: [Empty]; 
Ammonia: [Check]; 
Particulate matter: [Empty]; 
Volatile organic compounds: [Empty]. 

Minnesota: 
Hydrogen sulfide: [Check]; 
Ammonia: [Empty]; 
Particulate matter: [Empty]; 
Volatile organic compounds: [Empty]. 

Missouri: 
Hydrogen sulfide: [Check]; 
Ammonia: [Empty]; 
Particulate matter: [Check]; 
Volatile organic compounds: [Empty]. 

Nebraska: 
Hydrogen sulfide: [Check]; 
Ammonia: [Empty]; 
Particulate matter: [Empty]; 
Volatile organic compounds: [Empty]. 

North Dakota: 
Hydrogen sulfide: [Check]; 
Ammonia: [Empty]; 
Particulate matter: [Empty]; 
Volatile organic compounds: [Empty]. 

Source: State officials, as reported to GAO. 

[End of table] 

Specific examples of the types of regulations that the states have 
developed include the following: 

* Minnesota has established state emissions thresholds for hydrogen 
sulfide that apply to CAFOs. CAFO operators in the state must develop 
an air emissions control plan and must implement it if the Minnesota 
Pollution Control Agency detects elevated levels of hydrogen sulfide. 
According to state officials, once an operator reduces emissions, the 
agency re-monitors to ensure the emission levels remained below the 
state-established threshold.[Footnote 32] Minnesota may take legal 
action against CAFO operators violating this standard. For example, in 
June 2008, monitoring by the Minnesota Pollution Control Agency at a 
dairy operation recorded hydrogen sulfide levels above the state 
threshold and in cooperation with the State Attorney General, the 
agency, using state authorities, filed a lawsuit against the dairy's 
operator. 

* In 2003, California passed a law that authorized the state and local 
air districts to require animal feeding operations above a certain size 
to apply for clean air permits and develop a plan to decrease air 
emissions. For example, one air district in California--the San Joaquin 
Valley Air Pollution Control District with large clusters of animal 
feeding operations--developed a rule in 2006 to implement the law that 
required large animal feeding operations to apply for a permit that 
includes a plan for mitigating their emissions. According to air 
district officials, the district has implemented specific regulations 
for dairy animal feeding operations that require these operations to 
obtain five separate permits for components of their operations, 
including barns and land application of manure. The officials told us 
that these regulations were put in place, in part because the area is 
designated as a severe nonattainment area under the Clean Air Act and 
they are required to regulate a broader range of emission sources. 
According to state officials we spoke with, as a result of these more 
stringent state regulations, CAFOs in California may be relocating to 
other states--such as Texas and Iowa. 

Two Federal Court Decisions Have Affected EPA's and Some States' 
Ability to Regulate Water Pollutants Discharged by CAFOs: 

Two federal court decisions have affected EPA and some states' 
abilities to regulate CAFOs for water pollutants. The 2005 Waterkeeper 
Alliance Inc. v. EPA decision forced EPA to revise its 2003 rule for 
permitting CAFOs and abandon its approach of requiring all CAFO 
operators to obtain a permit. Although this court decision affected 
EPA's ability to regulate CAFOs, states' reaction to the Waterkeeper 
decision has varied: some states such as Minnesota continue to require 
all CAFOs to obtain permits while others such as Colorado have delayed 
developing new rules until EPA issues its final revised rule. In 
addition, the Supreme Court's 2006 decision--Rapanos v. United States-
-has made determination of Clean Water Act jurisdiction over certain 
types of waters more complex. According to EPA, this has required the 
agency to gather significantly more evidence to establish Clean Water 
Act jurisdiction in some enforcement cases. 

The Waterkeeper Decision Has Impacted EPA's Ability to Regulate CAFOs, 
but Has Not Had a Similar Impact on Some States: 

In its 2005 Waterkeeper decision, the U.S. Court of Appeals for the 
Second Circuit set aside a key provision of EPA's 2003 CAFO rule 
requiring every CAFO to apply for a NPDES permit. Under the 2003 rule, 
large numbers of previously unregulated CAFOs were required to apply 
for permits and would have been subject to monitoring and reporting 
requirements imposed by the permit as well as periodic inspections. 
According to EPA, the 2003 rule would have expanded the number of CAFOs 
requiring permits from an estimated 12,500 to an estimated 15,300, an 
increase of about 22 percent. According to EPA officials, when fully 
implemented, this requirement for all CAFOs with a potential to 
discharge to apply for permits would have provided EPA with more 
comprehensive information on the number and location of CAFOs and how 
they are operated and managed, thus allowing EPA to more effectively 
locate and inspect CAFOs nationwide. 

However, in 2003, both environmental and agricultural groups challenged 
EPA's 2003 rule. In the Waterkeeper case, environmental groups argued, 
among other things, that EPA's 2003 rule did not adequately provide for 
(1) public review and comment on a CAFO's nutrient management plan and 
(2) permitting authorities to review the CAFO's nutrient management 
plan. The court agreed with the environmental groups and instructed EPA 
to revise the rule accordingly. The agricultural groups challenged the 
2003 rule's CAFO permitting requirement, arguing that the agency 
exceeded its authority under the Clean Water Act by requiring CAFOs 
that were not discharging pollutants into federally regulated waters to 
apply for permits or demonstrate that they had no potential to 
discharge. The court also agreed with the agricultural groups and set 
aside the permitting requirements for CAFOs that did not actually 
discharge. Following the court's decision, many aspects of the 2003 
rule remained in effect, including EPA's revised regulatory definition 
of CAFOs and the expansion of the number of CAFOs needing permits by 
deleting a significant exception. 

In effect, the Waterkeeper decision returned EPA's permitting program 
to one in which CAFO operators are not required to apply for a NPDES 
permit unless they discharge, or propose discharging, into federally 
regulated waters. As a result, EPA must identify and prove that an 
operation has discharged or is discharging pollutants in order to 
require the operator to apply for a permit. To help identify 
unpermitted discharges from CAFOs, EPA officials stated that they have 
to rely on other methods that are not necessarily all-inclusive, such 
as citizens' complaints, drive-by observations, aerial flyovers, and 
state water quality assessments that identify water bodies impaired by 
pollutants associated with CAFOs. According to EPA officials, these 
methods have helped the agency identify some CAFOs that may be 
discharging as well as targeting inspections to such CAFOs. 

In response to the Waterkeeper decision, EPA proposed a new rule in 
June 2006 requiring that (1) only CAFO operators that discharge, or 
propose to discharge, apply for a permit; (2) permitting authorities 
review CAFO nutrient management plans and incorporate the terms of 
these plans into the permits; and (3) permitting authorities provide 
the public with an opportunity to review and comment on the nutrient 
management plans. According to EPA officials, the final rule is 
currently being reviewed by the Office of Management and Budget before 
it is formally published in the Federal Register. These officials said 
it is uncertain when the OMB review will be completed and the final 
rule issued. Estimates vary on how this rule, when implemented, will 
affect the number of CAFOs that will obtain a permit. EPA estimates 
that 25 percent fewer CAFOs will need to apply for a permit under the 
new rule than would have been required to apply for a permit under the 
2003 rule. In contrast, an association representing state water program 
officials believes that many fewer CAFOs than EPA estimates will 
voluntarily apply for a permit under the new 2006 rule, when it is 
finalized. 

The need to develop and implement a new rule that meets the Waterkeeper 
requirements has also resulted in delays in implementing the provisions 
of the 2003 rule that the Court upheld. Specifically, EPA has not yet 
implemented, among other things the expanded CAFO definitions, which 
cover operations such as dry-manure poultry operations. This is 
particularly significant since, according to a USDA official with 
extensive knowledge of the poultry industry and another agricultural 
expert that we spoke to, at least 90 percent of poultry operations use 
a dry-manure management system. An EPA Region 6 official told us that 
in Texas alone this expanded definition would result in about 1,500 
additional dry-manure poultry operations being covered under the new 
CAFO definition. 

Although the Waterkeeper decision has affected EPA's ability to 
regulate CAFOs' water pollutant discharges, this decision has not had 
the same impact on the ability of some of the states to regulate these 
operations. According to officials in the 47 states responding to our 
survey, the impact of the Waterkeeper decision on their ability to 
regulate water pollution from CAFOs has been mixed. As table 11 shows, 
the impacts of the Waterkeeper decision ranged from having little 
impact on state regulation of CAFOs to impairing state CAFO programs. 

Table 11: State Officials' Views of the Impact of the Waterkeeper 
Decision on Their CAFO Programs: 

Impact of Waterkeeper: Waterkeeper had little or no impact; 
Number of states reporting impact: 16. 

Impact of Waterkeeper: Reduced the number of CAFOs with permits; 
Number of states reporting impact: 15. 

Impact of Waterkeeper: Impaired state program; 
Number of states reporting impact: 10. 

Impact of Waterkeeper: Waiting for EPA to issue revised rule; 
Number of states reporting impact: 9. 

Impact of Waterkeeper: Prompted state legislation to require permits 
for CAFOs; 
Number of states reporting impact: 1. 

Source: GAO analysis of state official responses. 

Note: Some state officials identified more than one impact. 

[End of table] 

Officials from several of the states that told us that the Waterkeeper 
decision had little impact on their regulation of CAFOs, saying that 
this was primarily because their states had implemented CAFO 
regulations that were more stringent than those required under the 
Clean Water Act. For example, Minnesota officials stated that the 
Waterkeeper decision had no impact on their state's regulations because 
the state used its own authority to adopt regulations more stringent 
than EPA's regulations. Moreover, according to Minnesota officials, 
even after the Waterkeeper decision, the state has continued to require 
all CAFOs to obtain permits from the state environmental agency. 
Similarly, Kansas officials stated that the Waterkeeper decision had 
only minimal effects because the state has regulated CAFOs since the 
1960s. 

However, 34 states indicated that the Waterkeeper decision directly 
affected their state programs. Officials from 15 states told us that 
the number of CAFOs that had obtained permits since the Waterkeeper 
decision had decreased although none provided us with numbers on what 
this decrease had been. Similarly, officials in 10 states told us that 
the Waterkeeper decision had impaired their state's ability to regulate 
CAFOs because it discredited the program, created confusion or 
uncertainty, or made it difficult for them to determine which 
operations needed a permit. For example, according to the state 
official responsible for Indiana's CAFO permitting program, although 
the state has had a CAFO permitting program since 1971, it adopted 
EPA's 2003 CAFO Rule because the rule was more protective. However, 
when the Waterkeeper decision set aside portions of the 2003 rule, this 
official told us that the decision, in effect, discredited the state's 
regulatory program. In addition, officials from nine states who are 
responsible for their state's permitting program told us that their 
programs remain in limbo while they wait for EPA to issue its final 
revised rule. These state officials, including officials in Colorado, 
said that they will update their state rules once EPA's final rule is 
issued. 

Finally, state water pollution control officials have expressed some 
concerns that EPA's new 2006 rule will place a greater administrative 
burden on states than the 2003 rule would have. In an August 2006 
letter to EPA, the Association of State and Interstate Water Pollution 
Control Administrators noted that the "reactive" enforcement that EPA 
will now follow will require permitting authorities to significantly 
increase their enforcement efforts to achieve the level of 
environmental benefit that would have been provided by the 2003 rule. 
These officials believe that requiring EPA and the states to identify 
CAFOs that actually discharge pollutants into federally regulated water 
bodies will consume more resources than requiring all CAFOs to apply 
for a permit. 

The Rapanos Decision Has Affected EPA's Overall Ability to Regulate 
Pollutants Entering Federally Regulated Waters: 

The Supreme Court's 2006 Rapanos decision has also affected EPA's 
enforcement of the Clean Water Act because the agency believes that it 
must gather significantly more evidence to establish which waters are 
subject to the act's permitting requirements. At issue in the Rapanos 
decision was whether the Clean Water Act's wetlands permitting program 
applied to four specific wetlands that were adjacent to non-navigable 
tributaries of traditional navigable waters. The Court rejected the 
standards applied by the lower courts in determining whether wetlands 
at issue fell under the act's jurisdiction and, therefore, could be 
subject to permitting requirements. Although a majority of the justices 
rejected the standards applied by the lower courts, a majority could 
not agree on how to determine which waters would fall under the act's 
jurisdiction, and thus how far EPA could reach to regulate discharges 
of pollutants under the act. 

Although the Rapanos case arose in the context of a different permit
program, the scope of EPA’s pollutant discharge permit program
originates in the same Clean Water Act definition that was discussed in 
the decision. According to EPA enforcement officials, the agency may 
now be less likely to seek enforcement against a CAFO that it believes 
is discharging pollutants into a water body because it may be more 
difficult to prove that the water body is federally regulated. 
According to EPA officials, as a result of the Rapanos decision, the 
agency must now spend more resources developing an enforcement case 
because the agency must gather proof that the CAFO not only has 
illegally discharged pollutants, but that those discharges ultimately 
entered a federally regulated water body. These officials told us that 
the farther a CAFO is from a regulated water body, the more evidence 
they will need to prove that the discharges entered that water body. To 
ensure “nationwide consistency, reliability, and predictability in 
their administration of the statute,” EPA has issued national guidance 
to clarify the agency’s responsibilities in light of the Supreme 
Court’s decision. However, in a March 4, 2008, memorandum, EPA’s 
Assistant Administrator for Enforcement and Compliance Assurance stated 
that the Rapanos decision and EPA’s guidance has resulted in 
significant adverse impacts to the clean water enforcement program. 
According to the memorandum, the Rapanos decision and guidance 
negatively affected approximately 500 enforcement cases, including as 
many as 187 cases involving NPDES permits. In May 2007, Members of 
Congress, in both the House and Senate, introduced a bill entitled the 
Clean Water Restoration Act of 2007 to clearly define the scope of the 
Clean Water Act. As of August 2008, neither bill had been reported out 
of committee. 

Conclusions: 

For more than 30 years, EPA has regulated CAFOs under the Clean Water 
Act and during this time it has amassed a significant body of knowledge 
about the pollutants discharged by animal feeding operations and the 
potential impacts of these pollutants on human health and the 
environment. Despite its long-term regulation of CAFOs, EPA still lacks 
comprehensive and reliable data on the number, location, and size of 
the operations that have been issued permits and the amounts of 
discharges they release. As a result, EPA has neither the information 
it needs to assess the extent to which CAFOs may be contributing to 
water pollution, nor the information it needs to ensure compliance with 
the Clean Water Act. More recently, EPA has also begun to address 
concerns about air pollutants that are emitted by animal feeding 
operations. The Nationwide Air Emissions Monitoring Study, along with 
EPA's plans to develop air emissions estimating protocols, are 
important steps in providing much needed information on the amount of 
air pollutants emitted from animal feeding operations. However, 
questions about the sufficiency of the sites selected for the air 
emissions study and the quantity and quality of the data being 
collected could undermine EPA's efforts to develop air emissions 
protocols by 2011 as planned. Finally, while the study and resulting 
protocols are important first steps, a process-based model that more 
accurately predicts the total air emissions from an animal feeding 
operation is still needed. While EPA has indicated it intends to 
develop such a model, it has not yet established a strategy and 
timeline for this activity. 

Recommendations for Executive Action: 

In order to more effectively monitor and regulate CAFOs, we recommend 
that the Administrator of the Environmental Protection Agency should 
complete the agency's effort to develop a national inventory of 
permitted CAFOs and incorporate appropriate internal controls to ensure 
the quality of the data. 

In order to more effectively determine the extent of air emissions from 
animal feeding operations, the Administrator of the Environmental 
Protection Agency should: 

* reassess the current data collection efforts, including its internal 
controls, to ensure that the National Air Emissions Monitoring Study 
will provide the scientific and statistically valid data that EPA needs 
for developing its air emissions protocols; 

* provide stakeholders with information on the additional data that it 
plans to use to supplement the National Air Emissions Monitoring Study; 
and: 

* establish a strategy and timetable for developing a process-based 
model that will provide more sophisticated air emissions estimating 
methodologies for animal feeding operations. 

Agency Comments and Our Evaluation: 

We provided a draft of this report for review and comment to the EPA 
and the Secretary of USDA. We received written comments from EPA. USDA 
did not provide written comments, but did provide technical comments 
and clarifications, which we incorporated, as appropriate. 

EPA partially concurred with our conclusions and recommendations. In 
its written comments, EPA acknowledged that currently no national 
inventory of permitted CAFOs exists. The agency stated that it is 
currently working with its regions and the states to develop and 
implement a new national data system to collect and record facility- 
specific information on permitted CAFOs. We have revised our 
recommendation to reflect the actions that EPA has underway. In 
response to our recommendations that EPA reassess the current data 
collection effort, EPA stated that the agency has developed a quality 
assurance plan for the study and is continuously evaluating the 
National Air Emissions Monitoring Study. We are aware that EPA has 
developed a quality assurance plan for the data collected during the 
study. However, our recommendation also reflects other concerns with 
the study. For example, the monitoring sites selected may not represent 
a statistically valid sample or animal feeding operations that account 
for the differences in climatic conditions, manure-handling methods, 
and density of operations; and the study does not address other sources 
that can contribute significantly to emissions from animal feeding 
operations. EPA did not address these issues in its comments. 
Therefore, we continue to believe that EPA should reassess the ongoing 
effort to ensure that the study, as currently structured, will provide 
the data that EPA needs. 

In response to our recommendation that the agency identify the 
information that it plans to use to supplement the National Air 
Emissions Monitoring Study, EPA stated that it cannot yet identify the 
data that it will use to augment the data collected during the 
monitoring study. However, the agency indicated that it has begun 
discussions with USDA to identify ongoing research that is focused on 
agricultural air emissions and gaps that may still exist, but did not 
provide any additional information on when it plans to identify the 
supplemental data that it plans to use to augment the monitoring study. 
Until it does so, neither EPA nor stakeholders can be assured that 
these data, in combination with the emissions study data, will enable 
EPA to develop the planned protocols. The agency also agreed with our 
recommendation to establish a strategy and timetable for developing a 
process-based model and said that it has begun to evaluate what is 
needed to develop such a model. However, the agency did not provide any 
information on when it expects to complete plans for developing a 
process-based model. EPA also provided technical comments, which we 
have incorporated, as appropriate. EPA's written comments are provided 
in appendix V. 

As agreed with your offices, 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 to interested 
congressional committees, the Administrator of the Environmental 
Protection Agency, the Secretary of the United States Department of 
Agriculture and other interested parties. We also will make copies 
available to others upon request. In addition, the report will be 
available at no charge on GAO's Web site at [hyperlink, 
http://www.gao.gov]. 

If you or your staff have any questions regarding this report, please 
contact me at (202) 512-3841 or mittala@gao.gov. Contact points for our 
Office of Congressional Relations and Public Affairs may be found on 
the last page of this report. Key contributors to this report are 
listed in appendix VI. 

Signed by: 

Anu Mittal: 
Director, Natural Resources and Environment: 

List of Requesters: 

The Honorable John D. Dingell: 
Chairman: 
Committee on Energy and Commerce: 
House of Representatives: 

The Honorable James L. Oberstar: 
Chairman: 
Committee on Transportation and Infrastructure: 
House of Representatives: 

The Honorable Gene Green: 
Chairman: 
Subcommittee on Environment and Hazardous Materials: 
Committee on Energy and Commerce: 
House of Representatives: 

The Honorable Eddie Bernice Johnson: 
Chairwoman: 
Subcommittee on Water Resources and Environment: 
Committee on Transportation and Infrastructure: 
House of Representatives: 

The Honorable Hilda L. Solis: 
House of Representatives: 

[End of section] 

Appendix I: Objectives, Scope, and Methodology: 

For this report we were asked to determine the (1) trends in 
concentrated animal feeding operations (CAFOs) over the past 30 years; 
(2) amount of waste they generate; (3) findings of recent key academic, 
industry, and government research of the potential impacts of CAFOs on 
human health and the environment, and the extent to which the 
Environmental Protection Agency (EPA) has assessed the nature and 
severity of these identified impacts; (4) progress that EPA and states 
have made in regulating and controlling the air emissions of, and in 
developing protocols to measure, air pollutants from CAFOs that could 
affect air quality; and (5) extent to which recent court decisions have 
affected EPA and the states' ability to regulate CAFO discharges that 
impair water quality. 

In conducting our work, we reviewed laws and regulations and federal 
and state agencies' documents. We met with officials from EPA, the U.S. 
Department of Agriculture (USDA), the National Pork Producers Council, 
the National Pork Board, the National Cattlemen's Beef Association, the 
Environmental Integrity Project (a nonpartisan, nonprofit environmental 
advocacy group), the Sierra Club, California Association of Irritated 
Residents, Waterkeeper Alliance, Iowa Citizens for Community 
Improvement, Environmental Defense, National Association of Clean Air 
Agencies, Association of State and Interstate Water Pollution Control 
Administrators, as well as state officials. The National Chicken 
Council did not respond to our requests for information. Additionally, 
we visited CAFOs in eight states: Arkansas, California, Colorado, Iowa, 
Maryland, Minnesota, North Carolina, and Texas. We chose these states 
because they were geographically dispersed and contained numerous CAFOs 
representing multiple types of animals. 

For our analysis of trends in CAFOs over the past 30 years, we used 
USDA's Census of Agriculture data. We assessed the reliability of these 
data by reviewing USDA's documentation on the development, 
administration, and data quality program for the Census of Agriculture. 
We also electronically tested the data used in this study to determine 
if there were any missing data or anomalies in the dataset. 
Furthermore, we compared the results of our nationwide results for each 
year by animal sector to USDA's published reports. On the basis of 
these assessments, we determined the data to be sufficiently reliable 
for the purposes for which it was used in this report. In addition, 
respecting USDA's requirement to protect the privacy of individual 
farmers responding to the Census of Agriculture surveys, we conducted 
these analyses at USDA and worked with USDA to review our results and 
verify that no single operation could be identified from our analysis. 
[Footnote 33] 

From USDA's Census of Agriculture data, we analyzed the most recent 
data available for large farms raising animals from 1974 through 2002. 
[Footnote 34] We used these data on large farms as a proxy for CAFOs 
because no federal agency collects consistent data on these types of 
operations. USDA has periodically collected data on farms nationwide 
using the Census of Agriculture survey. Prior to 1982, these surveys 
were conducted every four years; whereas since 1982, the agency has 
administered the survey every five years (the most recent survey 
results, conducted in 2007, will not be available until February 2009). 
In analyzing Census data prior to 1982, we found that the categories 
reported by USDA were not consistent with EPA's minimum size threshold 
for large CAFOs: 2,500 hogs, 700 dairy or milk cows, 55,000 turkeys, 
1,000 beef cattle, 82,000 layers, and 125,000 broilers.[Footnote 35] 
For instance, the largest farm categories USDA reported for broilers 
prior to 1982 was farms with sales of 100,000 and more. Since sales 
data must be converted to an inventory number, we had to make 
adjustments for production cycles to determine the number of animals on 
a farm per day.[Footnote 36] Broiler farms complete six production 
cycles per year therefore, when we divided the USDA provided number of 
100,000 in broiler sales by 6 to account for the total number of 
possible production cycles, the USDA reported broiler sales represent a 
farm with an inventory of about 17,000 broilers. Farms of this size are 
much smaller than the 125,000 broiler CAFO threshold defined by EPA. 
Similarly, categories for farms raising other types of animals, in the 
pre-1982 USDA data, were also different than the EPA CAFO definitions 
for these types of operations. As a result, we used the time frame of 
1982 through 2002 because USDA could provide us with detailed 
electronic data that allowed us to apply EPA's CAFO thresholds to 
determine the trends in the overall number of large farms that raised 
animals and could be potentially considered a CAFO. For broilers and 
layers/pullets,[Footnote 37] we used EPA's CAFO minimum size threshold 
for dry-litter manure handling systems because these systems represent 
the majority of poultry operations. These thresholds are larger than 
for those poultry operations that have liquid manure handling systems. 

Because USDA does not report the average number of animals on a farm, 
we used USDA Census of Agriculture inventory, sales, and inventory plus 
sales data for this purpose. The choice of using inventory only, sales 
only, or inventory and sales data for a particular animal type depended 
on the wording of Census survey questions during the years we analyzed. 
When only sales data or inventory plus sales data were used, we 
adjusted these data using the appropriate USDA formulas to determine 
the average number of animals on a farm.[Footnote 38] When both 
inventory and sales were used for an animal type, we applied an 
approved USDA approach to determine the average number of animals on a 
farm. As a result, we made the following adjustments for each animal 
type: 

* For beef cattle, USDA only collected sales data for 1982 through 
1997. As a result, for beef cattle, we used sales of cattle on feed 
(2002 survey) or sales of fattened cattle (1982 through 1997 surveys) 
adjusted for the number of production cycles. This increased the 
likelihood that we were including cattle raised on CAFOs instead of 
operations that allow the cattle to graze on pastureland. 

* For dairy cows, we used the inventory of animals as of December 31 
for each Census year since these animals are maintained to produce milk 
and not specifically for slaughter. For dairy cows, we included the 
categories: lactating and nonlactating cows. 

* For hogs, the Census of Agriculture reported both inventory and sales 
data for hogs and pigs.[Footnote 39] These data were not reported by 
either the weight or age, so we used the total for all hogs and pigs of 
all ages. We used both the inventory and sales data for hogs and 
adjusted for the number of production or finish cycles. Hogs may be 
sold more than once because of the practice of selling feeder pigs at 
about 10-12 weeks of age to producers to be grown to typical slaughter 
size. For example, in 1997, about 25 percent of all hog and pig sales 
reported on the Census of Agriculture were feeder pigs.[Footnote 40] We 
adjusted the hog data to factor out these multiple sales. 

* For layers, we used survey responses of inventory as of December 31 
for layers 20 weeks old and older plus pullets for laying flock 
replacement. 

* For broilers, we used inventory and sales data from the categories: 
broilers, fryers, capons, roaster and other chickens raised for meat. 

* For turkeys, both inventory and sales data were used and included 
both hens and tom turkeys. 

We also reviewed EPA's data on the number of CAFOs that had been issued 
permits--these data are either collected by EPA's regional offices or 
from the states--for the period 2003 to 2008. We assessed the accuracy 
and reliability of these data by interviewing officials in 47 states 
and we asked them to verify the information that EPA had for the 
numbers of CAFOs permitted in their state.[Footnote 41] Based on the 
information we obtained from the state officials, we determined that 
EPA's data for permitted CAFOs was not reliable and could not be used 
to identify trends in permitted CAFOs over the 5-year period. 

To identify the amount of manure, including urine, a large CAFO is 
estimated to generate for each animal type, we used EPA's thresholds 
for the minimum number of animals that constitute a CAFO. To illustrate 
the size of a "typical" large farm for each animal type, we used the 
median for a large-sized farm. We used the median instead of the mean 
because we believe it provides a more representative measure for a 
typical large farm. We also present information on farms at the 75th 
percentile of all large farms for a particular animal type to represent 
larger farms.[Footnote 42] 

To estimate the amount of manure produced by each type of animal, we 
used engineering standards for manure production cited by the American 
Society of Agricultural and Biological Engineers (ASABE).[Footnote 43] 
These standards report the total amount of manure over the production 
cycle for hogs, beef cattle, turkeys, and broilers. In order to 
estimate the average pounds of manure per day, we divided the total 
manure produced over the production cycle by the number of days in the 
production cycle. Further, we converted the pounds of manure into tons 
of manure per farm per year. We adjusted the manure calculations for 
the following animal types: 

* For layers, the standards provided the average daily pounds of manure 
produced by layers. We multiplied the average pounds of manure per day 
times the average number of animals times 365 days to get manure 
produced per year. 

* For broilers, we determined the average daily pound of manure from 
the information provided in the standards. We multiplied the average 
pound of manure per day times the average number of animals times 365 
days to get manure per year. 

* For dairy cows, the standards provided the average daily pounds of 
manure produced by dairy cows. We multiplied the average pounds of 
manure per day times the average number of animals times 365 days to 
get manure per year. However, we adjusted the data to take into account 
the typical percentage of cows that are either lactating or dry 
(nonlactating) and applied the different amounts of manure produced by 
each type of dairy cow. 

* For turkeys, we adjusted the turkey statistics based on the ratio of 
hens to tom turkeys raised on farms and applied different amounts of 
manure due to the different sizes of the animals. 

* For hogs, the manure standards report manure produced by hogs 
covering a specific stage of production: feeder-pig-to-finish pigs-- 
beginning with a pig weighing on average about 27 pounds and resulting 
in a hog weighing 154 pounds. Estimates for other hog operation types 
such as nursery, farrow to feeder, and farrow to finish would therefore 
differ. Census of Agriculture data for 2002 indicate that about a third 
of all hogs sold were from the grow-to-finish (called finish only on 
the survey) operation type. The ASABE manure standards for this type of 
operation use 154 pounds as the finish weight. However, USDA reports 
that typical hog finish (slaughter) weights at the time of the 2002 
Census were about 260 pounds. For hogs only, we adjusted the ASABE 
manure estimates by 1.7 to account for the larger finish weights 
reported by USDA. We believe this is a conservative adjustment because 
manure produced by hogs weighing 154 to 260 pounds will be the maximum 
amount per day that ASABE used to calculate the average pounds produced 
for the hogs growing from about 27 pounds to 154 pounds. 

* For beef cattle, we used the manure standard for "beef-finishing 
cattle." This standard is for cattle fattened from about 740 pounds to 
about 1,200 pounds at marketing. Beef cattle (listed as cattle on feed) 
data from the Census are for cattle sold for slaughter and thus similar 
in weight to those for the standard. The reported manure results for 
beef cattle are for operations of this type only. 

In addition, the number of days on feed for hogs, turkeys, and broilers 
used for the ASABE manure standards does not take into account time 
between herds or flocks entering and leaving an operation; therefore, 
we adjusted the manure generated to account for the time between 
cycles. 

We recognize that all amounts of manure reported are estimates because 
amounts of manure per animal type vary by feeding programs, feeds used, 
climatic conditions, production techniques, and animal genetics, among 
other things. As feeds, animal genetics, and production techniques 
change in the future, these estimates might change--and may have 
changed since 2002--but USDA did not provide specific information on 
what changes have occurred and how those changes may have impacted the 
manure production on farms. We did not estimate the ability of the farm 
or surrounding farms to assimilate the manure if applied to pastures 
and crop land nor did we take into account various technologies to 
process and/or convert manure. Reported estimates of manure are for 
amounts produced. We did not determine whether these amounts were 
discharged into the air or streams and wetlands. Manure harvested from 
CAFOs for application to land might be less than that excreted by 
animals because of shrinkage due to evaporation. 

To provide a perspective of the amount of wastes generated by these 
large farms, we compared them with the amount of human sanitary waste 
generated in various cities. We selected certain cities on the basis of 
their population, as reported by the U.S. Census Bureau's Population 
Estimates for 2002, and calculated the amount of sanitary waste 
generated by the human population of those cities by applying estimates 
for human sanitary waste production. Human sanitary waste includes 
feces and urine but does not include wastes such as water from showers, 
washing dishes and clothes, and flushing toilets. We found two sources 
of information for average daily human sanitary waste.[Footnote 44] 
Because these sources provided different estimates (2.68 and 4.76 
pounds per person per day), we averaged the two amounts to use in our 
calculations of human sanitary waste produced for cities (3.72 pounds 
per person per day). All amounts of human sanitary waste reported are 
estimates because amounts will vary based on differences in age, 
dietary habits, activity levels, and climatic conditions, among other 
things. Human sanitary waste is a small portion of human discharge into 
sewage systems. Our reported estimates of human sanitary waste for a 
city are illustrative only and are not intended to be estimates of 
actual human sanitary waste entering a particular city's waste 
treatment system. These estimates are for a population the size of 
selected cities assuming that the residents do not commute outside the 
city boundaries and that nonresidents do not enter the city for work or 
other reasons. 

To identify the findings of recent key academic, industry, and 
government research on the potential impacts of CAFOs on human health 
and the environment, and the extent to which EPA has assessed the 
nature and severity of such impacts, we reviewed EPA's 2003 CAFO rule 
(for water impact studies) and the findings and supporting documents of 
the National Academy of Sciences study on air emissions from animal 
feeding operations (for air impact studies).[Footnote 45] In addition, 
we: 

* conducted library, online journal and Internet searches to identify 
recent studies; 

* consulted with EPA, USDA, state agencies, industry groups, 
environmental groups, and academia to help identify additional studies; 
and: 

* identified studies through citations in previously identified 
studies. 

We only included in our review studies that (1) were peer-reviewed or 
produced by a federal agency, (2) were new and original research 
completed since 2002, (3) had a clearly defined methodology, and (4) 
identified pollutants found in animal waste and/or their impacts. 
Through this effort, we found over 200 studies and identified 68 
studies that examined air and water quality issues associated with 
animal waste and met our criteria. We also classified these studies 
according to whether they: 

* found a direct link between pollutants from animal waste and impacts 
on human health or the environment; 

* did not find any impacts on human health or the environment from 
pollutants from animal waste; 

* found an indirect link between animal waste and human health or 
environmental impacts; or: 

* measured pollutants from animal waste otherwise known to cause human 
health or environmental impacts. 

The classification for each study involved two reviewers. If the 
reviewers disagreed on the classification, they turned to a third 
reviewer for resolution. Finally, we compared the findings from these 
studies with EPA assessments to date and interviewed EPA officials 
regarding these assessments. 

To determine the progress that EPA and states have made in regulating 
and controlling the air emissions of, and in developing protocols to 
measure, air pollutants from CAFOs, we reviewed relevant documents, 
interviewed officials responsible for the ongoing air monitoring study 
and visited several National Air Emissions Monitoring Study sites in 
North Carolina. Additionally, we interviewed industry and environmental 
groups, the umbrella association for state and local clean air 
agencies, and citizen groups about how EPA air emissions protocols 
affect them. Finally, we contacted state CAFO officials in all 50 
states to determine which states had developed air emission regulations 
applicable to CAFOs. Officials in 47 states responded.[Footnote 46] 
These 47 states account for an estimated 99 percent of large animal 
feeding operations that could be defined as CAFOs under EPA's 2003 
rule. 

Finally, to determine the extent to which recent court decisions have 
affected EPA and the states' ability to regulate CAFO discharges that 
impair water quality, we examined recent federal decisions, including 
the Waterkeeper Alliance Inc. v. EPA (Waterkeeper), and the Supreme 
Court's 2006 decision in Rapanos v. United States. We interviewed EPA 
officials about how these court decisions have affected their 
regulations. To better understand the bases for the lawsuits and what 
has occurred since the court decisions, we contacted plaintiffs and 
defendants involved in Waterkeeper and other court cases, including 
industry and environmental groups. To identify the impact of these 
cases on states regulations, we contacted state CAFO officials in all 
50 states to determine how the Waterkeeper decision affected their 
regulations. We asked the states if the Waterkeeper decision had 
affected their state's CAFO program. Using the responses we received 
from 47 states, we conducted content analyses and classified them into 
six categories, including if the decision (1) had little impact on the 
state program, (2) caused the state to wait for EPA guidance (3) 
impaired the state program, (4) proactively changed legislation, (5) 
reduced the number of CAFOs with permits, or (6) other. Some officials 
identified more than one impact. The responses in the "other" category 
included such responses as "not applicable," "because the state does 
not have delegated authority," and "we have spent a large amount of 
time studying the ruling and commenting on EPA proposed rules that were 
developed to satisfy the ruling." 

We conducted this performance audit between July 2007 and August 2008, 
in accordance with generally accepted government auditing standards. 
Those standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe that 
the evidence obtained provides a reasonable basis for our findings and 
conclusions based on our audit objectives. 

[End of section] 

Appendix II: EPA's Definition of Concentrated Animal Feeding 
Operations: 

EPA's National Pollutant Discharge Elimination System (NPDES) permit 
program regulates the discharge of pollutants from point sources to 
waters of the United States. The Clean Water Act defines point sources 
to include CAFOs. To be considered a CAFO, a facility must first be 
defined as an animal feeding operation, which is a lot or facility 
(other than an aquatic animal production facility) where the following 
conditions are met: 

* Animals have been, are, or will be stabled or confined and fed or 
maintained for a total of 45 days or more in any 12-month period. 

* Crops, vegetation, forage growth, or post-harvest residues are not 
sustained in the normal growing season over any portion of the lot or 
facility. 

Generally CAFOs must meet the above definition of an animal feeding 
operation and stable or confine a certain minimum number of animals at 
the operation. EPA classifies CAFOs as large, medium, or small, based 
on size. Table 12 shows the number of animals at a farm that meet EPA's 
definition of a large, medium, and small CAFO. 

Table 12: EPA Designation of Large, Medium, and Small CAFOs for Various 
Size Thresholds by Animal Type: 

Animal type: Cattle or cow/calf pairs; 
Size thresholds (number of animals): Large CAFOs: 1,000 or more; 
Size thresholds (number of animals): Medium CAFOs[A]: 300 - 999; 
Size thresholds (number of animals): Small CAFOs[B]: less than 300. 

Animal type: Mature dairy cows; 
Size thresholds (number of animals): Large CAFOs: 700 or more; 
Size thresholds (number of animals): Medium CAFOs[A]: 200 - 699; 
Size thresholds (number of animals): Small CAFOs[B]: less than 200. 

Animal type: Veal calves; 
Size thresholds (number of animals): Large CAFOs: 1,000 or more; 
Size thresholds (number of animals): Medium CAFOs[A]: 300 - 999; 
Size thresholds (number of animals): Small CAFOs[B]: less than 300. 

Animal type: Swine (weighing over 55 pounds); 
Size thresholds (number of animals): Large CAFOs: 2,500 or more; 
Size thresholds (number of animals): Medium CAFOs[A]: 750 - 2,499; 
Size thresholds (number of animals): Small CAFOs[B]: less than 750. 

Animal type: Swine (weighing less than 55 pounds); 
Size thresholds (number of animals): Large CAFOs: 10,000 or more; 
Size thresholds (number of animals): Medium CAFOs[A]: 3,000 - 9,999; 
Size thresholds (number of animals): Small CAFOs[B]: less than 3,000. 

Animal type: Horses; 
Size thresholds (number of animals): Large CAFOs: 500 or more; 
Size thresholds (number of animals): Medium CAFOs[A]: 150 - 499; 
Size thresholds (number of animals): Small CAFOs[B]: less than 150. 

Animal type: Sheep or lambs; 
Size thresholds (number of animals): Large CAFOs: 10,000 or more; 
Size thresholds (number of animals): Medium CAFOs[A]: 3,000 - 9,999; 
Size thresholds (number of animals): Small CAFOs[B]: less than 3,000. 

Animal type: Turkeys; 
Size thresholds (number of animals): Large CAFOs: 55,000 or more; 
Size thresholds (number of animals): Medium CAFOs[A]: 16,500 - 54,999; 
Size thresholds (number of animals): Small CAFOs[B]: less than 16,500. 

Animal type: Laying hens or broilers (liquid manure handling systems); 
Size thresholds (number of animals): Large CAFOs: 30,000 or more; 
Size thresholds (number of animals): Medium CAFOs[A]: 9,000 - 29,999; 
Size thresholds (number of animals): Small CAFOs[B]: less than 9,000. 

Animal type: Chickens other than laying hens (other than a liquid 
manure handling system); 
Size thresholds (number of animals): Large CAFOs: 125,000 or more; 
Size thresholds (number of animals): Medium CAFOs[A]: 37,500 - 124,999; 
Size thresholds (number of animals): Small CAFOs[B]: less than 37,500. 

Animal type: Laying hens (other than a liquid manure handling system); 
Size thresholds (number of animals): Large CAFOs: 82,000 or more; 
Size thresholds (number of animals): Medium CAFOs[A]: 25,000 - 81,999; 
Size thresholds (number of animals): Small CAFOs[B]: less than 25,000. 

Animal type: Ducks (other than a liquid manure handling system); 
Size thresholds (number of animals): Large CAFOs: 30,000 or more; 
Size thresholds (number of animals): Medium CAFOs[A]: 10,000 - 29,999; 
Size thresholds (number of animals): Small CAFOs[B]: less than 10,000. 

Animal type: Ducks (liquid manure handling systems); 
Size thresholds (number of animals): Large CAFOs: 5,000 or more; 
Size thresholds (number of animals): Medium CAFOs[A]: 1,500 - 4,999; 
Size thresholds (number of animals): Small CAFOs[B]: less than 1,500. 

Source: EPA. 

[A] Must also meet one of two "method of discharge" criteria to be 
defined as a CAFO or may be designated. 

[B] May be designated as a CAFO on a case-by-case basis. 

[End of table] 

In addition to size, EPA uses the following criteria to determine if a 
CAFO operator needs to apply for a NPDES permit. 

* A large CAFO confines at least the number of animals described in 
table 12. 

* A medium CAFO falls within the size range in table 12 and either: 

* discharged pollutants into federally regulated waters through a 
manmade ditch, flushing system, or similar manmade device; 

* discharged pollutants directly into federally regulated waters that 
originate outside of and pass over, across, or through the facility or 
otherwise come into contact with animals confined in the operation; or: 

* is designated as a CAFO by the permitting authority as a significant 
contributor of pollutants. 

* A small CAFO confines the number of animals described in table 12 and 
has been designated as a CAFO by the permitting authority as a 
significant contributor of pollutants. 

[End of section] 

Appendix III: Nationwide Trends in the Number of All Animal Farms and 
the Number of Animals Raised on Large Farms, 1982-2002: 

This appendix provides our analysis of USDA's data for trends on the 
number of all animal farms and the number of animals raised on large 
farms per day for all animal types for the period from 1982 through 
2002. 

Table 13: Nationwide Trends in the Number of All Farms That Raise 
Animals for All Animal Types, 1982 through 2002: 

Type of animal farm: Beef cattle[A]; 
1982: 215,465; 
1987: 173,961; 
1992: 133,795; 
1997: 99,654; 
2002: 98,061; 
Percentage change, 1982-2002: (54). 

Type of animal farm: Dairy cow; 
1982: 277,762; 
1987: 202,068; 
1992: 155,339; 
1997: 116,874; 
2002: 91,989; 
Percentage change, 1982-2002: (67). 

Type of animal farm: Hog[B]; 
1982: 347,699; 
1987: 256,595; 
1992: 202,811; 
1997: 114,289; 
2002: 89,542; 
Percentage change, 1982-2002: (74). 

Type of animal farm: Layer; 
1982: 218,114; 
1987: 146,056; 
1992: 89,507; 
1997: 74,073; 
2002: 104,974; 
Percentage change, 1982-2002: (52). 

Type of animal farm: Broiler; 
1982: 52,890; 
1987: 41,097; 
1992: 31,427; 
1997: 30,979; 
2002: 41,572; 
Percentage change, 1982-2002: (21). 

Type of animal farm: Turkey; 
1982: 24,701; 
1987: 19,195; 
1992: 13,767; 
1997: 12,129; 
2002: 16,999; 
Percentage change, 1982-2002: (31). 

Type of animal farm: Total of all animal farms[C]; 
1982: 1,136,631; 
1987: 838,972; 
1992: 626,646; 
1997: 447,998; 
2002: 443,137; 
Percentage change, 1982-2002: (61). 

Source: GAO analysis of USDA data. 

Notes: The phrase "all animal types" refers to the following animals: 
beef cattle, dairy cows, hogs, layers, broilers, and turkeys. 

The criteria for a large farm varied by animal type, consistent with 
EPA's CAFO thresholds, and represent the average number of animals on a 
farm per day. 

[A] Beef cattle includes only cattle on feed, not grazing on pasture, 
and sold weighing 500 pounds or more. 

[B] Hogs include swine of all sizes from birth to market size. 

[C] The number of large farms for all animal types is the total of 
large farms for each animal type and may include some farms multiple 
times if they were considered large for more than one animal type. 

[End of table] 

Table 14: Nationwide Trends in the Number of Animals Raised on Large 
Farms per Day for All Animal Types, 1982 through 2002: 

Type of animal farm: Beef cattle[A]; 
1982: 6,601,928; 
1987: 7,368,109; 
1992: 7,533,708; 
1997: 8,598,508; 
2002: 8,677,892; 
Percentage change, 1982-2002: 31. 

Type of animal farm: Dairy cow; 
1982: 632,583; 
1987: 860,878; 
1992: 1,300,616; 
1997: 2,049,814; 
2002: 3,183,086; 
Percentage change, 1982-2002: 403. 

Type of animal farm: Hog[B]; 
1982: 4,176,477; 
1987: 6,275,200; 
1992: 12,133,231; 
1997: 32,412,839; 
2002: 47,789,951; 
Percentage change, 1982-2002: 1,044. 

Type of animal farm: Layer; 
1982: 160,005,126; 
1987: 212,871,326; 
1992: 229,959,901; 
1997: 263,660,262; 
2002: 304,500,225; 
Percentage change, 1982-2002: 90. 

Type of animal farm: Broiler; 
1982: 52,140,827; 
1987: 102,198,894; 
1992: 170,873,560; 
1997: 298,222,567; 
2002: 457,461,691; 
Percentage change, 1982-2002: 777. 

Type of animal farm: Turkey; 
1982: 33,443,754; 
1987: 52,905,796; 
1992: 62,042,552; 
1997: 73,029,156; 
2002: 68,417,853; 
Percentage change, 1982-2002: 105. 

Type of animal farm: Total of all animal types[C]; 
1982: 257,000,695; 
1987: 382,480,203; 
1992: 483,843,568; 
1997: 677,973,146; 
2002: 890,030,698; 
Percentage change, 1982-2002: 246. 

Source: GAO analysis of USDA data. 

Notes: The phrase "all animal types" refers to the following animals: 
beef cattle, dairy cows, hogs, layers, broilers, and turkeys. 

The criteria for a large farm varied by animal type, consistent with 
EPA's CAFO thresholds, and represent the number of animals on a farm 
per production day. 

[A] Beef cattle include only cattle on feed, not grazing on pasture, 
and sold weighing 500 pounds or more. 

[B] Hogs include swine of all sizes from birth to market size. 

[C] The number of large farms for all animal types is the total of 
large farms for each animal type and may include some farms multiple 
times if they were considered large for more than one animal type. 

[End of table] 

[End of section] 

Appendix IV: Government-Sponsored or Peer-Reviewed Studies Completed 
Since 2002 on the Impacts of Pollutants from Animal Waste: 

Studies showing a direct impact: 

Study: Ankley, Gerald T., Kathleen M. Jensen, Elizabeth A. Makynen, 
Michael D. Kahl, Joseph J. Korte, Michael W. Hornung, Tala R. Henry, 
Jeffrey S. Denny, Richard L. Leino, Vickie S. Wilson, et al. "Effects 
of the Androgenic Growth Promoter 17-trenbolone on Fecundity and 
Reproductive Endocrinology of the Fathead Minnow." Environmental 
Toxicology and Chemistry. Vol. 22, no. 6 (2003):1,350-1,360; 
Sponsor[A]: EPA, University of Minnesota; 
Medium: Water; 
Pollutant(s): Hormones; 
Impact: Fertility of fish was significantly reduced by hormones and 
female fish developed male sex characteristics. 

Study: Clark, Clifford G., Lawrence Price, Rafiq Ahmed, David L. 
Woodward, Pasquale L. Melito, Frank G. Rodgers, Frances Jamieson, Bruce 
Ciebin, Aimin Li, and Andrea Ellis. "Characterization of Waterborne 
Outbreak-Associated Campylobacter jejuni, Walkerton, Ontario." Emerging 
Infectious Diseases. Vol. 9, no. 10 (2003):1,232-1,241; 
Sponsor[A]: Health Canada, Ontario Ministry of Health; 
Medium: Water; 
Pollutant(s): Bacteria; 
Impact: Cattle manure from a nearby farm entered the groundwater system 
and caused gastrointestinal illness and death in residents. 

Study: Diesel, Elizabeth A., Melissa L. Wilson, Ryan Mathur, Evan 
Teeters, David Lehmann, and Caitlan Ziatos. "Nutrient Loading Patterns 
on an Agriculturally Impacted Stream System in Huntingdon County 
Pennsylvania Over Three Summers." Northeastern Geology & Environmental 
Sciences. Vol. 29, no. 1 (2007):25-33; 
Sponsor[A]: Juniata College; 
Medium: Water; 
Pollutant(s): Nutrients; 
Impact: Excess nutrients from CAFO manure contributed significantly to 
impaired water quality and resulted in the inability to sustain fish 
populations. 

Study: Hill, Dagne D., William E. Owens, and Paul B. Tchounwou. "Impact 
of Animal Application on Runoff Water Quality in Field Experimental 
Plots." International Journal of Environmental Research and Public 
Health. Vol. 2, no. 2 (2005):314-321; 
Sponsor[A]: Jackson State University, NIH-Center for Environmental 
Health, Louisiana State University; 
Medium: Water; 
Pollutant(s): Nutrients, bacteria; 
Impact: Nutrients from manure spread on fields contributed to water 
pollution. 

Study: Jensen, Kathleen M., Elizabeth A. Makynen, Michael D. Kahl, and 
Gerald T. Ankley. "Effects of the Feedlot Contaminant 17-Trenbolone on 
Reproductive Endocrinology of the Fathead Minnow." Environmental 
Science & Technology. Vol. 40, no. 9 (2006): 3,112-3,117; 
Sponsor[A]: EPA; 
Medium: Water; 
Pollutant(s): Hormones; 
Impact: Fertility of fish was significantly reduced by hormones and 
female fish developed male sex characteristics. 

Study: Orlando, Edward F., Alan S. Kolok, Gerry A. Binzcik, Jennifer L. 
Gates, Megan K. Horton, Christy S. Lambright, L. Earl Gray, Jr., Ana M. 
Soto, and Louis J. Guillette, Jr. "Endocrine-Disrupting Effects of 
Cattle Feedlot Effluent on an Aquatic Sentinel Species, the Fathead 
Minnow." Environmental Health Perspectives. Vol. 112, no. 3 (2004):353- 
358; 
Sponsor[A]: University of Florida; St. Mary's College of Maryland, 
University of Nebraska, EPA, Tufts University; 
Medium: Water; 
Pollutant(s): Hormones; 
Impact: Male fish were demasculinized and there was defeminization of 
female fish. 

Study: Weldon, Mark B. and Keri C. Hornbuckle. "Concentrated Animal 
Feeding Operations, Row Crops, and Their Relationship to Nitrate in 
Eastern Iowa Rivers." Environmental Science & Technology. Vol. 40, no. 
10 (2006): 3,168-3,173; 
Sponsor[A]: University of Iowa; 
Medium: Water; 
Pollutant(s): Nitrogen; 
Impact: High concentrations of nutrients in waters are a result of CAFO 
manure and degrade water quality. 

Study: Mathisen, T., S. G. Von Essen, T. A. Wyatt, and D. J. Romberger. 
"Hog Barn Dust Extract Augments Lymphocyte Adhesion to Human Airway 
Epithelial Cells." Journal of Applied Physiology. Vol. 96, no. 5 
(2004):1,738-1,744; 
Sponsor[A]: Department of Veterans Affairs Medical Center, University 
of Nebraska Medical Center; 
Medium: Air; 
Pollutant(s): Dust; 
Impact: Dust from hog confinement facilities induces airway 
inflammation. 

Study: Romberger, D. J., V. Bodlak, S. G. Von Essen, T. Mathisen, and 
T. A. Wyatt. "Hog Barn Dust Extract Stimulates IL-8 And IL-6 Release in 
Human Bronchial Epithelial Cells Via PKC Activation." Journal of 
Applied Physiology. Vol. 93, no. 1 (2002):289-296; 
Sponsor[A]: Department of Veterans Affairs Medical Center, University 
of Nebraska Medical Center; 
Medium: Air; 
Pollutant(s): Dust; 
Impact: Dust from hog confinement facilities induces airway 
inflammation. 

Study: Schiffman, Susan S., Clare Studwell, Lawrence R. Landerman, 
Katherine Berman, and John S. Sundy. "Symptomatic Effects of Exposure 
to Diluted Air Sampled from a Swine Confinement Atmosphere on Healthy 
Human Subjects." Environmental Health Perspectives. Vol. 113, no. 5 
(2005):567-576; 
Sponsor[A]: Duke University; 
Medium: Air; 
Pollutant(s): Hydrogen sulfide, ammonia, total suspended particulates, 
endotoxin, odor, dust; 
Impact: Short-term exposure to emissions expected downwind from a swine 
confinement facility can induce headaches, eye irritation, and nausea. 

Study: Sigurdarson, Sigurdur T., Patrick T. O'Shaughnessy, Janet A. 
Watt, and Joel N. Kline. "Experimental Human Exposure Inhaled Grain 
Dust and Ammonia: Towards a Model of Concentrated Animal Feeding 
Operations." American Journal of Industrial Medicine.Vol. 46, issue 5 
(2004):345:348; 
Sponsor[A]: University of Iowa; 
Medium: Air; 
Pollutant(s): Dust, ammonia; 
Impact: Exposure to endotoxin-rich dust from CAFOs causes airflow 
obstruction in subjects with mild asthma. 

Study: Sundblad, B-M., B-M. Larsson, L. Palmberg, and K. Larsson. 
"Exhaled Nitric Oxide and Bronchial Responsiveness in Healthy Subjects 
Exposed to Organic Dust." European Respiratory Journal. Vol. 20, no. 2 
(2002): 426-431; 
Sponsor[A]: National Institute of Environmental Medicine, Sweden; 
Medium: Air; 
Pollutant(s): Dust; 
Impact: Airway inflammation is induced by exposure to a farming 
environment. 

Study: Wickens, K., et. Al. "Farm Residence and Exposures and the Risk 
of Allergic Diseases in New Zealand Children." Allergy. Vol. 57, no. 12 
(2002): 1,171-1,179; 
Sponsor[A]: University of Otago (New Zealand); Utrecht University (The 
Netherlands); 
Medium: Air; 
Pollutant(s): Dust; 
Impact: There was a greater prevalence of allergic disease for children 
on farms. 

Study: Wilson, Vickie S., Christy Lambright, Joe Ostby, and L.E. Gray, 
Jr. "In Vitro and in Vivo Effects of 17-Trenbolone: A Feedlot Effluent 
Contaminant." Toxicological Sciences. Vol. 70, no. 2 (2002): 202-211; 
Sponsor[A]: EPA; 
Medium: Water; 
Pollutant(s): Hormones; 
Impact: Hormones found in feedlot effluent caused reproductive 
malformations in laboratory rats and human cells. 

Study: Wyatt, Todd A., Rebecca E. Slager, Jane DeVasure, Brent W. 
Auvermann, Michael L. Mulhern, Susanna Von Essen, Tracy Mathisen, 
Anthony A. Floreani, and Debra J. Romberger. "Feedlot Dust Stimulation 
of Interleukin-6 And 8 Requires Protein Kinase C-Epsilon Human 
Bronchial Epithelial Cells." American Journal of Physiology-Lung 
Cellular and Molecular Physiology. Vol. 293, no. 5 (2007):1,163-1,170; 
Sponsor[A]: Nebraska Medical Center, Department of Veterans Affairs 
Medical Center, Texas A&M; 
Medium: Air; 
Pollutant(s): Dust; 
Impact: Dust extract from cattle feedlots stimulates airway 
inflammation at concentrations found downwind from the operation. 

Studies indicating no impact: 

Study: Hill, Dagne D., William E. Owens, and Paul B. Tchounwou. 
"Prevalence of Escherichia coli O157:H7 Bacterial Infections Associated 
With the Use of Animal Wastes in Louisiana for the Period 1996-2004." 
International Journal of Environmental Research and Public Health. Vol. 
3, no. 1 (2006): 107-113; 
Sponsor[A]: Grambling State University, Louisiana State University, 
Jackson State University; 
Medium: Water; 
Pollutant(s): Escherichia coli (not measured); 
Impact: Although some of the parishes surveyed had large amounts of 
animal waste generated each year, statistics did not show a 
correlations with Escherichia coli 0157:H7 bacterial infections. 

Study: Hill, Dagne D., William E. Owens, and Paul B. Tchounwou. 
"Prevalence of Selected Bacterial Infections Associated with the Use of 
Animal Waste in Louisiana." International Journal of Environmental 
Research and Public Health. Vol. 2, no. 1 (2005): 84-93; 
Sponsor[A]: Jackson State University, Louisiana State University; 
Medium: Water; 
Pollutant(s): Escherichia coli (not measured); 
Impact: Although the four parishes surveyed had large amounts of animal 
waste generated, statistics does not show a correlation between this 
and bacterial infections. 

Study: Krapac, I.G., W.S. Dey, W.R. Roy, C.A. Smyth, E. Storment, S.L. 
Sargent, and J.D. Steele. "Impacts of Swine Manure Pits on Groundwater 
Quality." Environmental Pollution. Vol. 120, issue 2 (2002): 475-492; 
Sponsor[A]: Illinois State Geological Survey, University of Illinois, 
Illinois Department of Agriculture; 
Medium: Water; 
Pollutant(s): Chloride, ammonium, phosphate, potassium, nitrate, 
bacteria; 
Impact: Groundwater near swine CAFOs has not been significantly 
impacted. 

Study: Mugel, Douglas N. "Ground-Water Quality and Effects of Poultry 
Confined Animal Feeding Operations on Shallow Ground Water, Upper Shoal 
Creek Basin, Southwest Missouri, 2000." U.S. Geological Survey Water- 
Resources Investigations Report 02-4125 (2002); 
Sponsor[A]: United States Geological Survey; 
Medium: Water; 
Pollutant(s): Nutrients, bacteria; 
Impact: The results do not indicate that poultry CAFOs are affecting 
the shallow ground water with respect to nutrients and fecal bacteria. 

Study: Braun-Fahrlander, Charlotte, Josef Riedler, Udo Herz, Waltraud 
Eder, Marco Waster, Leticia Grize, Soyoun Maisch, David Carr, Florian 
Gerlach, Albrecht Bufe. "Environmental Exposure to Endotoxin and its 
Relation to Asthma in School-Age Children." The New England Journal of 
Medicine. Vol. 347, no. 12 (2002): 869-877; 
Sponsor[A]: Institute of Social and Preventive Medicine (Switzerland), 
Children's Hospital (Austria), Philipps University (Germany), Ruhr 
University (Germany), University Children's Hospital (Switzerland), 
University of Munich (Germany); 
Medium: Air; 
Pollutant(s): Dust; 
Impact: Decreased risk of hay fever, asthma, and wheeze in children 
exposed to high levels of endotoxin in dust. 

Study: Elliott, L., K. Yeatts, and D. Loomis. "Ecological Associations 
Between Asthma Prevalence And Potential Exposure to Farming." European 
Respiratory Journal. Vol. 24, no. 6 (2004): 938-941; 
Sponsor[A]: University of North Carolina, Chapel Hill; 
Medium: Air; 
Pollutant(s): N/A; 
Impact: Findings are consistent with the hypothesis that certain farm 
exposures are protective against childhood asthma. 

Study: McGinn, S. M., H. H. Janzen, and T. Coates. "Atmospheric 
Pollutants and Trace Gases: Atmospheric Ammonia, Volatile Fatty Acids, 
and Other Odorants near Beef Feedlots." Journal of Environmental 
Quality. Vol. 32, no. 4 (2003):1,173-1,182; 
Sponsor[A]: Research Centre, Agriculture and Agri-Food Canada; 
Medium: Air; 
Pollutant(s): Ammonia, odor, organic compounds, total suspended 
particulates, dust; 
Impact: Odorants from feedlots were effectively dispersed. Emitted 
ammonia was deposited to the soil downwind. 

Studies showing an indirect link between pollutants and impacts: 

Study: Valcour, James E., Pascal Michel, Scott A. McEwen, and Jeffrey 
B. Wilson. "Associations between Indicators of Livestock Farming 
Intensity and Incidence of Human Shiga Toxin-Producing Escherichia coli 
Infection." Emerging Infectious Diseases. Vol. 8, no. 3 (2002): 252- 
257; 
Sponsor[A]: University of Guelph; Université de Montréal; Centre for 
Infectious Disease Prevention and Control-Health Canada; 
Medium: Water; 
Pollutant(s): Escherichia coli (not measured); 
Impact: The strongest associations with human Escherichia coli 
infection were the ratio of beef cattle to human population and the 
application of manure to the surface of agricultural land by a solid 
spreader and by a liquid spreader. 

Study: Wing, Steve, Stephanie Freedman, and Lawrence Band. "The 
Potential Impact of Flooding on Confined Animal Feeding Operations in 
Eastern North Carolina." Environmental Health Perspectives. Vol. 110, 
no. 4 (2002): 387-391; 
Sponsor[A]: University of North Carolina; 
Medium: Water; 
Pollutant(s): N/A; 
Impact: Flood events have a significant potential to degrade 
environmental health because of dispersion of wastes from industrial 
animal operations in areas with vulnerable populations. 

Study: Avery, Rachel C., Steve Wing, Stephen W. Marshall, and Susan S. 
Schiffman. "Odor from Industrial Hog Farming Operations and Mucosal 
Immune Function in Neighbors." Archives of Environmental Health. Vol. 
59, no. 2 (2004): 101-108; 
Sponsor[A]: University of North Carolina, Duke University; 
Medium: Air; 
Pollutant(s): N/A; 
Impact: This study suggests that malodor from industrial swine 
operations can affect the secretory immune system, although the reduced 
levels reported are still within normal range. 

Study: Bullers, Susan. "Environmental Stressors, Perceived Control, and 
Health: The Case of Residents Near Large-Scale Hog Farms in Eastern 
North Carolina." Human Ecology. Vol. 33, no. 1 (2005): 1-16; 
Sponsor[A]: University of North Carolina Wilmington; 
Medium: Air/Water; 
Pollutant(s): N/A; 
Impact: Residents living near large-scale hog farms in eastern North 
Carolina report symptoms related to respiratory and sinus problems and 
nausea. 

Study: Chénard, Liliane, Ambikaipakan Senthilselvan, Vaneeta K. Grover, 
Shelley P. Kirychuk, Joshua A. Lawson, Thomas S. Hurst, and James A. 
Dosman. "Lung Function and Farm Size Predict Healthy Worker Effect in 
Swine Farmers." Chest. Vol. 131, no. 1 (2007): 245-254; 
Sponsor[A]: Institute of Agriculture Rural and Environmental Health, 
University of Saskatchewan (Canada), University of Alberta (Canada), 
Canadian Institute of Health Research; 
Medium: Air; 
Pollutant(s): N/A; 
Impact: Some swine workers are less affected by swine air and continue 
in the profession. Other workers are more affected. 

Study: Chrischilles, Elizabeth, Richard Ahrens, Angela Kuehl, Kevin 
Kelly, Peter Thorne, Leon Burmeister, and James Merchant. "Asthma 
Prevalence and Morbidity Among Rural Iowa Schoolchildren." Journal of 
Allergy and Clinical Immunology. Vol. 113, no. 1 (2004): 66-71; 
Sponsor[A]: University of Iowa, EPA; 
Medium: Air; 
Pollutant(s): N/A; 
Impact: Among children who wheeze, farm and nonfarm children were 
equally likely to have been given a diagnosis of asthma and had 
comparable morbidity. 

Study: Dosman, J.A., J.A. Lawson, S.P. Kirychuk, Y. Cormier, J. Biem, 
and N. Koehncke. "Occupational Asthma in Newly Employed Workers in 
Intensive Swine Confinement Facilities." European Respiratory Journal. 
Vol. 24, no. 6 (2004): 698-702; 
Sponsor[A]: Institute of Agricultural Rural and Environmental Health, 
University of Saskatchewan (Canada), Laval University (Canada); 
Medium: Air; 
Pollutant(s): N/A; 
Impact: Newly employed workers in intensive swine confinement 
facilities reported development of acute onset of wheezing and cough 
suggestive of asthma. 

Study: Merchant, James A., Allison L. Naleway, Erik R. Svendsen, Kevin 
M. Kelly, Leon F. Burmeister, Ann M. Stromquist, Craig D. Taylor, Peter 
S. Thorne, Stephen J. Reynolds, Wayne T. Sanderson, and Elizabeth A. 
Chrischilles. "Asthma and Farm Exposures in a Cohort of Rural lowa 
Children." Environmental Health Perspectives. Vol. 113, No. 3 (2005): 
350-356; 
Sponsor[A]: University of Iowa, EPA, Colorado State University, Kaiser 
Permanente; 
Medium: Air; 
Pollutant(s): N/A; 
Impact: There was a high prevalence of asthma health outcome among farm 
children living on farms that raise swine and raise swine and add 
antibiotics. 

Study: Mirabelli, Maria C., Steve Wing, Stephen W. Marshall, and 
Timothy C. Wilcosky. "Asthma Symptoms Among Adolescents Who Attend 
Public Schools That Are Located Near Confined Swine Feeding 
Operations." Pediatrics. Vol. 118, no. 1 (2006): 66-75; 
Sponsor[A]: University of North Carolina, RTI International; 
Medium: Air; 
Pollutant(s): N/A; 
Impact: Estimated exposure to airborne pollution from confined swine 
feeding operations is associated with adolescents' wheezing symptoms. 

Study: Palmberg, Lena, Britt-Marie Larsson, Per Malmberg, and Kjell 
Larsson. "Airway Responses of Healthy Farmers and Nonfarmers to 
Exposure in a Swine Confinement Building." Scandinavian Journal of 
Work, Environment, and Health. Vol. 28, no. 4 (2002): 256-263; 
Sponsor[A]: National Institute of Environmental Medicine (Sweden), 
National Institute for Working Life (Sweden); 
Medium: Air; 
Pollutant(s): N/A; 
Impact: Altered lung function and bronchial responsiveness was found in 
nonfarming subjects. Only minor alterations were found in the farmers. 

Study: Radon, Katja, Anja Schulze, Vera Ehrenstein, Rob T. van Strien, 
Georg Praml, and Dennis Nowak. "Environmental Exposure to Confined 
Animal Feeding Operations and Respiratory Health of Neighboring 
Residents." Epidemiology. Vol. 18, no. 3 (2007): 300-308; 
Sponsor[A]: Institute for Occupational and Environmental Medicine 
(Germany), National Research Centre for Environment and Health 
(Germany), Boston University, Municipal Health Service Amersterdam; 
Medium: Air; 
Pollutant(s): N/A; 
Impact: Respiratory disease was found among resident living near 
confined animal feeding operations. 

Study: Sigurdarson, Sigurdur T. and Joel N. Kline. "School Proximity to 
Concentrated Animal Feeding Operations and Prevalence of Asthma in 
Students." Chest. Vol. 129, no. 6 (2006):1,486-1,491; 
Sponsor[A]: University of Iowa Carver College of Medicine, University 
of Iceland; 
Medium: Air; 
Pollutant(s): N/A; 
Impact: Children in the study school, located one-half mile from a 
CAFO, had a significantly increased prevalence of physician-diagnosed 
asthma. 

Studies measuring pollutants: 

Study: Anderson, M.E. and M.D. Sobsey. "Detection And Occurrence of 
Antimicrobially Resistant E. Coli In Groundwater on or Near Swine Farms 
In Eastern North Carolina." Water Science & Technology. Vol. 54, no. 3 
(2006): 211-218; 
Sponsor[A]: University of North Carolina; 
Medium: Water; 
Pollutant(s): Antibiotics; 
Impact: Antibiotic-resistant E. coli strains are present in 
groundwaters of swine farms. 

Study: Batt, Angela L., Daniel D. Snow, and Diana S. Aga. "Occurrence 
of Sulfonamide Antimicrobials in Private Water Wells in Washington 
Country, Idaho, USA." Chemosphere. Vol. 64, issue 11 (2006): 1,963- 
1,971; 
Sponsor[A]: State University of New York at Buffalo, University of 
Nebraska; 
Medium: Water; 
Pollutant(s): Antimicrobials, nitrate, ammonium; 
Impact: All six sampled wells were contaminated by veterinary 
antimicrobials and had elevated concentrations of nitrate and ammonium. 
Three wells had nitrate levels exceeding EPA thresholds. 

Study: Campagnolo, Enzo R., Kammy R. Johnson, Adam Karpati, Carol S. 
Rubin, Dana W. Kolpin, Michael T. Meyer, J. Emilio Esteban, Russell W. 
Currier, Kathleen Smith, Kendall M. Thu, and Michael McGeehin. 
"Antimicrobial Residues in Animal Waste and Water Resources Proximal to 
Large-Scale Swine and Poultry Feeding Operations." The Science of the 
Total Environment. Vol. 299, no. 1 (2002): 89-95; 
Sponsor[A]: CDC, U.S. Geological Survey, Iowa Department of Public 
Health, Ohio Department of Health, University of Iowa; 
Medium: Water; 
Pollutant(s): Antimicrobials; 
Impact: Multiple classes of antimicrobial compounds were detected in 
surface and groundwater samples collected proximal to the swine and 
poultry farms. 

Study: Durhan, Elizabeth J., Christy S. Lambright, Elizabeth A. 
Makynen, James Lazorchak, Phillip C. Hartig, Vickie S. Wilson, L. Earl 
Gray, and Gerald T. Ankley. "Identification of Metabolites of 
Trenbolone Acetate in Androgenic Runoff from a Beef Feedlot." 
Environmental Health Perspectives. Vol. 114, supp. 1 (2006):65-68; 
Sponsor[A]: EPA; 
Medium: Water; 
Pollutant(s): Hormones; 
Impact: Whole-water samples from the discharge contained detectable 
concentrations of hormones. 

Study: Gessel, Peter D., Neil C. Hansen, Sagar M. Goyal, Lee J. 
Johnston, and Judy Webb. "Persistence Of Zoonotic Pathogens in Surface 
Soil Treated With Different Rates of Liquid Pig Manure." Applied Soil 
Ecology. Vol. 25, issue 23 (2004): 237-243; 
Sponsor[A]: University of Minnesota; 
Medium: Water; 
Pollutant(s): Pathogens; 
Impact: Manure application rate was correlated positively with the 
persistence of fecal indicators but did not relate to survival of 
indicators with short survival times. 

Study: Haggard, Brian E. , Paul B. DeLaune, Douglas R. Smith, and 
Philip A. Moore, Jr. "Nutrient and B17-Estradiol Loss in Runoff Water 
From Poultry Litters." Journal of the American Water Resources 
Association. Vol. 41, no. 2 (2005):245-256; 
Sponsor[A]: USDA, University of Arkansas; 
Medium: Water; 
Pollutant(s): Nutrients, hormones; 
Impact: In general, poultry litter applications increased nutrient and 
hormone concentrations in runoff water. 

Study: Hutchins, Stephen R., Mark V. White, Felisa M. Hudson, and 
Dennis D. Fine. "Analysis of Lagoon Samples from Different Concentrated 
Animal Feeding Operations for Estrogens and Estrogen Conjugates." 
Environmental Science & Technology. Vol. 41, no. 3 (2007): 738-744; 
Sponsor[A]: EPA, Shaw Environmental and Infrastructure; 
Medium: Water; 
Pollutant(s): Hormones; 
Impact: Estrogen conjugates contribute significantly to the overall 
estrogen load, even in different types of CAFO lagoons. 

Study: Koike, S., I.G. Krapac, H.D. Oliver, A.C. Yannarell, J.C. Chee- 
Sanford, R.I. Aminov, and R.I. Makie. "Monitoring and Source Tracking 
of Tetracycline Resistance Genes in Lagoons and Groundwater Adjacent to 
Swine Production Facilities over a 3-Year Period." Applied and 
Environmental Microbiology. Vol. 73, no. 15 (2007): 4,813-4,823; 
Sponsor[A]: University of Illinois, USDA, Illinois State Geological 
Survey, Rowett Research Institute (UK); 
Medium: Water; 
Pollutant(s): Antibiotics; 
Impact: Antibiotic resistance genes in groundwater are affected by 
swine manure and also part of the indigenous gene pool. 

Study: Miller, David H. and Gerald T. Ankley. "Modeling Impacts On 
Populations: Fathead Minnow (Pimephales Promelas) Exposure to the 
Endocrine Disruptor 17ß-Trenbolone as a Case Study." Ecotoxicology and 
Environmental Safety. Vol. 59, issue 1 (2004): 1-9; 
Sponsor[A]: EPA; 
Medium: Water; 
Pollutant(s): Hormones; 
Impact: Model shows that if fathead minnow is exposed to continuous 
concentrations of hormone, there will be a risk of extinction. 

Study: Nelson, Nathan O., John E. Parsons, and Robert L. Mikkelsen. 
"Field-Scale Evaluation of Phosphorus Leaching in Acid Sandy Soils 
Receiving Swine Waste." Journal of Environmental Quality. Vol. 34, no. 
6 (2005): 2,024-2,035; 
Sponsor[A]: USDA, North Carolina State University; 
Medium: Water; 
Pollutant(s): Phosphorus; 
Impact: The results show that substantial quantities of phosphorus can 
be leached through soils with low phosphorus sorption capacities. 

Study: Peak, Nicholas, Knapp, Charles W, Richard K. Yang, Margery M. 
Hanfelt, Marilyn S. Smith, Diana S. Aga, and David W. Graham. 
"Abundance of Six Tetracycline Resistance Genes in Wastewater Lagoons 
at Cattle Feedlots With Different Antibiotic Use Strategies." 
Environmental Microbiology. Vol. 9, no. 1 (2007): 143-151; 
Sponsor[A]: University of Kansas, Kansas State University, State 
University of New York at Buffalo; 
Medium: Water; 
Pollutant(s): Antibiotic resistant genes; 
Impact: CAFOs using larger amounts of antibiotics had significantly 
higher detected resistance gene levels. 

Study: Sapkota, Amy R., Frank C. Curriero, Kristen E. Gibson, and 
Kellogg J. Schwab. "Antibiotic-Resistant Enterococci and Fecal 
Indicators in Surface Water and Groundwater Impacted by a Concentrated 
Swine Feeding Operation." Environmental Health Perspectives. Vol. 115, 
no. 7 (2007):1,040-1,045; 
Sponsor[A]: Johns Hopkins Bloomberg School of Public Health; University 
of Maryland; 
Medium: Water; 
Pollutant(s): Antibiotic resistant bacteria, fecal indicators; 
Impact: Detected elevated levels of fecal indicators and antibiotic-
resistant bacteria in water sources down gradient from a swine 
facility. 

Study: Soto, Ana M., Janine M. Calabro, Nancy V. Prechtl, Alice Y. Yau, 
Edward F. Orlando, Andreas Daxenberger, Alan S. Kolok, Louis J. 
Guillette, Jr., Bruno le Bizec, Iris G. Lange, and Carlos Sonnenschein. 
"Androgenic and Estrogenic Activity in Water Bodies Receiving Cattle 
Feedlot Effluent in Eastern Nebraska, USA." Environmental Health 
Perspectives. Vol. 112, no. 3 (2004):346-352; 
Sponsor[A]: Tufts University; Southwest Research Institute; St. Mary's 
College of Maryland; Universität München-Weihenstephan, Germany; 
University of Nebraska; University of Florida; Ecole Nationale 
Vétérinaire de Nantes, France; 
Medium: Water; 
Pollutant(s): Hormones; 
Impact: Feedlot effluents contain sufficient levels of hormonally 
active agents to warrant further investigation of possible effects on 
aquatic ecosystem health. 

Study: Thorsten, Christiana, Rudolf J. Schneider, Harald A. Farber, 
Dirk Skutlarek, Michael T. Meyer, and Heiner E. Goldbach. 
"Determination of Antibiotic Residues in Manure, Soil, and Surface 
Waters." Acta hydrochimica et hydrobiologica. Vol. 31, no. 1 (2003):36- 
44; 
Sponsor[A]: University of Bonn, Germany; U.S. Geological Survey; 
Medium: Water; 
Pollutant(s): Antibiotics; 
Impact: In each of the surface waters tested antibiotics could be 
detected. 

Study: Thurston-Enriquez, Jeanette A., John E. Gilley, and Bahman 
Eghball. "Microbial Quality of Runoff Following Land Application of 
Cattle Manure And Swine Slurry." Journal of Water and Health. vol. 3, 
no. 2 (2005): 157-171; 
Sponsor[A]: University of Nebraska; 
Medium: Water; 
Pollutant(s): Microbials; 
Impact: Large microbial loads could be released via heavy precipitation 
events and could have a significant impact on water bodies. 

Study: Toetz, Dale. "Nitrate in Ground and Surface Waters in the 
Vicinity of a Concentrated Animal Feeding Operation." Archives of 
Hydrobiology. Vol. 166, no. 1 (2006): 67-77; 
Sponsor[A]: Oklahoma State University; 
Medium: Water; 
Pollutant(s): Nitrogen; 
Impact: Drinking water was contaminated with CAFOs as the suspected 
source. 

Study: U.S. Department of Interior. U.S. Geological Survey. In 
cooperation with U.S. Environmental Protection Agency, National 
Exposure Research Laboratory. Geochemistry and Characteristics of 
Nitrogen Transport at a Confined Animal Feeding Operations in a Coastal 
Plain Agricultural Watershed, and Implications for Nutrient Loading in 
the Neuse River Basin, North Carolina, 1999-2002. Scientific 
Investigations Report 2004-5283, Reston, Va.: (2004); 
Sponsor[A]: U.S. Geological Survey, EPA; 
Medium: Water; 
Pollutant(s): Nitrogen; 
Impact: Large amounts of nitrogen moving in the estuary as a result of 
extreme events may potentially cause algal growths. 

Study: United State Geological Survey in cooperation with Virginia 
Department of Health. Water-Quality Data from Ground-and Surface-Water 
Sites near Concentrated Animal Feeding Operations (CAFOs) and non-CAFOs 
in the Shenandoah Valley and Eastern Shore of Virginia, January- 
February, 2004. Reston, Va (2005); 
Sponsor[A]: United State Geological Survey in cooperation with Virginia 
Department of Health; 
Medium: Water; 
Pollutant(s): Bacteria, antibiotics, trace metals, biological oxygen 
demand, nitrogen; 
Impact: N/A. 

Study: United States Geological Survey. Fractionation and 
Characterization of Organic Matter in Wastewater from a Swine Waste- 
Retention Basin. Scientific Investigations Report 2004-5217 (2004); 
Sponsor[A]: United States Geological Survey; 
Medium: Water; 
Pollutant(s): Organic matter; 
Impact: The bulk of the organic matter consists of microbial cellular 
constituents and their degradation products. 

Study: Chapin, Amy, Ana Rule, Kristen Gibson, Timothy Buckley, and 
Kellogg Schwab. "Airborne Multidrug-Resistant Bacteria Isolated from a 
Concentrated Swine Feeding Operation." Environmental Health 
Perspectives. Vol. 113, no. 2 (2005):137-142; 
Sponsor[A]: Johns Hopkins University; 
Medium: Air; 
Pollutant(s): Antibiotic resistant bacterial pathogens; 
Impact: Multidrug-resistant bacterial pathogens were detected in the 
air of a swine CAFO. 

Study: Donham, Kelley. J., Joung Ae Lee, Kendall Thu, and Stephen J. 
Reynolds. "Assessment of Air Quality at Neighbor Residences in the 
Vicinity Of Swine Production Facilities." Journal of Agromedicine. Vol. 
11, no. 3-4 (2006): 15-24; 
Sponsor[A]: University of Iowa, Northern Illinois University, and 
Colorado State University; 
Medium: Air; 
Pollutant(s): Hydrogen sulfide, ammonia, carbon dioxide, particulate 
matter; 
Impact: Average concentration of hydrogen sulfide exceeded EPA 
recommended community standards in all three areas assessed. 

Study: Gibbs, Shawn G., Christopher F. Green, Patrick M. Tarwater, 
Linda C. Mota, Kristina D. Mena, and Pasquale V. Scarpino. "Isolation 
of Antibiotic-Resistant Bacteria from the Air Plume Downwind of a Swine 
Confined or Concentrated Animal Feeding Operation." Environmental 
Health Perspectives. Vol. 114, no. 7 (2006):1,032-1,037; 
Sponsor[A]: University of Texas, University of Cincinnati; 
Medium: Air; 
Pollutant(s): Antibiotic-resistant bacteria; 
Impact: Bacterial concentrations with multiple antibiotic resistances 
or multidrug resistance were recovered inside and outside to 150 m 
downwind of a facility, even after antibiotic use was discontinued. 

Study: Harper, Lowry A., Ron R. Sharpe, Tim B. Parkin, Alex De 
Visscher, Oswald van Cleemput, and F. Michael Byers. "Nitrogen Cycling 
through Swine Production Systems: Ammonia, Dinitrogen, and Nitrous 
Oxide Emissions." Journal of Environmental Quality. Vol. 33, no. 4 
(2004): 1,189-1,201; 
Sponsor[A]: USDA, Ghent University (Belgium); 
Medium: Air; 
Pollutant(s): Nitrogen; 
Impact: In contrast with previous and current estimates of ammonia 
emissions from CAFOs, this study found smaller ammonia emissions from 
animal housing, lagoons, and fields. 

Study: Hamscher, Gerd, Heike Theresia Pawelzick, Silke Sczesny, Heinz 
Nau, and Jörg Hartung. "Antibiotics in Dust Originating from a Pig- 
Fattening Farm: A New Source of Health Hazard for Farmers?" 
Environmental Health Perspectives. Vol. 111, no. 13 (2003):1,590- 
1,594; 
Sponsor[A]: School of Veterinary Medicine Hannover, Germany; 
Medium: Air; 
Pollutant(s): Antibiotics; 
Impact: Five different antibiotics were detected in dust samples swine 
feeding operation. 

Study: Hoff, Steven J., Dwaine S. Bundy, Minda A. Nelson, Brian C. 
Zelle, Larry D. Jacobson, Albert J. Heber, Jinqin Ni, Yuanhui Zhang, 
Jacek A. Koziel, and David B. Beasley. "Emissions of Ammonia, Hydrogen 
Sulfide, and Odor before, during, and after Slurry Removal from a Deep- 
Pit Swine Finisher." Journal of the Air & Waste Management Association. 
Vol. 56, no. 5 (2006): 581-590; 
Sponsor[A]: Iowa State University, University of Minnesota, Purdue 
University, University of Illinois, North Carolina State University; 
Medium: Air; 
Pollutant(s): Ammonia, hydrogen sulfide, odor; 
Impact: Emissions of ammonia, hydrogen sulfide, and odor had large 
increases during slurry removal. A slurry removal even will result in 
acute exposure for animals and workers. 

Study: O'Connor, Rod, Mark O'Connor, Kurt Irgolic, Justin Sabrsula, 
Hakan Gurleyuk, Robert Brunette, Crystal Howard, Jennifer Garcia, John 
Brien, June Brien, and Jessica Brien. "Transformations, Air Transport, 
and Human Impact of Arsenic from Poultry Litter." Environmental 
Forensics. Vol. 6, no. 1 (2005): 83-89; 
Sponsor[A]: Chenard Consulting Services, Karl-Franzeas University 
(Austria), University of North Carolina, Frontier Geosciences, Aqua-
Tech Laboratories; 
Medium: Air; 
Pollutant(s): Arsenic; 
Impact: Levels of arsenic found in homes. This could represent a 
significant health risk. 

Study: Radon, Katja, Brigitta Danuser, Martin Iversen, Eduard Monso, 
Christoph Weber, Jorg Hartung, Kelley J. Donham, Urban Palmgren, and 
Dennis Nowak. "Air Contaminants in Different European Farming 
Environments." Annals of Agriculture and Environmental Medicine. Vol. 
9, no. 1 (2002): 41-48; 
Sponsor[A]: Ludwig-Maximilians-University (Germany), Swiss Federal 
Institute of Technology, Aarhus University Hospital (Denmark), Hospital 
Germans Trial I Pujol (Spain), School of Veterinary Medicine (Germany), 
University of Iowa, Pegasus Labor GmbH (Germany); 
Medium: Air; 
Pollutant(s): Dust, endotoxin, fungi; Impact: The exposure level found 
in this study might put the farmers at risk from respiratory diseases. 

Study: Razote, E.B., R.G. Maghirang, B.Z. Predicala, J.P. Murphy, B.W. 
Auvermann, J.P. Harner III, and W.L. Hargrove. "Laboratory Evaluation 
of the Dust-Emission Potential of Cattle Feedlot Surfaces." 
Transactions of the ASABE. Vol. 49, no. 4 (2006): 1,117-1,124; 
Sponsor[A]: Kansas State University, Prairie Swine Center, Inc. 
(Canada), Texas A&M University; 
Medium: Air; 
Pollutant(s): Particulate Matter; 
Impact: N/A. 

Study: Robarge, Wayne P., John T. Walker, Ronald B. McCulloch, and 
George Murray. "Atmospheric Concentrations of Ammonia and Ammonium at 
an Agricultural Site in the Southeast United States." Atmospheric 
Environment. Vol. 36, no. 10 (2002): 1,661-1,674; 
Sponsor[A]: North Carolina State University, EPA, URS Corporation, 
North Carolina Department of Environmental and Natural Resources; 
Medium: Air; 
Pollutant(s): Ammonia; 
Impact: Elevated ambient ammonia concentrations near an agricultural 
site. 

Study: United State Environmental Protection Agency. National Emission 
Inventory - Ammonia Emissions from Animal Husbandry Operations, Draft 
Report. Washington, D.C. (2004); 
Sponsor[A]: EPA; 
Medium: Air; 
Pollutant(s): Ammonia; 
Impact: N/A. 

Study: Walker, J.T., W.P. Robarge, Y. Wu, and T.P. Meyers. "Measurement 
of Bi-Directional Ammonia Fluxes Over Soybean Using Themodified Bowen- 
Ratio Technique." Agricultural and Forest Meteorology. Vol. 138, no. 1- 
4 (2006): 54-68; 
Sponsor[A]: EPA, North Carolina State University, NASA, NOAA; 
Medium: Air; 
Pollutant(s): Ammonia; 
Impact: In general, the net deposition flux was lower than expected. 

Study: Walker, John T., Wayne P. Robarge, Arun Shendrikar, and Hoke 
Kimball. "Inorganic Pm2.5 at a U.S. Agricultural Site." Environmental 
Pollution. Vol. 139, no. 2 (2006): 258-271; 
Sponsor[A]: EPA, North Carolina State University, North Carolina 
Department of Environment and Natural Resources; 
Medium: Air; 
Pollutant(s): Particulate matter; 
Impact: Model results show that reductions in atmospheric ammonia will 
have minimal effect on organic PM2.5 during summer and a moderate 
effect during winter. 

Study: Walker, J.T., Dave R. Whitall, Wayne P. Robarge, and Hans W. 
Pearl. "Ambient Ammonia and Ammonium Aerosol Across a Region of 
Variable Ammonia Emission Density." Atmospheric Environment. Vol. 38, 
no. 9 (2004): 1,235-1,246; 
Sponsor[A]: EPA, NOAA, North Carolina State University, University of 
North Carolina; 
Medium: Air; 
Pollutant(s): Ammonia, ammonium; 
Impact: Agricultural ammonia emissions influence local ambient 
concentrations of ammonia and PM2.5. 

Study: Wilson, Sacoby M. and Marc L. Serre. "Examination of Atmospheric 
Ammonia Levels Near Hog Cafos, Homes, and Schools In Eastern North 
Carolina." Atmospheric Environment. Vol. 41, issue 23 (2007): 4,977- 
4,987; 
Sponsor[A]: University of Michigan, Ann Arbor; University of North 
Carolina at Chapel Hill; 
Medium: Air; 
Pollutant(s): Ammonia; 
Impact: Distance to one or more CAFOs is the key variable in 
controlling atmospheric ammonia at the community level in Eastern N.C. 

Study: Muller-Suur, C., P.H. Larsson, K. Larsson, J. Grunewald. 
"Lymphocyte Activation After Exposure to Swine Dust: A Role Of Humoral 
Mediators and Phagocytic Cells." European Respiratory Journal. Vol. 19, 
issue 1 (2002): 104-107; 
Sponsor[A]: [Empty]; 
Medium: Air; 
Pollutant(s): Dust; 
Impact: About immune system response. 

Study: Charavaryamath, Chandrashekhar, Kyathanahalli S Janardhan, Hugh 
G Townsend, Philip Willson, and Baljit Singh. "Multiple Exposures to 
Swine Barn Air Induce Lung Inflammation and Airway Hyper- 
Responsiveness." Respiratory Research. Vol. 6, no. 1 (2005):50-66; 
Sponsor[A]: University of Saskatchewan, Canada; 
Medium: Air; 
Pollutant(s): Endotoxin; 
Impact: Does not address human impacts. 

Study: Eduard, Wijnand, Ernst Omenaas, Per Sigvald Bakke, Jeroen 
Douwes, and Dick Heederik. "Atopic and Non-atopic Asthma in a Farming 
and a General Population." American Journal of Industrial Medicine. 
Vol. 46, issue 4 (2004): 396-399; 
Sponsor[A]: National Institute of Occupational Health (Norway), 
University of Bergen (Norway), University of Wellington (New Zealand); 
Medium: Air; 
Pollutant(s): N/A; 
Impact: Protective effect of the farm environment on asthma. 

Source: GAO's analysis of identified studies. 

[A] Sponsor refers to the organization under whose auspices the 
research was conducted or with whom the primary researchers were 
affiliated. 

[End of table] 

[End of section] 

Appendix V: Comments from the Environmental Protection Agency: 

United States Environmental Protection Agency: 
Office Of Air And Radiation: 
Washington, D.C. 20460: 

August 26, 2008: 

Ms. Anu K. Mittal: 
Director, Natural Resources and Environment: 
U.S. Government Accountability Office: 
Washington, DC 20548: 

Dear Ms. Mittal: 

The U. S. Environmental Protection Agency (EPA) appreciates the 
opportunity to review and comment on the Government Accountability 
Office (GAO) draft report, "Concentrated Animal Feeding Operations: EPA 
Needs More Information and a Clearly Defined Strategy to Protect Air 
and Water Quality from Pollutants of Concern" (GAO-08-944). EPA agrees 
in part with the conclusions and recommendations in the draft report. 
Below I provide brief comments and clarification on EPA's position 
regarding the recommendations. Additional technical comments and 
clarifications on draft report language have been provided in a 
separate document. 

Recommendation 1: 

In order to more effectively monitor and regulate CAFOs, we recommend 
that the Administrator of the Environmental Protection Agency develop a 
comprehensive inventory of CAFOs nationwide and incorporate appropriate 
internal controls to ensure the quality of data. The inventory should 
be periodically updated to capture changes that have occurred in the 
animal production industry. 

In Waterkeeper Alliance v. EPA, the Second Circuit Court of Appeals 
found that EPA did not have the authority to permit non-discharging 
CAFOs. EPA is finalizing amendments to a rule that would require CAFOs 
that discharge to apply for National Pollutant Discharge Elimination 
System (NPDES) permits. EPA's Office of Water has supported the Office 
of Enforcement and Compliance Assurance in working with EPA Regions and 
States to develop and implement a new national NPDES data system to 
collect and record facility-specific information on permitted CAFOs. 
This data system. called the Integrated Compliance Information System 
(ICIS), allows EPA to have an updated national inventory of permitted 
CAFOs, including number, location, size, and permitted discharges. 
National data on inspections, enforcement actions, and other NPDES 
program information will also be collected. Although this information 
is currently collected by permitting authorities for permitted CAFOs, 
there is no comprehensive national database where the information is 
stored. While ICIS-NPDES has the functionality to provide complete data 
on the CAFO universe, the data must be consistently collected and 
entered into ICIS by states and EPA regions. Thus, EPA is developing a 
proposed rule package that would set national requirements for certain 
NPDES program information being provided by authorized states to ICIS. 
This information may include an obligation for authorized states to 
provide national data for identifying facilities that have been issued, 
or applied for, a CAFO permit, as well as facilities that should have 
applied for a CAFO permit based on an inspection or enforcement action. 

Recommendation 2: 

In order to more effectively determine the extent of air emissions from 
animal feeding operations, the Administrator of the Environmental 
Protection Agency should: 

a. reassess the current data collection efforts, including its internal 
controls, to ensure that the national air emissions monitoring study 
will provide the scientific and statistically valid data that EPA needs 
for developing its air emissions protocols; 

The Agency is continuously evaluating the national air monitoring 
study, both the data that are being collected, and the methods that are 
being used to collect these data. This continuous evaluation has 
resulted in changes being made and documented in numerous Standard 
Operating Procedures and also to the overall project Quality Assurance 
Plan. These changes range from items such as utilization of a different 
instrument than originally intended to changes in the maintenance 
schedule for certain items. All these changes have been identified 
through discussions with the monitoring contractor or through the 
audits described below. 

The Agency is currently involved in a program of conducting Technical 
Systems Audits at each monitoring site to ensure the contractor is 
performing according to the approved Quality Assurance Project Plan. 
Each site is scheduled to be audited once during the first year of data 
collection and once during the second year. To date, the Agency has 
performed the first year audits at all but six of the monitoring sites 
with those remaining sites scheduled to be audited by the end of 
September 2008. 

b. Provide information to stakeholders on the additional data that it 
plans to use to supplement the national air emissions monitoring study 

With respect to the supplemental data that will be used to augment the 
data collected from the national air emissions monitoring study, we are 
not able to identify at this time which individual studies we plan to' 
use. However, the Agency has been meeting with the Department of 
Agriculture (USDA), with representatives from their offices of 
Agricultural Research Service, Cooperative State Research, Education, 
and Extension Service, and the Economic Research Service to help 
identify what research is ongoing and where gaps may still exist. We 
recently held a two-day workshop in Research Triangle Park, North 
Carolina with USDA representatives that focused on agricultural air 
emissions. Day 2 of the workshop focused exclusively on CAFO research. 

c. Establish a strategy and timetable for developing a process-based 
model that will provide more sophisticated air emissions estimating 
methodologies for animal feeding operations. 

The Agency has begun to evaluate what is necessary to develop a process-
based model for estimating emissions from animal feeding operations. 
The emissions estimating methodology that will be developed from the 
monitoring study will assess the wide range of process information that 
is being collected in the monitoring study. To the extent that these 
parameters have a significant and logical impact on emissions. they 
will be included in the development of a process-based model. This will 
be performed through the use of a statistical software application that 
performs a multiple regression analysis of the process parameters and 
the measured emissions. 

As part of the joint EPA/USDA workshop, there were several 
presentations on process based models that are currently being 
developed throughout the country and discussions on what data needs 
still exist to complete these models. These discussions with both USDA 
and modeling experts will continue as we flesh out our plans for 
developing a process based model. 

Once again, thank you for the opportunity to respond. 

Sincerely, 

Signed by: 
Robert J. Meyers: 
Principal Deputy Assistant Administrator: 

[End of section] 

Appendix VI: GAO Contact and Staff Acknowledgments: 

GAO Contact: 

Anu Mittal (202) 512-3841 or mittala@gao.gov: 

Staff Acknowledgments: 

In addition to the individual named above, Sherry L. McDonald, 
Assistant Director; Kevin Bray; Yecenia C. Camarillo; Wendy Dye; Paul 
Hobart; Cathy Hurley; Holly L. Sasso; James W. Turkett; and Greg 
Wilmoth made key contributions to this report. Also contributing to 
this report were Elizabeth Beardsley, Ben N. Shouse, and Carol 
Herrnstadt Shulman. 

[End of section] 

Footnotes: 

[1] 40 C.F.R. § 122.23(b). 

[2] Federally regulated waterways include waters of the United States 
as defined in 33 C.F.R. § 328.3(a)(1)-(7) and may include rivers, 
wetlands, impoundments, the territorial seas, and waters used in 
interstate commerce. 

[3] Particulate matter is a complex mixture of extremely small 
particles and liquid droplets. Particulate matter can be made up of a 
number of components, including acids (such as nitrates and sulfates), 
organic chemicals, metals, and soil or dust particles. 

[4] Each of these hazardous substances has a reportable quantity of 100 
pounds in a 24-hour period. 

[5] Human sanitary waste includes feces and urine but does not include 
wastes such as water from showers, washing dishes and clothes, and 
flushing toilets. 

[6] National Academy of Sciences, Air Emissions from Animal Feeding 
Operations: Current Knowledge, Future Needs (Washington, D.C.: National 
Academies Press, 2003). 

[7] EPA has retained program authority for Alaska, Idaho, 
Massachusetts, New Hampshire, and New Mexico. Oklahoma has been 
authorized to issue permits for most sources but not for CAFOs. 

[8] GAO, Livestock Agriculture: Increased EPA Oversight Will Improve 
Environmental Program for Concentrated Animal Feeding Operations, 
[hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-03-285] (Washington, 
D.C.: Jan. 16, 2003). 

[9] In July 2007, EPA extended these deadlines to February 27, 2009. 

[10] The Clean Air Act, 42 U.S.C. §§7401-7671q. 

[11] CERCLA, Pub. L. No. 96-510, 94 Stat. 2767 (codified as amended at 
42 U.S.C. §§9601-9675) and EPCRA, Pub. L. No. 99-499, Tit. III, 100 
Stat. 1728 (codified as amended at 42 U.S.C. §§11001-11050). 

[12] The median is the point above and below which half of the cases 
exist. For large farms that raise animals, half of the farms of a 
particular animal type have more animals than the median farm and half 
have fewer animals. 

[13] Human sanitary waste includes urine and feces only; it does not 
include any other household sewage wastes such as water from washing 
dishes or clothes or water used for showers or flushing. 

[14] EPA officials told us that the agency has identified a hog farm of 
this size. 

[15] EPA officials told us that the agency has identified a cattle farm 
of this size. 

[16] R. L. Kellogg, C.H. Lander, D. C. Moffitt, and N. Gollehon. Manure 
Nutrients Relative to the Capacity of Cropland and Pastureland to 
Assimilate Nutrients: Spatial and Temporal Trends for the United 
States. (Washington, D.C.: December 2000). 

[17] GAO, Animal Agriculture: Information on Waste Management and Water 
Quality Issues, [hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO/RCED-
95-200BR] (Washington, D.C.: June 1995). 

[18] Sponsors are agencies, organizations, or universities responsible 
for conducting the study and not necessarily the group funding the 
study. 

[19] EPA, Office of Water, Feedlots Point Source Category Study 
(Washington, D.C.: 1999). 

[20] National Academies of Sciences, Air Emissions from Animal Feeding 
Operations: Current Knowledge, Future Needs (Washington, D.C.: National 
Academies Press, 2003). 

[21] Section 302.4 of title 40 of the Code of Federal Regulations notes 
that the reportable quantity for ammonia is 100 pounds per 24 hours. 

[22] Section 302.4 of title 40 of the Code of Federal Regulations notes 
that the reportable quantity for hydrogen sulfide is 100 pounds per 24 
hours. 

[23] The total penalty is capped at $10,000 for a participant having 10 
or fewer farms to $100,000 for a participant having over 200 farms. 

[24] Since announcing the Air Compliance Agreement, EPA has proposed 
exempting such releases from the CERCLA and EPCRA reporting 
requirements. The exemption, proposed in December 2007, has not been 
finalized. 

[25] Any farm more than 10 times larger than EPA's established size 
thresholds for CAFOs must, within 120 days of receiving an executed 
copy of the agreement, provide the National Response Center with a 
written statement noting the facility's location, estimating air 
emissions of ammonia, and stating that it will notify the Center of 
reportable releases when emission rates are determined by the 
monitoring study. 

[26] EPA placed certain conditions and limits on its agreement not to 
sue animal feeding operations participating in the Air Compliance 
Agreement. For example, EPA can continue to pursue cases that present 
an imminent and substantial endangerment to public health, welfare, or 
the environment. In addition, EPA's agreement not to sue only covers 
emissions from agricultural livestock and livestock waste and does not 
extend to generators or land application of animal waste. 

[27] The Agricultural Air Quality Task Force, created in accordance 
with the 1996 farm bill, is charged with advising the Secretary of 
Agriculture with respect to providing oversight and coordination 
related to agricultural air quality, and consists of leaders in 
farming, industry, health, and science. 

[28] The National Association of SARA Title III Program Officials. The 
Superfund Amendments and Reauthorization Act (SARA) amended CERCLA on 
October 17, 1986, after the first 6 years of the program. 

[29] Sierra Club v. Seaboard Farms Inc., 387 F.3d 1167 (10th Cir. 
2004). 

[30] In commenting on a draft of this report, EPA noted that it had a 
very limited time to respond to the court's request. 

[31] Sierra Club v. Tyson Foods, Inc., 299 F. Supp. 2d 693 (W.D. Ky. 
2003). 

[32] The standard is: 50 ppb average over 1/2 hour not to be exceeded 
more than two times per year; 30 ppb average over 1/2 hour not to be 
exceeded more than two times in any 5 consecutive days. 

[33] In order to adjust the data for survey undercoverage and 
nonresponse, we used the official USDA statistical weights. However, we 
were unable to calculate the confidence intervals around the reported 
estimates because the Census of Agriculture's documentation does not 
provide the necessary information to determine the statistical error 
associated with subpopulation estimates. 

[34] We included a farm, for the purposes of calculating the number of 
farms overall and for each animal type, only when it reported, on the 
Census of Agriculture survey, either sales or inventory numbers for a 
particular animal type. 

[35] By minimum size threshold, we mean the minimum number of animals 
required for classification as a large CAFO without consideration of 
other factors, such as whether the animal feeding operation is a 
significant contributor of pollutants to federally regulated waters, or 
whether pollutants are discharged into federally regulated waters from 
the operation through a manmade ditch. 

[36] A production cycle is the length of time an animal is fed before 
being sold plus time between "crops." For example, the feeding period 
for a broiler is about 48 days. Including time for cleaning barns 
between cycles, restocking, etc., a broiler farm has about 6 production 
cycles per year. We used the number of cycles per animal type provided 
in "Manure Nutrients Relative to the Capacity of Cropland and 
Pastureland to Assimilate Nutrients," USDA, December 2000. 

[37] A pullet is a replacement hen for laying eggs that is less than 1 
year of age. 

[38] "Manure Nutrients Relative to the Capacity of Cropland and 
Pastureland to Assimilate Nutrients," USDA, December 2000. 

[39] The term "hogs" includes all production stages unless otherwise 
stated. 

[40] 1997 was the last Census of Agriculture survey that asked for 
sales of feeder pigs. The 2002 survey asked for hogs "sold or moved 
from this operation, including feeder pigs." In many hog contract 
operations, the farmer does not own the pigs being fed. GAO did not 
determine what effect changing the survey wording had on the change in 
total hogs sold between 1997 and 2002 nor whether the sales of feeder 
pigs as a percentage of total swine sales changed from 1997 to 2002. 

[41] The three states that did not provide information on their state 
CAFO programs were Connecticut, Nevada, and Vermont. 

[42] We do not report the largest farm for each particular animal type 
to avoid disclosing information that would allow the identification of 
the person who supplied the particular information to USDA. Federal law 
prohibits such disclosure. 

[43] "Manure Production and Characteristics" (St. Joseph, Mich.: March 
2005). Manure is "as-excreted" and excludes bedding, waste feed, 
dilution water, biochemical degradation of solids, or dissipation of 
gases. 

[44] Metcalf and Eddy, Inc., "Wastewater Engineering: Treatment, 
Disposal, and Reuse," 3RD Edition, (New York, N.Y.: 1991) and Parker, 
D. and Gallagher, S. K., "Distribution of Human Waste Samples in 
Relation to Sizing Waste Processing in Space," in "Second Conference on 
Lunar Bases and Space Activities of the 21st Century," Volume 2 (NASA 
Conferences Publication 3166: 1992). 

[45] National Academy of Sciences, Air Emissions from Animal Feeding 
Operations: Current Knowledge, Future Needs (Washington, D.C.: National 
Academies Press, 2003). 

[46] The three states that did not provide information on their CAFO 
programs were Connecticut, Nevada, and Vermont. 

[End of section] 

GAO's Mission: 

The Government Accountability 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. 

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each 
weekday, GAO posts newly released reports, testimony, and 
correspondence on its Web site. To have GAO e-mail you a list of newly 
posted products every afternoon, go to [hyperlink, http://www.gao.gov] 
and select "E-mail Updates." 

Order by Mail or Phone: 

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. Government Accountability 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: 

To Report Fraud, Waste, and Abuse in Federal Programs: 

Contact: 

Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: 
E-mail: fraudnet@gao.gov: 
Automated answering system: (800) 424-5454 or (202) 512-7470: 

Congressional Relations: 

Ralph Dawn, Managing Director, dawnr@gao.gov: 
(202) 512-4400: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7125: 
Washington, D.C. 20548: 

Public Affairs: 

Chuck Young, Managing Director, youngc1@gao.gov: 
(202) 512-4800: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7149: 
Washington, D.C. 20548: