United States
Environmental Protection
Agency
Office of Water (4303)
Washington, DC 20460
EPA-821-R-01-002
January 2001
v°xEPA Environmental and Economic
Benefit Analysis of Proposed
Revisions to the National
Pollutant Discharge Elimination
System Regulation and the
Effluent Guidelines for
Concentrated Animal Feeding
Operations
-------
ENVIRONMENTAL AND ECONOMIC BENEFIT ANALYSIS OF THE PROPOSED
REVISIONS TO THE NATIONAL POLLUTANT DISCHARGE ELIMINATION
SYSTEM REGULATION AND THE EFFLUENT GUIDELINES FOR
CONCENTRATED ANIMAL FEEDING OPERATIONS
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue NW
Washington, DC 20460
January, 2001
-------
TABLE OF CONTENTS
INTRODUCTION AND SUMMARY CHAPTER 1
1.1 Definition of CAFOs 1-2
1.2 Current Issues Related to CAFOs 1-3
1.2.1 Potential Environmental Impacts of CAFOs 1-4
1.2.1.1 Water Quality Impairments 1-4
1.2.1.2 Ecological Impacts 1-5
1.2.1.3 Human Health Effects 1-5
1.2.2 Recent Industry Trends 1-5
1.2.2.1 Increased Production and Industry Concentration 1-6
1.2.2.2 Location of Animal Operations Closer to Consumer Markets 1-6
1.2.2.3 Advances in Agriculture Production Practices to
Manage and Dispose Manure 1-7
1.3 Proposed Changes to CAFO Regulations 1-7
1.3.1 Changes to NPDES Regulations 1-7
1.3.2 Changes to ELGs 1-9
1.3.3 Number of Regulated Operations 1-10
1.4 Analytic Methods and Results 1-10
1.5 Organization of Report 1-11
1.6 References 1-13
POTENTIAL IMPACTS OF AFOs ON
ENVIRONMENTAL QUALITY AND HUMAN HEALTH CHAPTER 2
2.1 Pathways for the Release of Pollutants from AFOs 2-2
2.1.1 Overland Discharge 2-4
2.1.1.1 Surface Runoff 2-4
2.1.1.2 Soil Erosion 2-4
2.1.1.3 Acute Events 2-5
-------
TABLE OF CONTENTS
(continued)
2.1.2 Leaching to Groundwater 2-6
2.1.3 Discharges to the Air and Subsequent Deposition 2-6
2.2 Potential Ecological Hazards Posed by AFO Pollutants 2-7
2.2.1 Nutrients and Eutrophication 2-9
2.2.1.1 Nitrogen and Nitrogen Compounds 2-9
2.2.1.2 Phosphorus 2-10
2.2.1.3 Eutrophication 2-12
2.2.2 Pathogens 2-12
2.2.3 Organic Compounds and Biochemical Oxygen Demand (BOD) 2-14
2.2.4 Solids and Siltation 2-15
2.2.5 Salts and Trace Elements 2-16
2.2.6 Odorous/Volatile Compounds 2-17
2.2.7 Other Pollutants and Ecosystem Imbalances 2-17
2.3 Human Health Impacts Related to AFO Pollutants 2-18
2.3.1 Health Impacts Associated with Nitrates 2-19
2.3.2 Health Impacts Associated with Algal Blooms 2-20
2.3.3 Health Impacts Associated with Pathogens 2-21
2.3.4 Health Impacts Associated with Trace Elements and Salts 2-22
2.3.5 Other Health Impacts 2-22
2.4 References 2-24
CONCEPTUAL FRAMEWORK AND OVERVIEW OF METHODS CHAPTER 3
3.1 Possible Environmental Improvements and Resulting Benefits 3-1
3.2 Specific Benefits Analyzed 3-3
3.3 Predicting Change in Environmental Quality and Resulting Beneficial Use 3-4
3.4 Valuing Benefits 3-6
3.4.1 Overview of Economic Valuation 3-6
3.4.2 Primary Approaches for Measuring Benefits 3-7
-------
TABLE OF CONTENTS
(continued)
3.4.3 Valuation of CAFO Regulatory Benefits Based on Previous Studies 3-8
3.4.4 Aggregating Benefits 3-9
3.5 Summary 3-10
3.6 References 3-11
MODELING OF IMPROVEMENTS IN SURFACE WATER QUALITY
AND BENEFITS OF ACHIEVING RECREATIONAL USE LEVELS CHAPTER 4
4.1 Introduction and Overview 4-1
4.2 Model Facility Analysis 4-2
4.3 Edge-of-Field Loadings Analysis 4-6
4.3.1 Loadings from Manure Application 4-6
4.3.2 Loadings from Lagoons and Other Storage Structures 4-7
4.3.3 Loadings from Feedlots 4-8
4.3.4 Model Loadings Under Regulatory Scenarios 4-8
4.4 Analysis of AFO/CAFO Distribution 4-9
4.4.1 Approach 4-9
4.4.2 Estimated Number of AFOs and CAFOs 4-11
4.4.3 Geographic Placement of Facilities 4-13
4.5 Surface Water Modeling 4-13
4.5.1 Defining the Hydrologic Network 4-15
4.5.2 Distributing AFOs and CAFOs to Agricultural Land 4-16
4.5.3 Calculating AFO/CAFO-Related Loadings to Waterbodies 4-16
4.5.4 Loadings from Other Sources 4-16
4.5.5 Fate and Transport Modeling 4-17
4.5.6 Estimated Changes in Loadings 4-17
4.6 Valuation of Water Quality Changes 4-17
4.6.1 Support of Designated Uses 4-22
4.6.2 Application of CV Study 4-23
4.6.3 Estimated Benefits 4-25
in
-------
TABLE OF CONTENTS
(continued)
4.7 References 4-26
Appendix 4-A: NWPCAM Calculation of the Economic Benefits of
Improved Surface Water Quality 4A-1
REDUCED INCIDENCE OF FISH KILLS CHAPTER 5
5.1 Introduction 5-1
5.2 Analytic Approach 5-2
5.2.1 Data Sources and Limitations 5-2
5.2.2 Predicted Change in Fish Kills Under Alternate CAFO Regulations 5-4
5.2.2.1 Baseline Scenario 5-4
5.2.2.2 Regulatory Scenarios 5-7
5.2.3 Valuation of Predicted Reduction in Fish Kills 5-8
5.3 Results 5-10
5.4 Limitations and Caveats 5-10
5.5 References 5-11
Appendix 5-A: Calculation of Annual Benefits Using Minimum and
Maximum Fish Replacement Values 5A-1
IMPROVED COMMERCIAL SHELLFISHING CHAPTER 6
6.1 Introduction 6-1
6.2 Analytic Approach 6-1
6.2.1 Data on Shellfish Harvest Restrictions Attributed to AFOs 6-1
6.2.2 Estimated Impact on Shellfish Harvests 6-4
6.2.2.1 Baseline Annual Shellfish Landings 6-5
6.2.2.2 Estimated Acreage of Harvested Waters 6-5
IV
-------
TABLE OF CONTENTS
(continued)
6.2.2.3 Average Annual Yield of Harvested Waters 6-6
6.2.2.4 Characterization of Waters that are
Unharvested Due to Pollution from AFOs 6-6
6.2.2.5 Estimated Impact of Pollution from AFOs on
Commercial Shellfish Landings 6-7
6.2.3 Estimated Impact of Alternate Regulations on
Commercial Shellfish Harvests 6-8
6.2.4 Valuation of Predicted Change in Shellfish Harvests 6-9
6.2.4.1 Characterization of Consumer Demand for Shellfish 6-11
6.2.4.2 Determining the Change in Consumer Surplus
Associated with Increased Harvests 6-11
6.3 Results 6-13
6.4 Limitations and Caveats 6-14
6.5 References 6-15
REDUCED CONTAMINATION OF PRIVATE WELLS CHAPTER 7
7.1 Introduction 7-1
7.2 Analytic Approach 7-3
7.2.1 Relationship Between Well Nitrate Concentrations and
Nitrogen Loadings 7-3
7.2.1.1 Included Variables and Data Sources 7-3
7.2.1.2 Omitted Variables 7-6
7.2.2 Modeling of Well Nitrate Concentrations Under Alternate
Regulatory Scenarios 7-6
7.2.3 Discrete Changes from above the MCL to below the MCL 7-7
7.2.4 Incremental Changes below the MCL 7-8
7.2.5 Valuation of Predicted Reductions in Well Nitrate Concentrations 7-9
7.2.5. IPoe and Bishop (1992) 7-11
7.2.5.2 Crutchfield et al. (1997) 7-12
7.2.5.3 De Zoysa (1995) 7-13
7.2.5.4 Adjustments to the Values 7-13
-------
TABLE OF CONTENTS
(continued)
7.2.5.5 Timing of Benefits 7-14
7.3 Results 7-15
7.4 Limitations and Caveats 7-15
7.5 References 7-19
Appendix 7-A: Model Variables 7A-1
Appendix 7-B: The Gamma Model 7B-1
Appendix 7-C: Literature Search and Evaluation 7C-1
INTEGRATION OF RESULTS CHAPTER 8
8.1 Introduction 8-1
8.2 Integration of Analytic Results 8-1
8.3 Present Value of Benefits 8-2
8.4 Annualized Benefits Estimates 8-6
8.5 Limitations of the Analysis and Implications for Characterizing Benefits 8-8
Appendix 8-A: Impact of Alternative Time Frames on Present Value and
Annualized Benefits Estimates 8A-1
Appendix 8-B: Calculation of Present Values 8B-1
Appendix 8-C: Calculation of Annualized Benefits 8C-1
VI
-------
INTRODUCTION AND SUMMARY CHAPTER 1
The U.S. Environmental Protection Agency (EPA) is revising and updating the two primary
regulations that ensure that manure, wastewater, and other process waters generated by concentrated
animal feeding operations (CAFOs) do not impair water quality. EPA's proposed regulatory changes
affect the existing National Pollutant Discharge Elimination System (NPDES) provisions that define
and establish permit requirements for CAFOs, and the existing effluent limitations guidelines (ELGs)
for feedlots, which establish the technology-based effluent discharge standard that is applied to
specified CAFOs. Both of these existing regulations were originally promulgated in the 1970s. EPA
is revising the regulations to address changes that have occurred in the animal industry sectors over
the last 25 years, to clarify and improve implementation of CAFO requirements, and to improve the
environmental protection achieved under these rules.
This report addresses the environmental and economic benefits of several alternative
regulatory scenarios, including two scenarios that EPA is proposing. It examines in detail four
environmental quality improvements that would result from the regulatory changes: improvements
in the suitability of freshwater resources for fishing and swimming; reduced incidence offish kills;
improved commercial shellfishing; and reduced contamination of private wells. Because these are
not the only beneficial impacts of the regulatory scenarios considered by EPA — and because, in
general, EPA takes a conservative approach to quantifying the benefits analyzed — the Agency
believes that this report presents a lower-bound estimate of the beneficial impacts of the proposed
scenarios.
This chapter first defines and describes animal feeding operations and CAFOs, then briefly
summarizes the environmental problems and industry changes associated with animal feeding
operations that EPA is addressing with its proposed regulations. Finally, the chapter outlines the
regulatory changes and alternatives that EPA is considering, and provides a summary of the methods
and results of the more detailed benefits analyses presented in the remaining chapters.
1-1
-------
1.1 DEFINITION OF CAFOS
The term CAFO is a regulatory designation that describes certain animal feeding operations
(AFOs). AFOs are defined by federal regulation as lots or facilities where 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 and crops, vegetation forage growth, or post-harvest residues are not sustained in the normal
growing season over any portion of the lot or facility" (40 CFR 122.23(b)(l)). AFOs congregate
animals on a small land area where feed must be brought to the animals. Winter feeding of animals
on pasture or rangeland is not normally considered an AFO.
Current EPA regulations employ a three-tier structure to identify AFOs that are subject to
regulation as CAFOs. Tier 1 facilities include any animal feeding operation where more than 1,000
"animal units" (AUs) are confined; such facilities are by definition CAFOs unless discharges from
the operation occur only as the result of a 25-year, 24-hour (or more severe) storm event.1 Tier 2
facilities include AFOs that confine 301 to 1000 AUs; these facilities are defined as CAFOs if:
• Pollutants are discharged into navigable waters through a manmade ditch,
flushing system, or other similar man-made device; or
Pollutants are discharged directly into waters that originate outside of and
pass over, across, or through the facility or come into direct contact with the
confined animals.
The regulatory definition of a CAFO does not extend to operations with 300 or fewer AUs (i.e., Tier
3 facilities). Under certain circumstances, however (e.g., a facility causing significant surface water
impairment), a permitting authority may designate such facilities as CAFOs.
Current CAFO regulations address only those facilities with wet-manure management
systems; this eliminates most poultry operations from regulation under the Clean Water Act because
they use dry manure management systems. In addition, the current definition of CAFO includes only
swine over 55 pounds and mature dairy cattle, assuming that immature swine and heifers would be
raised in the same operations as adults. As a result, the regulatory definition does not address the
"stand-alone" immature swine or heifer operations that have proliferated in the last two decades.
USDA reports that there were 1.2 million livestock and poultry operations in the United
States in 1997. This number includes all operations that raise beef or dairy cattle, hogs, chickens
(broilers or layers), and turkeys, and includes both confinement and non-confinement (i.e., grazing
1 Animal units are defined in EPA's current regulations at 40 CFR 122 and vary by animal
type. An AU is considered equivalent to one beef cow.
1-2
-------
and rangefed) production.2 Of these, EPA estimates that there are about 376,000 AFOs that raise
or house animals in confinement, as defined by the existing regulations. For many of the animal
sectors, it is not possible to estimate from available data what proportion of the total livestock
operations have feedlots (i.e., confinement) and what proportion are grazing operations only. For
analytical purposes, EPA has therefore assumed that all dairy, hog, and poultry operations are AFOs.
Exhibit 1-1 summarizes the estimated total number of AFOs of all sizes in each of the four major
livestock categories, based on 1997 data. EPA estimates that only a subset of these AFOs will be
regulated as CAFOs.
Exhibit 1-1
NUMBER OF ANIMAL FEEDING OPERATIONS
(based on 1997 data)
Sector
Beef operations, including both cattle and veal operations.
Dairy operations, including both milk and heifer operations.
Hog operations, including both "farrow to finish" and "grower to finish" operations.
Poultry operations, including broilers, layers (both wet and dry operations) and turkeys.
Sum Total
Total AFOs1
Total AFOs
106,930
118,130
117,860
123,750
466,670
375,740
Source: EPA estimates derived from published USDA/NASS data, including 1997 Census of Agriculture. For more
information, see Technical Development Document of Proposed Effluent Limitations Guidelines for Animal
Feeding Operations.
1 "Total AFOs" eliminates double counting of operations with mixed animal types. Based on survey level
Census data, operations with mixed animal types account for roughly 20 percent of total AFOs.
1.2 CURRENT ISSUES RELATED TO CAFOS
AFOs (including CAFOs) produce and manage large amounts of animal waste, most in the
form of manure. USDA estimates that 291 billion pounds (132 million metric tons) of "as excreted"
manure were generated in 1997 from major livestock and poultry operations. Despite the existing
ELG and NPDES regulations that define CAFOs and regulate their discharges, the management of
animal wastes at AFOs has continued to be associated with environmental problems, including large
spills of manure, fish kills, and outbreaks ofPftesteria. In addition, industry changes in recent years
may contribute to and exacerbate the problems caused by releases of manure from AFOs. EPA is
revising the existing regulations with the following goals:
2 EPA's proposed regulatory changes do not address certain types of animal confinement
operations, such as farms that raise sheep, lambs, goats, horses, and other miscellaneous animal
species, as well as nontraditional animals, such as bison and various exotic species.
1-3
-------
To address reports of continued discharge and runoff of manure and nutrients
from livestock and poultry farms;
To update the existing regulations to reflect structural changes in these
industries over the last few decades; and
To improve the effectiveness of the CAFO regulation.
Below we summarize the potential environmental impacts of manure releases from AFOs,
and outline the recent industry changes that may exacerbate these impacts.
1.2.1 Potential Environmental Impacts of CAFOs
Manure management practices at AFOs can include storage in piles or in open waste lagoons,
followed by land application to agricultural fields as fertilizer. While some discharges from
regulated CAFOs are governed as point sources, unregulated releases of manure from waste piles
or lagoons and overapplication of manure to agricultural lands can affect nearby surface and
groundwater. National and local studies have confirmed the presence of manure pollutants in surface
waters. Once contaminants from manure have reached surface waters they can cause a variety of
ecological and human health problems, including water quality impairments, ecological impacts, and
human health effects from recreational exposure or from contaminated drinking water.
1.2.1.1 Water Quality Impairments
EPA's 1998 National Water Quality Inventory., prepared under Section 305(b) of the Clean
Water Act, identifies agriculture (including irrigated and non-irrigated crop production, rangeland,
feedlots, pastureland, and animal holding areas) as the leading contributor to identified water quality
impairments in the nation's rivers and lakes, and the fifth leading contributor to identified water
quality impairments in the nation's estuaries. The report also identifies the key pollutants and
stressors that impair the nation's waters. Among the most problematic pollutants are several -
including pathogens, nutrients, and oxygen depleting substances - that are associated commonly,
although not exclusively, with animal waste.3
3 The National Water Quality Inventory 1998 Report to Congress notes that 28 states and
tribes reported impairment by agricultural subcategory. Specifically, these states and tribes reported
that animal feeding operations degraded 16 percent of impaired river miles; range and pasture
grazing degraded 11 and 6 percent of impaired river miles, respectively; and irrigated and non-
irrigated crop production degraded 18 and 27 percent of impaired river miles, respectively.
1-4
-------
1.2.1.2 Ecological Impacts
The most dramatic ecological impacts associated with manure pollutants in surface waters
are massive fish kills. Incomplete records indicate that every year dozens offish kills associated with
AFOs result in the deaths of hundreds of thousands offish. In addition, manure pollutants such as
nutrients and suspended solids can seriously disrupt aquatic systems by over-enriching water (in the
case of nutrients) or by increasing turbidity (in the case of solids). Excess nutrients cause
fast-growing algae blooms that reduce the penetration of sunlight in the water column, and reduce
the amount of available oxygen in the water, reducing fish and shellfish habitat and affecting fish
and invertebrates. Manure pollutants can also encourage the growth of toxic organisms, including
Pfiesteria, which has also been associated with fish kills and fish disease events. Reduction in
biodiversity due to animal feeding operations has also been documented; for example, a study of
three Indiana stream systems found fewer fish and more limited diversity offish species downstream
of CAFOs than were found downstream of study reference sites.
1.2.1.3 Human Health Effects
Manure contains over 100 human pathogens; contact with some of these pathogens during
recreational activities in surface water can result in infections of the skin, eye, ear, nose, and throat.
Eutrophication due to excess nutrients can also promote blooms of a variety of organisms that are
toxic to humans either through ingestion or contact. This includes the estuarine dinoflagellate
Pfiesteriapiscicida. While Pftesteria is primarily associated with fish kills and fish disease events,
the organism has also been linked with human health impacts through dermal exposure. Finally,
even with no visible signs of algae blooms, shellfish such as oysters, clams and mussels can carry
toxins produced by some types of algae in their tissue. These can affect people who eat
contaminated shellfish.
Contaminants from manure, including nitrogen, algae, and pathogens, can also affect human
health through drinking water sources and can result in increased drinking water treatment costs.
For example, nitrogen in manure can be transported to drinking water as nitrates, which are
associated with human health risks. EPA has identified nitrate as the most widespread agricultural
contaminant in drinking water wells. Algae blooms triggered by nutrient pollution can affect
drinking water by clogging treatment plant intakes, producing objectionable tastes and odors, and
reacting with the chlorine used to disinfect drinking water to produce harmful chlorinated byproducts
(e.g., trihalomethanes).
1.2.2 Recent Industry Trends
Since EPA promulgated the existing ELG and NPDES regulations governing CAFOs in the
1970s, a number of trends in the livestock and poultry industries have influenced the nature of
pollution from AFOs and the potential for contamination of surface and groundwater. These trends
1-5
-------
include a combination of industry growth and concentration of animals on fewer, larger farms;
location of farms closer to population centers; and advances in farm production practices and waste
management techniques. The changes in the industry have limited the effectiveness of the current
regulations that define and govern releases from CAFOs.
1.2.2.1 Increased Production and Industry Concentration
U.S. livestock and poultry production has risen sharply since the 1970s, resulting in an
increase in the amount of manure and wastewater generated annually. The Census of Agriculture
reports 1997 turkey sales of 299 million birds, compared to 141 million sold in 1978. Sales of
broilers increased to 6.4 billion in 1997 from 2.5 billion in 1974.4 Red meat production also rose
during the 1974-1997 period; the number of hogs and pigs sold in 1997 totaled 142.6 million,
compared to 79.9 million in 1974.
As production has increased, the U.S. livestock and poultry sectors have also consolidated
animal production into a smaller number of larger-scale, highly specialized operations that
concentrate more animals (and manure) in a single location. At the same time, significant gains in
production efficiency have increased per-animal yields and the rate of turnover of animals between
farm and market. These large AFOs can present considerable environmental risks, because they often
do not have an adequate land base for manure disposal through land application. As a result, large
facilities must incur the risks associated with storing significant volumes of manure, or must attempt
to maximize the application of manure to the limited land they have available. By comparison,
smaller AFOs manage fewer animals and tend to concentrate less manure nutrients at a single
location. These operations are more likely to have sufficient cropland and fertilizer needs to land
apply manure nutrients generated at a livestock or poultry business.
1.2.2.2 Location of Animal Operations Closer to Consumer Markets
Since the 1970s, the combined forces of population growth and re-location of operations
closer to consumer markets and processing sectors have resulted in more AFOs located near densely
populated areas. Surface waters in these areas face additional stresses from urban runoff and other
point sources. The proximity of large AFOs to human populations thus increases the potential for
human health impacts and ecological damage if manure and wastewater at AFOs is improperly
discharged.
4 This more than two-fold increase in the number of broilers raised annually signals the need
to review the existing CAFO regulations, which effectively do not cover broiler operations since
virtually no such operations use wet manure management systems.
1-6
-------
1.2.2.3 Advances in Agriculture Production Practices to Manage and Dispose Manure
Continued research by USDA, state agencies and universities has led to advances in
technologies and management practices that minimize the potential environmental degradation
attributable to discharge and runoff of manure and wastewater. Today, there are many more
practicable options to properly collect, store, treat, transport, and utilize manure and wastewater than
there were in the 1970s, when the existing regulations were instituted. As a result, current
regulations do not reflect the full range of management practices and technologies that may be
implemented to achieve greater protection of the environment (e.g., by more effectively treating
certain constituents present in animal manure or by converting manure into a more marketable form).
In addition, during the time since promulgation of the existing regulation, certain practices have
proven to be relatively less protective of the environment. There is documented evidence that
lagoons may leak if not properly maintained, and evidence of overapplication of manure and nutrient
saturation of soils in some parts of the country.
1.3 PROPOSED CHANGES TO CAFO REGULATIONS
In response to persistent reports of environmental problems, and to changes in the industries
and technologies associated with AFOs, EPA is proposing changes to both the NPDES regulations
for CAFOs and the ELG regulations for feedlots. Proposed changes to the NPDES regulations for
CAFOs affect which animal feeding operations are defined as CAFOs and are therefore subject to
the NPDES permit program. Changes to the ELG regulations for feedlots affect which technology-
based requirements will apply to CAFOs.
EPA's analysis of the benefits of revised regulations considers four alternatives for the
NPDES definition of a CAFO (described as Scenarios 1, 2/3, 4a, and 4b), combined with two
alternative ELG regulations (Options 1 and 2), yielding a total of eight regulatory scenarios. EPA
is co-proposing two of these scenarios. The first incorporates NPDES scenario 2/3 and ELG Option
2; this scenario would preserve the current three-tier structure for identifying facilities that are
CAFOs (though with revised conditions for identifying CAFOs within the tiers), and would revise
the ELG to establish a phosphorus-based manure application limit. The second proposed scenario
incorporates NPDES scenario 4a and ELG Option 2; this scenario would replace the current three-
tier structure with a two-tier structure, and would also incorporate a phosphorus-based ELG.
Specific proposed changes are described in more detail below, and are summarized in Exhibit 1-2.
1.3.1 Changes to NPDES Regulations
EPA considered four regulatory scenarios that reflect changes to the current approach to
determining which facilities are CAFOs that are subject to NPDES requirements. Scenario 1 would
retain the existing three-tier structure for identifying CAFOs (described in section 1.1). Scenario 2/3
would also retain the current three-tier structure, but would revise the conditions within the tiers for
determining which facilities are CAFOs. Scenario 4a would replace the current three-tier structure
1-7
-------
with a two-tier structure that would alter the definition of a CAFO to include all AFOs with 500 or
more AUs; operations with fewer than 500 AUs would be regulated at the discretion of the
permitting authority (similar to the current Tier 3 facilities). Finally, Scenario 4b would change to
a two-tier structure similar to that in Scenario 4a, but would define as a CAFO any operation with
300 or more animal units.5
As noted above, EPA has chosen to co-propose NPDES Scenario 2/3 and NPDES Scenario
4a. In doing so, the Agency is soliciting comments on regulatory approaches that, on a national
basis, yield similar environmental benefits, but offer different administrative benefits and have
differing impacts on regulated industry sectors. Specifically:
Scenario 2/3 would apply a three-tier structure combined with a risk-based
approach to identify which AFOs pose a potential to discharge. This scenario
would automatically define all operations over 1,000 AUs as CAFOs. AFOs
with between 300 and 1,000 AUs would be required to either apply for an
NPDES permit or certify to the permitting authority that they do not meet any
of the conditions that define a CAFO. An advantage of this approach is that
it would offer states flexibility in developing requirements and programs that
could reduce the number of facilities needing NPDES permits. A potential
disadvantage, however, is the complexity associated with administering this
approach, as well as the cost associated with extending the
certification/application requirement to facilities as small as 300 AUs.
In contrast to Scenario 2/3, the two-tier structure of Scenario 4a would define
all operations with at least 500 AUs as CAFOs. As such, all facilities with
at least 500 AUs would be required to obtain and comply with an NPDES
permit; operations with fewer than 500 AUs would be subject to permitting
only if designated by the permitting authority as a significant contributor of
pollution. An advantage of this approach is that it simplifies the structure of
the regulations and supports EPA's goal of clarifying their scope. In addition,
operations with at least 300 but fewer than 500 AUs would not automatically
incur permitting or certification costs; however, the potential benefits
associated with a more flexible, risk-based approach to permitting operations
with between 500 and 1,000 AUs would be foregone.
5 Each of the regulatory scenarios analyzed also reflects several proposed structural changes
that would revise the CAFO definition and permit requirements under the NPDES permit program.
For example, EPA is proposing to include dry poultry and stand-alone immature swine and heifer
operations as AFOs; this change would increase the number of facilities that meet the definition of
a CAFO and must obtain an NPDES permit. In addition, EPA is proposing several clarifications
designed to assure that all facilities meeting the CAFO definition obtain an NPDES permit.
Similarly, EPA is proposing changes to the ELGs for feedlots that would clarify the development
of technical standards for manure storage and land application operations. These changes from the
current baseline are reflected in all of the regulatory scenarios.
1-8
-------
Exhibit 1-2
REGULATORY SCENARIOS CONSIDERED IN
THE BENEFITS ANALYSIS
Regulatory
Scenario
Baseline
Option 1-
Scenario 1
Option 1-
Scenario 2/3
Option 1-
Scenario 4a
Option 1-
Scenario 4b
Option 2-
Scenario 1
Option 2-
Scenario
2/3*
Option 2-
Scenario 4a*
Option 2-
Scenario 4b
NPDES Revisions
CAFOs include any AFO with over 1,000 AUs, as well as AFOs with
fewer AUs that meet certain requirements.
Baseline scenario plus dry poultry and immature swine and heifer
operations.
New NPDES conditions for identifying CAFOs among AFOs with 300
- 1000 AUs, plus dry poultry and immature swine and heifer operations.
CAFOs include all AFOs with 500 or more AUs, plus dry poultry,
immature swine and heifer operations.
CAFOs include all AFOs with 300 or more AUs, plus dry poultry,
immature swine and heifer operations.
Baseline scenario plus dry poultry and immature swine and heifer
operations.
New NPDES conditions for identifying CAFOs among AFOs with 300
- 1000 AUs, plus dry poultry and immature swine and heifer operations.
CAFOs include all AFOs with 500 or more AUs, plus dry poultry,
immature swine and heifer operations.
CAFOs include all AFOs with 300 or more AUs, plus dry poultry,
immature swine and heifer operations.
ELG Revisions
Manure application
not regulated
Nitrogen-based
manure application
Nitrogen-based
manure application
Nitrogen-based
manure application
Nitrogen-based
manure application
Phosphorus-based
manure application
Phosphorus-based
manure application
Phosphorus-based
manure application
Phosphorus-based
manure application
* Proposed scenarios.
1.3.2 Changes to ELGs
EPA's proposed changes to the effluent limitation guidelines would include a technical
standard for nutrient-based land application of manure. The Agency is considering two regulatory
options. Option 1 would limit manure application to a nitrogen-based agronomic application rate
(i.e., manure application could not exceed the soil and crop demand for the nitrogen within the
manure). Option 2 would limit manure application to a phosphorus-based agronomic application
rate (i.e., manure application could not exceed the soil and crop demand for the phosphorus within
the manure).
1-9
-------
EPA's proposed regulatory scenarios both reflect the phosphorus standard. Because manure
is phosphorus rich, nitrogen-based manure application is likely to result in application of phosphorus
in excess of crop requirements. Although excess phosphorus does not usually harm crops and is
often adsorbed by soils, the capacity of soil to adsorb phosphorus will vary by soil type. Recent
observations indicate that soils can and do become saturated with phosphorus. When saturation
occurs, continued application of phosphorus in excess of what can be used by the crop and soil will
result in phosphorus leaving the field with storm water via leaching or runoff; eutrophication of
surface waters can result.
1.3.3 Number of Regulated Operations
EPA has estimated the likely number of AFOs that would be regulated under the revised
definition of CAFO in each of the four NPDES scenarios (i.e., Scenarios 1, 2/3, 4a, and 4b). EPA
analyzed data from the USDA's 1997 Census of Agriculture to identify AFOs and CAFOs. EPA first
determined the number of operations that raise animals under confinement by using available data
on the total number of livestock and poultry facilities. Next, EPA determined the number of CAFOs
based on the number of facilities that discharge or have the potential to discharge to U.S. waters and
which meet a minimum size threshold (i.e., number of animals) as defined by the regulatory options.
Exhibit 1-3 shows the number of CAFOs estimated for each scenario.
1.4 ANALYTIC METHODS AND RESULTS
To determine the economic benefits of the regulatory scenarios, EPA performed four separate
analyses of expected changes in environmental quality that would likely result from reduced AFO
pollution. These include:
Improvements in Water Quality and Suitability for Recreational
Activities: this analysis estimates the economic value of improvements in
inland surface water quality that would increase opportunities for recreational
fishing and swimming;
• Reduced Incidence of Fish Kills: this analysis estimates the economic
value of a potential reduction in the number of fish kills caused by AFO-
related waste;
Improved Commercial Shellfishing: this analysis characterizes the impact
of pollution from AFOs on access to commercial shellfish growing waters,
and values the potential increase in commercial shellfish harvests that may
result from improved control of that pollution; and
1-10
-------
Exhibit 1-3
ESTIMATED NUMBER OF CAFOS UNDER ALTERNATIVE REGULATORY SCENARIOS*
Production
Sector
Beef
Dairy
Heifers
Veal
Swine
Layers
Broilers
Turkeys
Total
Currently
Regulated
2,220
3,150
620
20
5,260
470
620
50
12,410
NPDES
Scenario 1
2,290
3,560
590
20
5,630
870
4,320
420
17,700
NPDES
Scenario 2/3
2,720
5,430
830
70
7,520
1,420
13,830
1,680
33,500
NPDES
Scenario 4a
3,080
3,760
800
90
8,550
1,640
9,780
1,280
28,980
NPDES
Scenario 4b
4,080
7,140
1,050
210
14,370
2,050
14,140
2,100
45,140
* AFOs with more than one animal type are counted more than once; numbers have been rounded to nearest
ten.
Reduced Contamination of Private Wells: this analysis examines the
impact of the revised regulations on groundwater quality, and values
predicted improvements in the quality of aquifers that supply private wells.
Exhibit 1-4 summarizes the results of these four studies for each of the regulatory scenarios.
It is important to note that these results are not intended to represent the total value of all benefits
associated with a reduction in AFO pollutants; they include only the subset of benefits that is
addressed by EPA's analyses. Moreover, EPA's analyses generally take a conservative approach to
quantifying benefits; therefore, the results are likely to reflect conservative estimates of the specific
benefits that EPA has examined.
1.5 ORGANIZATION OF REPORT
The remainder of this report presents EPA's analysis of the benefits expected under each of
the regulatory scenarios considered. Specifically:
• Chapter 2 provides a detailed description of the potential impacts of AFOs
on environmental quality and human health;
1-11
-------
Exhibit 1-4
ESTIMATED ANNUALIZED BENEFITS OF CHANGES IN REGULATION OF CAFOS
(1999 dollars, millions)
Regulatory Scenario
Option 1- Scenario 1
Option 1- Scenario 2/3
Option 1- Scenario 4a
Option 1- Scenario 4b
Option 2- Scenario 1
Option 2- Scenario 2/3*
Option 2- Scenario 4a*
Option 2- Scenario 4b
Recreational
and Non-use
Benefits
$4.9
$6.3
$5.5
$7.2
$87.6
$127.1
$108.5
$145.0
Reduced
Fish Kills
$0.1 -$0.2
$0.1 -$0.3
$0.1 -$0.3
$0.1 -$0.3
$0.2 - $0.3
$0.2 - $0.4
$0.2 - $0.4
$0.2 - $0.4
Improved
Shellfishing
$0.1 -$1.8
$0.2 - $2.4
$0.2 - $2.2
$0.2 - $2.6
$0.2- $2.1
$0.2 - $2.7
$0.2 - $2.4
$0.2 - $3.0
Reduced
Private Well
Pollution
$33. 3 -$49.0
$33. 3 -$49.1
$35. 5 -$52.2
$35. 5 -$52.2
$35.4 -$52.1
$35.4 -$52.1
$36.6 -$53. 9
$36.6 -$53.9
Total
$38.4 -$55. 9
$39.9 -$58.0
$41.2 -$60.2
$43.0 -$62.3
$123.3- $142.1
$163.0 -$182.3
$145.5 -$165.1
$182.1 -$202.2
* Proposed scenarios.
Chapter 3 describes the range of benefits that would result from decreased
AFO loadings, and outlines EPA's general approach to quantifying and
valuing the subset of benefits analyzed;
Chapter 4 assesses the value of changes in surface water quality that would
result from a reduction in AFO loadings, focusing on changes in the quality
of freshwater resources that would improve their suitability for fishing and
swimming;
Chapter 5 assesses the value of reducing the incidence of fish kills
attributable to pollution from AFOs;
Chapter 6 assesses the value of improved commercial shellfishing resulting
from decreased AFO loadings;
Chapter 7 assesses the value of reduced contamination of private wells
associated with reductions in the pollution of groundwater by AFOs; and
Chapter 8 provides the summary and conclusions of the benefits analysis.
1-12
-------
1.6 REFERENCES
USDA/USEPA (U.S. Department of Agriculture and U.S. Environmental Protection Agency). 1999.
Unified National Strategy for Animal Feeding Operations, Section 4.2. Available on EPA's
web site at: http://www.epa.gov/owm/fmafost.htnrfl.0.
1-13
-------
POTENTIAL IMPACTS OF AFOS ON
ENVIRONMENTAL QUALITY AND HUMAN HEALTH CHAPTER 2
Animal manure, the primary cause of pollution related to AFOs, contains a variety of
pollutants that can cause environmental degradation, particularly when released to surface waters in
large quantities.6 Documented releases from AFOs have been associated with a number of adverse
human health and ecological impacts, including fish kills, disease outbreaks, and degradation of
water quality and aquatic life.
EPA's 1998 National Water Quality Inventory, prepared under Section 305(b) of the Clean
Water Act, presents recent information on impaired water bodies nationwide. The Inventory
identifies agriculture (including irrigated and non-irrigated crop production, range grazing, pasture
grazing, and animal feeding operations) as the leading contributor to identified water quality
impairments in the nation's rivers and lakes, and the fifth leading contributor to identified water
quality impairments in the nation's estuaries. The report also identifies the key pollutants and
stressors that impair the nation's waters. Among the most problematic pollutants are several -
including pathogens, nutrients, and oxygen depleting substances - that are associated commonly,
although not exclusively, with animal waste (USEPA, 2000).7
The animal waste management practices and pollutant transport pathways that can lead to
contamination of surface waters are well known. Animal wastes at AFOs are typically managed by
land application and/or storage in waste piles or lagoons. Land application and storage of manure
are centuries-old farming practices. In small or low-density farming operations these methods pose
6 This document uses the term manure to refer to both "solid" manure and urine, since these
wastes are typically managed together. Additional animal wastes associated with AFOs (e.g., hair,
feathers, bedding material and carcasses) are identified separately in the discussion.
7 The National Water Quality Inventory 1998 Report to Congress notes that 28 states and
tribes reported impairment by agricultural subcategory. Specifically, these states and tribes reported
that animal feeding operations degraded 16 percent of impaired river miles; range and pasture
grazing degraded 11 and 6 percent of impaired river miles, respectively; and irrigated and non-
irrigated crop production degraded 18 and 27 percent of impaired river miles, respectively.
2-1
-------
minimal pollution potential. AFOs, however, manage large amounts of manure in a concentrated
area. Under these circumstances, the following waste management failures pose an increased
potential for pollution:
Over-application of manure: While land application of manure can provide
valuable nutrients to soil and crops, the capacity of soil and crops to absorb
nutrients over any given period is limited. Excess manure applied to
cropland can damage crops and soil, and is more likely to run off into surface
waters or be released to air through volatilization or erosion (for example,
through spray application)..
Runoff from uncovered manure piles: Manure piles are frequently used
for temporary storage of animal wastes. Precipitation may wash pollutants
from uncovered manure piles into nearby surface waters.
• Lagoon failures: AFOs frequently store large quantities of manure in
lagoons prior to land application or other disposal. While lagoons are
designed to prevent the release of wastes into the environment, they are
subject to various types of failure, including spills due to overfilling;
washouts in floods; liner failures; failures of dikes, pipes, or other above-
ground structures; and accidental and intentional operator-related releases.
This chapter briefly describes the pathways, pollutants, and environmental and human health effects
associated with releases from AFOs. More detailed information is available in Environmental
Assessment of the Proposed Effluent Limitation Guidelines for Concentrated Animal Feeding
Operations (available in Section 8.1 of the Record).
2.1 PATHWAYS FOR THE RELEASE OF POLLUTANTS FROM AFOS
Pollutants in animal wastes can reach surface waters by several pathways, including overland
discharge, migration through groundwater, and atmospheric deposition. The most common pathway
is overland discharge, which includes surface runoff (i.e., land-applied or piled manure that is
washed into surface waters by rain), soil erosion, and acute events such as spills or impoundment
failures. Contamination can also occur when pollutants leach through soil into groundwater and then
to surface water through groundwater recharge. In addition, airborne pollutants created by
volatilization or by spray-application of manure to land can contaminate surface water through
atmospheric deposition. Exhibit 2-1 illustrates the various pathways by which AFO releases can
affect surface waters and groundwater. The following discussion describes these pathways in
greater detail.
2-2
-------
Waste
Lagoon
Exhibit 2-1
PATHWAYS FOR AFO-RELATED POLLUTANTS
Atmospheric
deposition
Land
Application
Site
Groundwater
recharge
2-3
-------
2.1.1 Overland Discharge
Contamination from manure often reaches surface water though overland discharge; that is,
by flowing directly into surface waters from land application sites or lagoons. There are three
distinct types of overland discharge: surface runoff, soil erosion, and direct discharge of manure to
surface water during acute events. For example, a single flood event might include lagoon
"washouts," soil erosion and surface runoff. This section describes the various types of overland
discharge in more detail.
2.1.1.1 Surface Runoff
Surface runoff occurs whenever rainfall or snowmelt is not absorbed by soil and flows
overland to surface waters.8 Runoff from land application sites or manure piles can transport
pollutants to surface waters, especially if rainfall occurs soon after application, if manure is over-
applied, or if it is misapplied.9 The potential for runoff of animal wastes varies considerably with
climate, soil conditions, and management practices. For example, manure applied to saturated or
frozen soils is more likely to runoff the soil surface (ODNR, 1997). Other factors that promote
runoff to surface waters are steep land slope, high rainfall, low soil porosity or permeability, and
close proximity to surface waters.
Surface runoff is a particularly significant transport mechanism for water soluble pollutants,
including nitrogen compounds. However, runoff can also carry solids. Runoff of manure pollutants
has been identified as a factor in a number of documented impacts from AFOs, including hog, cattle,
and chicken operations. For example, in 1994, multiple runoff problems were cited for a hog
operation in Minnesota, and in 1996 runoff from manure spread on land was identified at hog and
chicken operations in Ohio. In 1996 and 1997, runoff problems were identified for several cattle
operations in numerous counties in Minnesota (CWAA, 1998; ODNR, 1997).
2.1.1.2 Soil Erosion
In addition to simple surface runoff, pollutants from animal wastes can enter surface water
through erosion, in which the soil surface itself is worn away by the action of water or wind. Soil
erosion often occurs in conjunction with surface runoff as part of rainfall events, but it represents a
transport mechanism for additional pollutants that are strongly sorbed (i.e., chemically bound) to
8 Surface discharges can also result from direct contact between confined animals and surface
waters. Certain animals, particularly cattle, will wade into the surface waters to drink, and will often
urinate and defecate there as well. This practice is now restricted for CAFOs, but may still occur at
other types of AFOs.
9 Experiments show that for all animal wastes, application rates have a significant effect on
runoff concentrations of pollutants. See Daniel et al, 1995.
2-4
-------
soils. The most important of these pollutants is phosphorus. Because of its tendency to sorb to soils,
many agricultural phosphorus control measures focus on soil erosion control. However, soils do not
have infinite adsorption capacity for phosphorus or other pollutants, and dissolved pollutants
(including phosphates) can still enter waterways through runoff even if soil erosion is controlled
(NRC, 1993).
In spite of control efforts, soil
erosion remains a serious challenge for
agriculture. For example, in 1997 the
USDA Natural Resources Conservation
Service (NRCS) reviewed the connection
between manure production, soil erosion,
and water quality in a watershed in South
Carolina. NRCS calculated that soil
erosion from the 13,000 acres of cropland
in the watershed ranged from 9.6 to 41.5
tons per acre per year. The report further
found that manure and erosion-related
pollutants such as bacteria, nutrients, and
sediment are the primary contaminants
affecting streams and ponds in the
watershed (USEPA, 1997).
Catastrophic Release of Manure:
New River, North Carolina, 1995
On June 21, 1995, a breach in the dike of a 30 million
gallon hog waste lagoon discharged over 25 million
gallons of waste into tributaries of the New River in
Onslow County, North Carolina.
Within a week of the event, North Carolina state officials
estimated that roughly 2,600 fish were destroyed, though
monitoring indicated that oxygen levels had recovered in
the river within a week of the event. JoAnne Burkholder,
a North Carolina State University marine scientist,
noted that the initial waste deluge probably smothered
many fish. Others were killed more slowly by declining
oxygen levels and the toxic effects of ammonia and
bacteria in the water.
Two days after the spill scientists sampling in some of
the affected areas found ammonia levels of about 20
times the lethal limit for most fish.
Though oxygen levels recovered rapidly, Burkholder
noted that it could take years for the upper New
ecosystem to fully recover and support the range offish,
clams and other creatures that existed before the spill.
In addition to immediate problems, longer term
problems caused by the breach would include rains
churning up settled pollution and potential algae
blooms.
State environmental officials also confirmed that high
levels of fecal coliform bacteria were detected in the
river, and Onslow County healthy officials posted
warnings in public recreation areas to prevent people
from swimming. According to local newspaper reports,
in some places fecal coliform levels were 10,000 times
the state standard for swimming.
Sources: War rick and Stith, 199 5 b; War rick 199 5b,
1995c, 199 5 d.
2.1.1.3 Acute Events
In addition to surface runoff and
erosion, acute events such as spills,
floods, or other lagoon or application
failures can affect surface waters. Unlike
runoff and erosion, which generally affect
land-applied wastes, acute events
frequently affect waste management
lagoons. Spills can result from
mechanical malfunctions (e.g., pump
failures, manure irrigation gun
malfunctions, and failures in pipes or
retaining walls), overfilling, or washouts
during flood events. There are even
indications that some operators discharge
wastes into surface waters deliberately in
order to reduce the volume of waste in ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
overfull lagoons (CWAA, 1998). Acute
events frequently result in large waste discharges and are often associated with immediate ecological
2-5
-------
effects such as fish kills. Furthermore, In addition to immediate fish kills, large releases can be
linked with eutrophication, sedimentation, and the growth of pathogens. All of these impacts can
also cause acute mortality in fish and other aquatic species.
2.1.2 Leaching to Groundwater
Pollutants from animal waste can migrate to groundwater and subsequently contaminate
surface waters through the process of "groundwater recharge," in which hydrological connections
between aquifers and surface waters allow transfer of water (and pollutants). Groundwater
contamination itself can result from leaching of land-applied pollutants into the soil, or from leaking
lagoons. Although most lagoons are lined with clay or are designed to be "self-sealed" by manure
solids that prevent infiltration of pollutants into groundwater, these methods are not always
effective. For example, a survey of hog and poultry lagoons in the Carolinas found that the contents
of nearly two-thirds of the 36 lagoons sampled had leaked into the groundwater (Meadows, 1995).
Similarly, clay-lined lagoons can crack or break as they age, and are susceptible to burrowing worms.
In a three-year study of clay-lined swine lagoons on the Delmarva Peninsula, researchers found that
leachate from lagoons located in well-drained loamy sand adversely affected groundwater quality
(Ritter ef or/., 1990).
Surface water contamination from groundwater is most likely to occur in areas with high soil
permeability and shallow water tables, and is most likely to involve water soluble contaminants such
as nitrogen (Smith et al, 1997). Overall, the potential for contamination by this pathway may be
considerable. For example, in the Chesapeake Bay watershed, the USGS estimates that about half
of the nitrogen loads from all sources to non-tidal streams and rivers originates from groundwater
(ASCE, 1998). In addition, about 40 percent of the average annual stream flow in the United States
results from groundwater recharge (USEPA, 1993).
2.1.3 Discharges to the Air and Subsequent Deposition
Discharges to the air from AFOs include both volatile pollutants (e.g., ammonia and various
by-products of manure decomposition) and particulate matter such as dried manure, feed, hair, and
feathers. The degree of volatilization of pollutants from manure depends on environmental
conditions and the manure management system employed. For example, spray application of manure
increases the potential for volatilization, as does the practice of spreading manure on the land
without incorporating it into the soil. Volatilization is also affected by climate and soil conditions,
(e.g., soil acidity and moisture content), and is reduced by the presence of growing plants (Follett,
1995).
2-6
-------
Particulate matter from manure forms an organic dust made up of dried manure, feed, and
epithelial cells. These airborne particles can contain adsorbed gases, endotoxin (the toxic protoplasm
liberated when a microorganism dies and disintegrates), and possibly steroids from animal waste.
According to information presented to the Centers for Disease Control, at least 50 percent of the dust
emissions from swine operations are believed to be respirable and may therefore be associated with
inhalation-related human health effects (Thu, 1998).10
In addition to creating the potential for air-related health effects, both volatilized pollutants
and particulate matter can contaminate nearby surface waters through atmospheric deposition.
Volatilization of the ammonia in urine, in particular, has been linked with atmospheric deposition
of nitrogen (Lander et al, 1998). While it is not clear what percentage of total deposition of
pollutants can be linked to AFOs, EPA's 1998 National Water Quality Inventory indicates that
atmospheric deposition from all sources is the third greatest cause of water quality impairment for
estuaries, and the fifth greatest cause of water quality impairment for lakes, reservoirs and ponds.
2.2 POTENTIAL ECOLOGICAL HAZARDS
POSED BY AFO POLLUTANTS
The primary pollutants associated with animal waste are nutrients (particularly nitrogen and
phosphorus), organic matter, solids, pathogens, and odorous/volatile compounds. Animal waste is
also a source of salts and trace elements and, to a lesser extent, antibiotics, pesticides, and hormones.
The concentration of particular pollutants in manure varies with animal species, the size, maturity,
and health of the individual animal, and the composition (e.g., protein content) of animal feed.11 The
range of pollutants associated with manure is evident in a 1991 U.S. Fish and Wildlife Service
(USFWS) report on suspected impacts from cattle feedlots on Tierra Blanca Creek in the Texas
Panhandle. The impacts the USFWS reported included elevated concentrations of ammonia,
coliform bacteria, chloride, nitrogen, and volatile suspended solids, as well as reduced concentrations
of dissolved oxygen. In addition, USFWS found elevated concentrations of the feed additives copper
and zinc in creek sediment (USFWS, 1991).
The ecological impacts of animal waste releases to surface water can range from minor,
temporary fluctuations in water quality (e.g., associated with limited surface runoff) to chronic
degradation of ecosystems (e.g., associated with consistently poor management practices such as
over-application), to dramatic impacts such as extensive fish or wildlife kills (e.g., associated with
acute events such as spills or toxic algae blooms). In some cases, individual pollutants associated
with animal waste are the clear and direct cause of observable ecological effects. In other cases,
10 "Respirable" generally refers to particles less than 10 microns in diameter, or PM10; these
particles are responsible for the majority of human health effects related to air pollution because they
are small enough to travel through the nasal passage and into the lungs.
11 For more detailed discussion of the pollutants associated with animal waste, see Phillips
etal, 1992.
2-7
-------
ecological effects such as declines in aquatic populations are the result of complex systemic changes
that are linked directly or indirectly to pollution from AFOs.
Exhibit 2-2 lists the key pollutants associated with AFO waste, and notes their potential
impacts. The remainder of this section describes in more detail the relationship between AFO
pollutants and observed ecological effects. Section 2.3 focuses on the specific impacts of AFO
pollutants on human health.
Exhibit 2-2
KEY POLLUTANTS IN ANIMAL WASTE
Pollutant
Description of Pollutant Forms
in Animal Waste
Pathways
Potential Impacts
Nutrients
Nitrogen
Phosphorus
Potassium
Organic
Compounds
Solids
Pathogens
Salts
Trace Elements
Exists in fresh manure in organic (e.g.,
ammonia in urea) and inorganic forms (e.g.,
ammonium and nitrate). Microbes transform
organic nitrogen to inorganic forms that are
absorbed by plants.
Exists in both organic (water soluble) and
inorganic forms. As manure ages, phosphorus
mineralizes to inorganic phosphate compounds
that are absorbed by plants.
Most potassium in manure is in an inorganic
form available for absorption by plants; it can
also be stored in soil for future uptake.
Carbon-based compounds in manure that are
decomposed by surface water micro-
organisms. Creates biochemical oxygen
demand, or BOD, because decomposition
consumes dissolved oxygen in the water.
Includes manure itself and other elements
(e.g., feed, bedding, hair, feathers, and
corpses).
Includes range of disease-causing organisms,
including bacteria, viruses, protozoa, fungi,
and algae. Some pathogens are found in
manure, others grow in surface water due to
increased nutrients and organic matter.
Includes soluble salts containing cations
sodium and potassium (from undigested feed),
calcium and magnesium; and anions chloride,
sulfate, bicarbonate, carbonate, and nitrate.
Includes feed additives arsenic, copper,
selenium, zinc, cadmium; and trace metals
molybdenum, nickel, lead, iron, manganese,
aluminum, and boron (pesticide ingredients).
• Overland discharge
• Leachate into
groundwater
• Atmospheric deposition
as ammonia
• Overland discharge
• Leachate into
groundwater (water
soluble forms)
• Overland discharge
• Leachate into
groundwater
• Overland discharge
• Overland discharge
• Atmospheric deposition
• Overland discharge
• Growth in waters with
high nutrient, organic
materials
• Algal by-products
• Overland discharge
• Leachate into
groundwater
• Overland discharge
• Eutrophication
• Animal, human
health effects
• Eutrophication
• Eutrophication
• Increased salinity
• Depletion of
dissolved oxygen
• Reduction in aquatic
life
• Turbidity
• Siltation
• Animal, human
health effects
• Reduction in aquatic
life
• Human health
effects
• Toxicity at high
levels
2-8
-------
Exhibit 2-2
KEY POLLUTANTS IN ANIMAL WASTE
Pollutant
Odorous,
Volatile
Compounds
Other
Pollutants
Description of Pollutant Forms
in Animal Waste
Includes carbon dioxide, methane, hydrogen
sulfide, and ammonia gases generated during
decomposition of waste.
Includes pesticides, antibiotics, and hormones
used in feeding operations.
Pathways
• Inhalation
• Atmospheric deposition
of ammonia
• Overland discharge
Potential Impacts
• Human health
effects
• Eutrophication
• Impacts unknown
2.2.1 Nutrients and Eutrophication
EPA's 1998 National Water Quality Inventory indicates that nutrients from all sources
comprise the leading stressor in impaired lakes, ponds, and reservoirs, and are among the most
frequent stressors in impaired rivers, streams, and estuaries. Nutrients are naturally occurring
elements that are necessary for plant growth. However, when excess nutrients enter surface waters
they can stimulate overgrowth of algae and bacteria, changing ecosystems in a process called
"eutrophication." In addition, nutrients (nitrogen, in particular) in high concentrations can be toxic
to animals and humans.
The two nutrients of most concern related to AFOs are nitrogen and phosphorus.12 Each of
these elements exists in several forms in the environment, and is involved in several phases of uptake
and digestion by animals and plants. This section briefly describes the processes by which nitrogen
and phosphorus enter aquatic ecosystems, then discusses the process and impacts of eutrophication.
2.2.1.1
Nitrogen and Nitrogen Compounds
Nitrogen, an element essential to plant growth, moves through the environment in a series
of chemical reactions known as the nitrogen cycle. Nitrogen in manure exists in both organic forms
(e.g., urea) and inorganic forms (e.g., ammonium and nitrate) (NCAES, 1982). In fresh manure, 60
to 90 percent of total nitrogen is present in the organic form. Inorganic nitrogen can enter the
environment by volatilizing in the form of ammonia, or through soil or water microbe processes that
transform organic nitrogen to an inorganic form that can be used by plants (i.e., as fertilizer). Both
ammonia and ammonium are toxic to aquatic life, and ammonia in particular reduces the dissolved
oxygen in surface waters that is necessary for aquatic animals. Nitrites pose additional risks to
12 Potassium contributes to the salinity of animal manure, which may in turn contribute
salinity to surface water polluted by manure. Actual or anticipated levels of potassium in surface
water and groundwater, however, are unlikely to pose hazards to human health or aquatic life. For
more information see Wetzel, 1983.
2-9
-------
aquatic life: if sediments are enriched with nutrients, nitrite concentrations in the water may be
raised enough to cause nitrite poisoning or "brown blood disease" in fish (USDA, 1992).
A 1975 study found that up to 50
percent or more of the nitrogen in fresh
manure may be in ammonia form or
converted to ammonia relatively quickly
once manure is excreted (Vanderholm,
1975). Ammonia is highly volatile, and
ammonia losses from animal feeding
operations can be considerable. In North
Carolina, animal agriculture is responsible
for over 90 percent of all ammonia
emissions; ammonia composes more than 40
percent of the total estimated nitrogen
emissions from all sources. Once airborne,
these volatile pollutants may be deposited
onto nearby streams, rivers, and lakes. Data
from Sampson County, North Carolina show
that "ammonia rain" has increased as the hog
industry has grown, with ammonia levels in
rain more than doubling between 1985 and
1995(Anejaera/., 1998).
National Study of Nitrogen Sources to Watersheds
In 1994, the USGS analyzed potential
nitrogen sources to 107 watersheds, including
manure (from both confined and unconflned animals),
fertilizers, point sources, and atmospheric deposition.
While the study found that proportions of nitrogen
originating from various sources differ according to
climate, hydrologic conditions, land use, population,
and physical geography, results for selected
watersheds for the 1987 base year showed that in
some instances, nitrogen from manure represents a
large portion of the total nitrogen added to the
watershed. For example, in nine study watersheds
more than 25 percent of nitrogen originates from
manure.
Source: Puckett, 1994.
Ammonia is highly toxic to aquatic life and is a leading cause offish kills. In a May 1997
incident in Wabasha County, Minnesota, ammonia in a dairy cattle manure discharge killed 16,500
minnows and white suckers (CWAA, 1998). In addition, ammonia and other pollutants in manure
exert a direct biochemical oxygen demand (BOD) on the receiving water. As ammonia is oxidized,
dissolved oxygen is consumed. Moderate depressions of dissolved oxygen are associated with
reduced species diversity, while more severe depressions can produce fish kills (USFWS, 1991).
2.2.1.2
Phosphorus
Like nitrogen, phosphorus is necessary for the growth of plants, but is damaging in excess
amounts. Phosphorus exists in solid and dissolved phases, in both organic and inorganic forms. Over
70 percent of the phosphorus in animal manure is in the organic form (USD A, 1992). As manure
ages, phosphorus mineralizes to inorganic phosphate compounds which are available to plants.
Organic phosphorus compounds are generally water soluble and may leach through soil to
groundwater or runoff into surface waters. In contrast, inorganic phosphorus tends to adhere to soils
and is less likely to leach into groundwater, though it can reach surface waters through erosion or
over-application. A report by the Agricultural Research Service noted that phosphorus bound to
eroded sediment particles makes up 60 to 90 percent of phosphorus transported in surface runoff
from cultivated land (USDA/ARS, 1999). Animal wastes typically have lower nitrogen: phosphorus
2-10
-------
ratios than crop requirements. The application of manure at a nitrogen-based agronomic rate can
therefore result in application of phosphorus at several times the agronomic rate. Soil test data in
the United States confirm that many soils in areas dominated by animal-based agriculture exhibit
excessive levels of phosphorus (Sims, 1995).
Available Nitrogen and Phosphorus
1998 U.S. Department of Agriculture Study
In 1998, the USDA studied the amount of manure nitrogen and phosphorus produced by confined
animals relative to crop uptake potential. USDA evaluated the quantity of nutrients available from
recoverable livestock manure relative to crop growth requirements, by county, based on data from the
1992 Census of Agriculture. The analyses did not consider manure from grazing animals in pasture.
When calculating available nutrients, USDA also corrected for unrecoverable manure, nutrient losses that
occur during storage and treatment, and losses to the environment that can occur through runoff, erosion,
leaching to groundwater, and volatilization (especially for nitrogen in the form of ammonia). Considering
typical management systems, USDA estimates that average manure nitrogen losses range from 31 to 50
percent for poultry, 60 to 70 percent for cattle (including the beef and dairy categories), and 75 percent
for swine. The typical phosphorus loss is 15 percent.
USDA's study examined the potential for available manure nitrogen and phosphorus generated
to meet or exceed plant uptake in each of the 3,141 mainland counties, considering harvested non-legume
cropland and hayland. Based on the analysis of 1992 conditions, available manure nitrogen exceeds crop
system needs in 266 counties, and available manure phosphorus exceeds crop system needs in 485
counties. The relative excess of phosphorus compared to nitrogen is expected because manure is typically
nitrogen-deficient relative to crop needs. Therefore, when manure is applied to meet a crop's nitrogen
requirement, phosphorus is typically over-applied with respect to crop requirements (Sims, 1995).
These analyses do not evaluate environmental transport of applied manure nutrients. Therefore,
an excess of nutrients does not necessarily indicate that a water quality problem exists; likewise, a lack
of excess nutrients does not imply the absence of water quality problems. Nevertheless, the analyses
provide a general indicator of excess nutrients on a broad basis.
Source: Lander etal, 1998.
2-11
-------
2.2.1.3 Eutrophication
Eutrophication is a process in which excess phosphorus or nitrogen over-enriches water
bodies and disrupts aquatic ecosystems. Excess nutrients cause overgrowth of plants, including fast-
growing algae "blooms." Eutrophication can affect the population diversity, abundance, and biomass
of phytoplankton and zooplankton, and can increase the mortality rates of aquatic species (USEPA,
1991). Even when algae are not themselves directly harmful to aquatic life, floating algal mats can
reduce the penetration of sunlight in the water column and limit growth of seagrass beds and other
submerged vegetation. Reduction in submerged aquatic vegetation adversely affects both fish and
shellfish populations, and is the leading cause of biological decline in Chesapeake Bay (Carpenter
et a/., 1998). The 1998 National Water Quality Inventory indicates that excess algal growth alone
is the seventh leading stressor in lakes, ponds, and reservoirs.
Increased algal growth can also raise the pH of water bodies as algae consume dissolved
carbon dioxide to support photosynthesis. This elevated pH can harm the gills of aquatic organisms.
The pH may then drop rapidly at night, when algal photosynthesis stops. In extreme cases, such pH
fluctuations can severely stress aquatic species. In addition, excess nitrogen can contribute to water
quality decline by increasing the acidity of surface waters (USEPA, 1995, 1991).
Damage from eutrophication increases when algae blooms die and are digested by bacteria
in a decomposition process that depletes the level of oxygen in the water. Dissolved oxygen is
necessary for the survival of aquatic life in a healthy ecosystem, and depressed levels of dissolved
oxygen can cause widespread morbidity and mortality among aquatic species. Algal decay and
night-time respiration can lower the dissolved oxygen content of a water body to levels insufficient
to support fish and invertebrates. Severe reductions in dissolved oxygen can result in dramatic fish
kills (Carpenter etal, 1998).
In addition to reducing plant diversity and dissolved oxygen, eutrophication can encourage
the growth of toxic microorganisms such as cyanobacteria (a toxic algae) and the dinoflagellate
Pfiesteriapiscicida. These organisms can be toxic to both wildlife and humans. Researchers have
documented stimulation of Pfiesteria growth by swine effluent spills, and have shown that the
organism's growth can be highly stimulated by both inorganic and organic nitrogen and phosphorus
enrichment (NCSU, 1998).
2.2.2 Pathogens
Pathogens are organisms that cause disease in humans and other species; they include certain
species of bacteria, viruses, protozoa, fungi, and algae. Animal waste itself contains hundreds of
species of microorganisms, including bacteria, viruses, protozoa, and parasites (USDA, 1998;
Jackson etal., 1987; Boyd, 1990). Pathogens may be transmitted directly from manure to surface
water, and pathogens already in surface water may increase in number due to loadings of animal
manure nutrients and organic matter. Of particular concern are certain pathogens associated with
algae blooms. EPA's 1998 National Water Quality Inventory focuses on bacterial pathogens and
2-12
-------
notes that they are the leading stressors in impaired estuaries and the second most prevalent stressors
in impaired rivers and streams.
Over 150 pathogens in livestock manure are associated with risks to humans; these include
the bacteria Escheria coli and Salmonella species, and the protozoa Cryptosporidium parvum and
Giardia species. A recent study by the USDA revealed that about half the cattle at the nation's
feedlots carry E. coli (NAS, 2000). The pathogens C. parvum, Giardia, and E. coli are able to
survive and remain infectious in the environment for long periods of time (Stehman, 2000). In
addition, some bacteria in livestock waste cause avian botulism and avian cholera, which have in the
past killed tens of thousands of migratory waterfowl annually (USEPA, 1993).
Eutrophication is associated
with blooms of a variety of organisms
that can be toxic to fish. This includes
the estuarine dinoflagellate Pfiesteria
piscicida, which is believed to be the
primary cause of many major fish kills
and fish disease events in North
Carolina estuaries and coastal areas, as
well as in Maryland and Virginia
tributaries to the Chesapeake Bay
(NCSU, 1998; USEPA, 1993). In
1997, hog operations were linked to a
Pfiesteria piscicida outbreak in North
Carolina rivers in which 450,000 fish
died (U.S. Senate, 1997). That same
year, poultry operation wastes caused
Pfiesteria outbreaks that killed tens of
thousands offish in Maryland waters,
including the Pokomoke River, King's
Creek, and Chesapeake Bay (Shields,
1997; Shields and Meyer, 1997; New
York Times, 1997).
The generation of toxins
associated with eutrophication can also
threaten other species. In freshwater,
cyanobacterial toxins have caused
many incidents of poisoning of wild
and domestic animals that have
consumed contaminated waters (Health
Canada Environmental Health
Program, 1998; Carpenter ef or/., 1998).
1995 Algae Blooms and Pfiesteria Outbreaks:
Neuse River, North Carolina
Algae blooms andpfiesteria outbreaks on the Neuse
River in North Carolina during the summer and fall of 1995
were the identified causes of three major fish kills and the
suspected causes of several incidents of human illness.
Heavy rains in June of 1995 caused overflows of
wastewater treatment plants and hog lagoons in the
watershed. Within weeks, large mats of algae and aquatic
weeds were reported near the town of New Bern on the
Trent River, a tributary of the Neuse. By July, historically
low levels of dissolved oxygen were recorded in a stretch of
the Neuse downstream from New Bern, coinciding with the
deaths of over 100,000 fish. A second fish kill in August on
another Neuse tributary numbered in the thousands.
In September and October a third major fish kill
occurred along a 35-mile stretch of the Neuse River itself;
the dead fish were covered with sores, and the cause of the
outbreak was determined to be the dinoflagellate pfiesteria.
After multiple reports of similar welts and sores on the
bodies of those who went swimming or fishing in
contaminated areas, state officials declared a health
warning for the area, urging people not to swim, boat, or
fish in the affected area. In addition, the area was closed to
commercial fishing for two weeks.
Source: Leavenworth, 1995a, 1995b.
In coastal waters, visible algae blooms known as red or
brown tides have caused significant mortality in marine mammals. Even when algae blooms are not
2-13
-------
visible, shellfish such as oysters, clams and mussels can carry the toxins from certain algae in their
tissue. Shellfish are filter feeders, and pass large volumes of water over their gills to obtain
nutrients. As a result, they can concentrate a broad range of microorganisms in their tissues, and
provide a pathway for pathogen transmission from surface water to higher trophic organisms (Chai
et al., 1994). Information is becoming available to assess the health effects of contaminated shellfish
on wildlife receptors. In 1998, the death of over 400 California sea lions was linked to ingestion of
mussels contaminated by a bloom of toxic algae (Scholin et a/., 2000). Previous incidents associated
the deaths of manatees and whales with toxic and harmful algae blooms (Anderson, 1998).
In August 1997, the National Oceanic and Atmospheric Administration (NOAA) released
The 1995 National Shellfish Register of Classified Growing Waters. The register characterizes the
status of 4,230 shellfish-growing water areas in 21 coastal states, reflecting an assessment of nearly
25 million acres of estuarine and non-estuarine waters. NOAA found that 3,404 shellfish areas had
some level of impairment. Of these, 110 (3 percent) were impaired to varying degrees by feedlots,
and 280 (8 percent) were impaired by "other agriculture," which could include land where manure
is applied (NOAA, 1997).
2.2.3 Organic Compounds and Biochemical Oxygen Demand (BOD)
Livestock manures contain many carbon-based, biodegradable compounds. Once these
compounds reach surface water, they are decomposed by aquatic bacteria and other microorganisms.
During this process dissolved oxygen is consumed, which in turn reduces the amount of oxygen
available for aquatic animals. EPA's 1998 National Water Quality Inventory indicates that oxygen-
depleting substances are the second leading stressor in estuaries. They are also the fourth leading
stressor both in impaired rivers and streams and in impaired lakes, ponds, and reservoirs.
Carbon compounds and associated biochemical oxygen demand (BOD) can deplete oxygen
and affect the health of aquatic ecosystems in the absence of any other pollutants (e.g., due to
decaying vegetation).13 When carbon compounds enter aquatic ecosystems in conjunction with
nutrients (which is generally the case in manure-related pollution), the impacts of BOD are
compounded by eutrophication and the presence and growth of pathogens. The result is often a rapid
decrease in biodiversity. A study of three Indiana stream systems documents such a reduction in
biodiversity due to AFOs (Hoosier Environmental Council, 1997). The study found that waters
13 Biochemical oxygen demand (BOD) is an indirect measure of the concentration of
biodegradable substances present in an aqueous solution. Anaerobic lagoon effluent from AFOs
typically contains BOD values 10 to 200 times higher than treated domestic sewage. See NCAES,
1982; USD A, 1992; USD A/NRCS, 1992/1996.
2-14
-------
downstream of animal feedlots (mainly hog and dairy operations) contained fewer fish and a limited
number of species offish in comparison with reference sites. It also found excessive algal growth,
altered oxygen content, and increased levels of ammonia, turbidity, pH, and total dissolved solids.
2.2.4 Solids and Siltation
Solids from animal manure include the manure itself and any other elements that have been
mixed with it, such as spilled feed, bedding, hair, feathers, and corpses. Smaller solids with less
weight remain in the water column as "suspended solids" while heavier solids sink to the bottom of
receiving waters in the gradual process of "siltation."
Solids entering surface water can degrade aquatic ecosystems to the point of non-viability.
Suspended particles can reduce the depth to which sunlight can penetrate, decreasing photosynthetic
activity and the resulting oxygen production by plants and phytoplankton. The increased turbidity
also limits the growth of aquatic plants, which serve as critical habitat for fish, crabs, shellfish, and
other aquatic organisms upon which these animals feed. In addition, suspended particles can clog
fish gills, reduce visibility for sight feeders, and disrupt migration by interfering with fish's ability
to navigate using chemical signals (Goldman and Home, 1983; Abt, 1993). EPA's 1998 National
Water Quality Inventory indicates that suspended solids from all sources are the fifth leading stressor
in lakes, ponds, and reservoirs.
A major source of siltation is
erosion from agricultural lands, including Arkansas Water Quality Inventory Report:
AFOs, cropland, and grazing lands Agricultural Activities and Turbidity
(USEPA, 1992b). Silt can contain heavier A1 , inn.,,, ~ ,. T
. . ,, ., ., . . Arkansas 1996 Water Quality Inventory Report
manure particles as well as the soil particles ,. , , ,,. ^ ; . ,
r r discussed a sub-watershed m northwestern Arkansas.
earned by erosion. Such sediment can Landuses in that area, primarily poultry production
smother fish eggs and otherwise interrupt and pasture management, are major sources of
the reproduction of aquatic species (Boyd, nutrients and chronic high turbidity, and water in the
1990). It can also alter or destroy habitat area only partially supports aquatic life.
for benthic organisms. Solids can also
degrade drinking water sources, thereby Source: USEP A, 1993.
increasing treatment costs. The 1998 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
National Water Quality Inventory indicates
that siltation from all sources (including agriculture and non-agriculture) is the leading stressor in
impaired rivers and the third greatest stressor in impaired lakes, ponds, and reservoirs.
2-15
-------
2.2.5 Salts and Trace Elements
Animal manure contains a number of salts and trace elements such as metals. While these
contaminants do not directly alter or interfere with ecosystem processes such as oxygen availability,
they are toxic in high concentrations, both to animals and plants. For example, bottom feeding birds
may be susceptible to metal toxicity because they are attracted to shallow feedlot wastewater ponds
and waters adjacent to feedlots. In addition, metals can remain in aquatic ecosystems for long
periods of time because of adsorption to suspended or bed sediments or uptake by aquatic biota.
The salinity of animal manure is due to the presence of dissolved mineral salts. In particular,
significant concentrations of soluble salts containing sodium and potassium remain from undigested
feed that passes unabsorbed through animals.14 Salinity tends to increase as the volume of manure
decreases during decomposition, and can have an adverse effect on aquatic life and drinking water
supplies (Gresham etal., 1990). Repeated application of manure can lead to increased soil salinity
in the root zone and on top of the soil, where it can damage crops; to reduce salinity farmers apply
excess water, and salts are washed into surface waters in runoff. In fresh waters, increasing salinity
can disrupt the balance of the ecosystem, making it difficult for resident species of plants and
animals to remain. For example, laboratory experiments have linked increased salinity with
inhibited growth and slowed molting in mallard ducklings (USFWS, 1992).
Trace elements in manure can include arsenic, copper, selenium, zinc, cadmium,
molybdenum, nickel, lead, iron, manganese, aluminum, and boron. Of these, arsenic, copper,
selenium, and zinc are often added to animal feed as growth stimulants or biocides (Sims, 1995).
Trace metals may also end up in manure through use of pesticides that are applied to livestock to
suppress houseflies and other pests (USDA/ARS, 1998).
A recent Iowa investigation of chemical and microbial contamination near large scale swine
operations demonstrated the presence of trace elements not only in manure lagoons used to store
swine waste before it is land applied, but also in drainage ditches, agricultural drainage wells, tile
line inlets and outlets, and an adjacent river (CDCP, 1998). Similarly, USFWS has reported on
suspected impacts from a large number of cattle feedlots on Tierra Blanca Creek, upstream of the
Buffalo Lake National Wildlife Refuge in the Texas Panhandle. USFWS found elevated
concentrations of the feed additives copper and zinc in the creek sediment (USFWS, 1991).
14 See Boyd, 1990 and NCAES, 1982. Other major cations contributing to manure salinity
are calcium and magnesium; the major anions are chloride, sulfate, bicarbonate, carbonate, and
nitrate. SeeNRC, 1993.
2-16
-------
2.2.6 Odorous/Volatile Compounds
Sources of volatile compounds and odor from AFOs include animal confinement buildings,
manure piles, waste lagoons, and land application sites, where decomposition of animal wastes by
microorganisms produces gases. The four main gases generated are carbon dioxide, methane,
hydrogen sulfide, and ammonia. Aerobic conditions yield mainly carbon dioxide, while anaerobic
conditions that dominate in typical, unaerated animal waste lagoons generate both methane and
carbon dioxide. Anaerobic conditions are also associated with the generation of hydrogen sulfide
and about 40 other odorous compounds, including volatile fatty acids, phenols, mercaptans,
aromatics, sulfides, and various esters, carbonyls, and amines (USDA, 1992; Bouzaher et a/, 1993).
Volatile compounds affect aquatic ecosystems through atmospheric deposition; ammonia
(discussed in Section 2.2.1.1) is the most important AFO-related volatile because it is itself toxic and
also contributes to eutrophication as a source of nitrogen. Other compounds are less clearly
associated with broad ecological impacts, but may have localized impacts.
2.2.7 Other Pollutants and Ecosystem Imbalances
In addition to the pollutants discussed above, pesticides, antibiotics, and hormones used in
animal feeding operations may exist in animal wastes and may be present in increased levels in the
environment (USDA/ARS, 1998). These compounds may pose risks such as chronic aquatic toxicity
(from pesticides) and reproductive impairment (from hormones). While there is limited information
on the quantities of these compounds that reach surface waters from AFOs, some research suggests
that manure-related runoff may be a significant source of these contaminants.
Pesticides: Pesticides are used to suppress houseflies and other livestock
pests. There is little information on the rate at which pesticides in manure
enter surface water, but a 1999 literature review by the University of
Minnesota notes a 1994 study that links quantities of cyromazine (used to
control flies in poultry litter) in runoff to the rate of manure application and
rainfall intensity. The review also identifies a 1995 study finding that roughly
one percent of all applied pesticides enter surface water. The impacts of
these compounds on aquatic ecosystems are unclear, but there is some
concern that pesticides may contribute to endocrine disruption (Mulla, 1999).
• Hormones: Animal operations use a variety of hormones such as steroids
(e.g., estrogen, progesterone, testosterone) and proteins (e.g., prolactin,
growth hormone) to improve animal health and productivity. Studies have
identified hormones in animal manures. Naturally high hormone
concentrations in birds contribute to higher hormone levels in poultry
manure, including measurable amounts of estrogen and testosterone. When
present in high concentrations, hormones in the environment are linked to
reduced fertility, mutations, and the death offish. There is evidence that fish
2-17
-------
in some streams are experiencing endocrine disruption (Shore et al., 1995;
Mulla, 1999).15
Antibiotics The majority of livestock (roughly 60 to 80 percent) receive
antibiotics during their productive life span. Some of these agents are used
only therapeutically (e.g., to treat illness), but in both the swine and poultry
industries, most antibiotics are administered as feed additives to promote
growth or to improve feed conversion efficiency. Essentially all of an
antibiotic administered is eventually excreted, either unchanged or in
metabolite form (Tetra Tech, 2000). Little information is available regarding
the concentrations of antibiotics in animal wastes, or on the fate and transport
of antibiotics in the environment. However, the key concern related to
antibiotics in animal manure is the potential emergence of antibiotic-resistant
pathogens in surface and drinking water. As antibiotics use has increased,
more strains of antibiotic resistant pathogens are emerging (Mulla, 1999).
Finally, manure pollutants of all types can affect terrestrial as well as aquatic ecosystems.
Over-application of manure, in particular, can have terrestrial effects. High oxygen depletion rates
due to microbial activity have been reported in manure-amended agricultural soils. In addition,
elevated microbial populations can affect crop growth by competing with plant roots for soil oxygen
and nutrients. Trace elements (e.g., feed additives such as arsenic, copper, and selenium) and salts
in animal manure can accumulate in soil and become toxic to plants (USDA, 1992 and USFWS,
1991).
2.3 HUMAN HEALTH IMPACTS RELATED TO AFO POLLUTANTS
Human health impacts from manure-related contaminants are primarily associated with
drinking contaminated water, contact with contaminated water, and consuming contaminated
shellfish. The most common causes of health effects are ingestion of nitrates in drinking water,
ingestion of water containing pathogens from manure, and contact with or ingestion of harmful algae
or toxic algal by-products. The ingestion of elevated concentrations of trace elements (e.g., arsenic,
copper, selenium, and zinc) may also affect human health, and certain gases associated with AFOs
may pose inhalation risks for nearby residents.
15 The presence of estrogen and estrogen-like compounds in surface water has been the focus
of recent research. While their ultimate fate in the environment is unknown, studies indicate that no
common soil or fecal bacteria can metabolize estrogen (Shore et al., 1995). Estradiol, an estrogen
hormone, was found in runoff from a field receiving poultry litter at concentrations up to 3.5
micrograms per litre (ug/L). Fish exposed to 0.25 ug/L of estradiol can undergo gender changes, and
exposures at levels above 10 ug/L can be fatal (Mulla, 1999).
2-18
-------
While some recorded human health effects stem from contamination of public drinking water
supplies and ingestion of shellfish, more frequently health effects are caused by contamination of
private wells, or recreational ingestion or contact. Public water supplies are generally protected by
monitoring and treatment, though contaminants and algae blooms may increase treatment costs and
affect system operation. Ingestion of contaminated shellfish is reduced by monitoring and closure
of shellfish beds in response to excessive levels of contaminants.
2.3.1 Health Impacts Associated with Nitrates
Nitrogen in manure is easily transformed into nitrate form, which can be transported to
drinking water sources (e.g., through leaching to groundwater) and presents a range of health risks.
EPA found that nitrate is the most widespread agricultural contaminant in drinking water wells, and
estimates that 4.5 million people served by wells are exposed to elevated nitrate levels (USEPA,
1990). Elevated nitrate levels can cause nitrate poisoning, particularly in infants (this is known as
methemoglobinemia or "blue baby syndrome"), in which potentially fatal oxygen starvation gives
a "blue" appearance to the skin. In addition to blue baby syndrome, low blood oxygen due to
methemoglobinemia has also been linked to birth defects, miscarriages, and poor health in humans
and animals.16
Reported cases of methemoglobinemia are most often associated with wells that were
privately dug and that may have been badly positioned in relation to the disposal of human and
animal excreta (Addiscott etal., 1991). Reported cases of methemoglobinemia are rare, though the
incidence of actual cases may be greater than the number reported. Studies in South Dakota and
Nebraska have indicated that most cases of methemoglobinemia are not reported. Under-reporting
may result from the fact that methemoglobinemia can be difficult to detect in infants because its
symptoms are similar to other conditions. In addition, doctors are not always required to report it
(Michel, 1996; Meyer, 1994).
In 1995, several private wells in North Carolina were found to be contaminated with nitrates
at levels 10 times higher than the health standard; this contamination was linked with a nearby hog
operation (Warrick 1995c, 1995d). In 1982, nitrate levels greater that 10 milligrams per liter were
found in 32 percent of the wells in Sussex County, Delaware; these levels were associated with local
poultry operations (Chapman, 1996). In southeastern Delaware and the Eastern Shore of Maryland,
where poultry production is prominent, over 20 percent of wells were found to have nitrate levels
16 See USEPA, 1991. In addition, studies in Australia found an increased risk of congenital
malformations with consumption of high-nitrate groundwater. Nitrate- and nitrite-containing
compounds also have the ability to cause hypotension or circulatory collapse. Nitrate metabolites
such as N-nitroso compounds (especially nitrosamines) have been linked to severe human health
effects such as gastric cancer. See Bruning-Fann and Kaneene, 1993.
2-19
-------
exceeding EPA's maximum contaminant level (MCL) (Ritter et a/., 1989). Nitrate is not removed
by conventional drinking water treatment processes. Its removal requires additional, relatively
expensive treatment units.
2.3.2 Health Impacts Associated with Algal Blooms
Eutrophication can affect human health by encouraging the formation of algal blooms. Some
algae release toxins as they die and may affect human health through dermal contact or through
consumption of contaminated water or shellfish. In marine ecosystems, algal blooms such as red
tides form toxic byproducts that can affect human health through recreational contact or consumption
of contaminated shellfish (Thomann and Muller, 1987). In freshwater, blooms of cyanobacteria
(blue-green algae) may pose a serious health hazard to those who consume the water. When
cyanobacterial blooms die or are ingested, they release water-soluble compounds that are toxic to
the nervous system and liver (Carpenter et a/., 1998).
Non-toxic algae blooms triggered by nutrient pollution can also affect drinking water by
clogging treatment plant intakes and by producing objectionable tastes and odors. In addition,
increased algae in drinking water sources can increase production of harmful chlorinated byproducts
(e.g., trihalomethanes) by reacting with chlorine used to disinfect drinking water.
Impacts of Manure Pollutants on Water Treatment Costs
Public water providers may incur considerable expenses associated with removing manure-related
contaminants and algae from public water supplies. For example:
• In California's Chino Basin, it could cost over $1 million per year to remove the nitrates from
drinking water due to loadings from local dairies.
• In Wisconsin, the City ofOshkosh has spent an extra $30,000 per year on copper sulfate to kill
the algae in the water it draws from Lake Winnebago. The thick mats of algae in the lake have
been attributed to excess nutrients from manure, commercial fertilizers, and soil.
• In Tulsa, Oklahoma, excessive algal growth in Lake Eucha is associated with poultry farming.
The city spends $100,000 per year to address taste and odor problems in the drinking water.
Sources: For more details on these examples, see USEPA, 1993; Behm, 1989; Lassek, 1998; and
Lassek, 1997.
2-20
-------
2.3.3 Health Impacts Associated with Pathogens
Over 150 pathogens in livestock manure are associated with risks to humans (Juranek, 1991;
CAST, 1992). Although human contact can occur through contaminated drinking water, adequate
treatment of public water supplies generally prevents exposure. Most exposure occurs through
incidental ingestion during recreation in contaminated waters or through ingestion of contaminated
shellfish (Stelma and McCabe, 1992). Relatively few microbial agents are responsible for the
majority of human disease outbreaks from water-based exposure routes. Intestinal infections are the
most common type of waterborne infection, but contact recreation with pathogens can also result in
infections of the skin, eye, ear, nose, and throat (Juranek, 1995; and Stehman, 2000). In 1989, ear
and skin infections and intestinal illnesses were reported in swimmers as a result of discharges from
a dairy operation in Wisconsin (Behm, 1989).
A study for the period 1989 to 1996 revealed that Cryptosporidium parvum (a pathogen
associated with cows) was one of the leading causes of infectious water-borne disease outbreaks in
which an agent was identified. C. parvum can produce gastrointestinal illnesses such as
cryptosporidiosis, with symptoms that include severe diarrhea (Stehman, 2000). While otherwise
healthy people typically recover quickly from illnesses such as cryptosporidiosis, these diseases can
be fatal in certain subpopulations, including children, the elderly, people with HIV infection,
chemotherapy patients, and those taking medications that suppress the immune system.17 In
Milwaukee, Wisconsin in 1993, C. parvum contamination of a public water supply caused more than
100 deaths and an estimated 403,000 illnesses. The source was not identified, but speculated sources
include runoff from cow manure application sites (Gasman, 1996). More recently, a May, 2000
outbreak of Escherichia coli O157:H7 in Walkerton, Ontario resulted in at least seven deaths and
1,000 cases of intestinal problems; public health officials theorize that flood waters washed manure
contaminated with E. coli into the town's drinking water well (Brooke, 2000).
Algae blooms are associated with a variety of organisms that are toxic to humans, including
the algae associated with "red tide" and a number dinoflagellates. One pathogen of particular
concern is the estuarine dinoflagellate Pfiesteria piscicida. While Pfiesteria is primarily associated
with fish kills and fish disease events, the organism has also been linked with human health impacts
through dermal or inhalation exposure. Researchers working with dilute toxic cultures of Pfiesteria
have exhibited symptoms such as skin sores, severe headaches, blurred vision, nausea/vomiting,
sustained difficulty breathing, kidney and liver dysfunction, acute short-term memory loss, and
severe cognitive impairment. In addition, people with heavy environmental exposure have exhibited
symptoms as well. In a 1998 study, such environmental exposure was definitively linked with
cognitive impairment, and less consistently linked with physical symptoms (NCSU, 1998; Morris
etal., 1998).
17 By the year 2010, about 20% of the human population (especially infants, the elderly, and
those with compromised immune systems) will be classified as particularly vulnerable to the health
effects of pathogens (Mulla, 1999).
2-21
-------
While many soil types prevent most pathogens from reaching aquifers, groundwater in areas
of sandy soils, limestone formations, or sinkholes is more vulnerable to contamination. Private
wells, in particular, are prone to contamination because they tend to be shallower than public wells
and therefore more susceptible to contaminants leaching from the surface.18 While the general extent
of groundwater contamination from AFOs is unknown, there are incidents that indicate a connection
between livestock waste and contaminated well water. For example, in cow pasture areas of Door
County, Wisconsin, where a thin topsoil layer is underlain by fractured limestone bedrock,
groundwater wells have commonly been shut down due to high bacteria levels (Behm, 1989).
2.3.4 Health Impacts Associated with Trace Elements and Salts
Trace elements in manure include feed additives such as zinc, arsenic, copper, and selenium.
While these are necessary nutrients, they are toxic at elevated concentrations, and tend to persist in
the environment and to bioconcentrate in plant and animal tissues. Trace elements are associated
with a variety of illnesses. For example, over exposure to selenium can cause liver dysfunction and
loss of hair and nails, while ingestion of too much zinc can produce changes in copper and iron
balances, particularly copper deficiency anemia (IRIS, 2000).
Total concentrations of trace elements in animal manures have been reported as comparable
to those in some municipal sludges, with typical values well below the maximum concentrations that
EPA allows in land-applied sewage sludge (Sims, 1995). Based on this information, trace elements
in agronomically applied manures should pose little risk to human health and the environment.
However, repeated application of manures above agronomic rates could result in exceedances of the
cumulative metal loading rates that EPA considers safe, potentially affecting human health and the
environment. There is some evidence that this is happening. For example, in 1995, zinc and copper
were found building to potentially harmful levels on the fields of a North Carolina hog farm
(Warrick and Stith, 1995b).
Salts in manure can also affect the salinity of drinking water. Increased salts in drinking
water can in turn increase blood pressure in salt-sensitive individuals, increasing the risk of stroke
and heart attack (Anderson, 1998; Boyd, 1990).
2.3.5 Other Health Impacts
Potential health effects associated with other contaminants in manure include inhalation-
related risks associated with volatile organic chemicals and odors, and the effects of hormones,
antibiotics, and pesticides that are found in animal feed.
18 In a 1997 survey of drinking water standard violations in six states over a four-year period,
the U.S. General Accounting Office reported that bacterial standard violations occurred in up to 6
percent of community water systems each year and in up to 42 percent of private wells. See
USGAO, 1997.
2-22
-------
Volatile Compounds
In 1996, the Minnesota Department of Health found levels of hydrogen sulfide gas at
residences near AFOs that were high enough to cause symptoms such as headaches, nausea,
vomiting, eye irritation, respiratory problems (including shallow breathing and coughing), achy
joints, dizziness, fatigue, sore throats, swollen glands, tightness in the chest, irritability, insomnia,
and blackouts (Hoosier Environmental Council, 1997). In an Iowa study, neighbors within two miles
of a 4,000-sow swine facility reported more physical and mental health symptoms than a control
group (Thu, 1998). These symptoms included chronic bronchitis, hyperactive airways, mucus
membrane irritation, headache, nausea, tension, anger, fatigue, and confusion. Odor is itself a
significant concern because of its documented effect on moods, such as increased tension,
depression, and fatigue (Schiffman et a/., 1995). Heavy odors are the most common complaint from
neighbors of swine operations (Agricultural Animal Waste Task Force, 1996).
Pesticides
Various ingredients in pesticides have been linked to a variety of human health effects, such
as systemic toxicity and endocrine disruption (see below). However, information linking pesticide
levels in surface and drinking water to human exposure and to animal manure is currently limited.
It is therefore unclear what health risks are posed by pesticide concentrations in AFO wastes.
Hormones and Endocrine Disruption
Hormones in the environment can act as endocrine disrupters, altering hormone pathways
that regulate reproductive processes in both human and animal populations. Estrogen hormones
have been implicated in the drastic reduction in sperm counts among European and North American
men (Sharpe and Skakkebaek, 1993) and widespread reproductive disorders in a variety of wildlife
(Colburn et al., 1993). A number of agricultural chemicals have also been demonstrated to cause
endocrine disruption as well, including pesticides (Shore et al, 1995). The effects of these chemicals
on the environment and their impacts on human health through environmental exposures are not
completely understood, but they are currently being studied for evidence that they cause
neurobiological, developmental, reproductive, and carcinogenic effects (Tetra Tech, 2000). No
studies exist on the human health impact of hormones from manure watersheds.
Antibiotics and Antibiotic Resistance
While antibiotics themselves are not generally associated with human health impacts,
antibiotic resistance poses a significant health threat. In April 2000, the New England Journal of
Medicine published an article that discussed the case of a 12-year old boy infected with a strain of
Salmonella that was resistant to no fewer than 13 antimicrobial agents (Fey, 2000). The cause of the
child's illness is believed to be exposure to the cattle on his family's Nebraska ranch. The Centers
2-23
-------
for Disease Control, the Food and Drug Administration, and the National Institutes of Health issued
a draft action plan in June, 2000, to address the increase in antibiotic resistant diseases (CDCP,
2000). The plan is intended to combat antimicrobial resistance through surveys, prevention and
control activities, research, and product development. Some actions are already underway.
2.4 REFERENCES
Abt Associates Inc. 1993. Human Health Risk Assessment for the Use and Disposal of Sewage
Sludge: Benefits of Regulation. Prepared for Health and Ecological Criteria Division, Office
of Science and Technology, Office of Water, U.S. Environmental Protection Agency.
January.
Addiscott, T.M., A.P. Whitmore, and D.S. Powlson. 1991. Farming, Fertilizers, and the Nitrate
Problem. Rothamsted Experimental Station. C-A-B International: Oxon, United Kingdom.
Anderson, Don. 1998. "Why are outbreaks of Pfiesteria and red tides suddenly threatening our
oceans?" Scientific American online at: http://www.sciam.com/askexpert/environment/
environment2 l/environment21 .html
Aneja, Viney, George C. Murray, and James Southerland. April 1998. Atmospheric Nitrogen
Compounds: Emissions, Transport, Transformation, Deposition, and Assessment. EM, Air
& Waste Management Association's Magazine for Environmental Managers, 22-25.
Available at: online.awma.org/em/april98/features/aneja/aneja.htm
ASCE (American Society of Civil Engineers). February 1998. Bay Pollution Flows Underground.
Civil Engineering, p. 13-14.
Behm, Donald. 1989. "Ill Waters: The Fouling of Wisconsin's Lakes and Streams." The
Milwaukee Journal Sentinel. Special Report: A series of articles and an editorial published
Novembers - 10, 1989.
Bouzaher, A., P.G. Lakshminarayan, S.R. Johnson, T. Jones, and R. Jones. 1993. The Economic
and Environmental Indicators for Evaluating the National Pilot Project on Livestock and the
Environment, Livestock Series Report 1. Center for Agricultural and Rural Development
(CARD) at Iowa State University and Texas Institute for Applied Environmental Research
at Tarleton State University. Staff Report 93-SR 64. October.
Boyd, W. 1990. Effects of Poultry Waste on Ground and Surface Water. In Blake, J., and R.M.
Hulet, eds. Proceedings of the 1990 National Poultry Waste Management Symposium.
October.
Brooke, James. 2000. Few left untouched after deadly E. coli flows through an Ontario town's
water. The New York Times. July 10, 2000.
2-24
-------
Bruning-Fann, C. S. and J. B. Kaneene. 1993. The Effects of Nitrate, Nitrite, and N-Nitroso
Compounds on Human Health: A Review. Vet. Human. Toxicol. 35(6). December.
Carpenter, Stephen, N.F. Caraco, D.L. Correll, R.W. Howarth, A.N. Sharpley, and V.H. Smith.
1998. Nonpoint Pollution of Surface Waters with Phosphorus and Nitrogen. Issues in
Ecology, Number 3. Published by the Ecological Society of America, Washington, D.C.
Summer. (Also available at: esa.sdsc.edu/).
Gasman, Elizabeth A. 1996. Chemical and Microbiological Consequences of Anaerobic Digestion
of Livestock Manure, A Literature Review. Interstate Commission on the Potomac River
Basin, ICPRB Report #96-6.
CAST (Council for Agricultural Science and Technology). 1992. Water Quality: Agriculture's
Role. Report 120. December.
CDCP (Centers for Disease Control and Prevention). 1998. Report to the State of Iowa Department
of Public Health on the Investigation of the Chemical andMicrobial Constituents of Ground
and Surface water Proximal to Large-Scale Swine Operations.
CDCP (Centers for Disease Control and Prevention). 2000. A Public Health Action Plan to Combat
Antimicrobial Resistance. June Draft, available at:
http://www.cdc.gov/drugresistance/actionplan/aractionplan.pdf
Chai, T.J., Tj. Han and R.R. Cockey. 1994. Microbiological quality of shellfish-growing waters in
Chesapeake Bay. J. Food Protect. 57:229-234.
Chapman, S.L. 1996. Soil and Solid Poultry Waste Nutrient Management and Water Quality.
Poultry Science 75:862-866.
CWAA (Clean Water Action Alliance). 1998. Minnesota Manure Spills and Runoff.
Daniel, T.C., D.R. Edwards, and DJ. Nichols. 1995. Edge-of-Field Losses of Surf ace-Applied
Animal Manure. In: Animal Waste and the Land-Water Interface, edited by Kenneth Steele.
Fey, Paul D.; Safranek, Thomas J.; Rupp, MarkE.; Dunne, Eileen F.; Ribot, Efrain; Iwen, Peter C.;
Bradford, Patricia A.; Angulo, Frederick J.; Hinrichs, Steven H. Ceftriaxone. 2000.
"Resistant Salmonella Infection Acquired by a Child from Cattle." The New England
Journal of Medicine; 342 (17), pp 1242-1249.
Follett, Ronald F. 1995. Fate and Transport of Nutrients: Nitrogen. Working Paper No. 7. U.S.
Department of Agriculture, Agricultural Research Service, Soil-Plant-Nutrient Research
Unit, Fort Collins, Colorado. September.
Goldman, C., and A. Home. 1983. Limnology. New York: McGraw-Hill Publishing, Co.
2-25
-------
Gresham, C.W., et al. 1990. Composting of Poultry Litter, Leaves, and Newspaper. Rodale
Research Center.
Health Canada Environmental Health Program. 1998. Blue-green Algae (Cyanobacteria) and their
Toxins. Publication of the Canadian federal government. Available on the Internet at
http: //www. he- sc. gc. ca/ehp/ehd/catal ogue/general/iy h/algea. htm.
Hoosier Environmental Council. 1997. Internet home page. Available at: www.envirolink.org/
orgs/hecweb/monitorspring97/confmed.htm.
IRIS (Integrated Risk Information System). 2000. Chemical Files and Background Documents and
Papers. Washington, DC. Available at: www.epa.gov/iris/subst/index.html. March.
Jackson, G., et al. 1987. Agricultural Management Practices to Minimize Groundwater
Contamination. University of Wisconsin Environmental Resources Center (July).
Jordan, Herbert C., and Robert E. Graves. 1987. Poultry Manure Management, Pennsylvania State
University. Prepared under the direction of the Manure Management Work Group of the
Agricultural Advisory Committee to the PA Department of Environmental Resources.
Juranek, D.D. 1995. Cryptosporidiosis: sources of infection and guidelines for prevention. Clin
InfectDis. 21 (Sup. 1): S57-61.
Lander, Charles H., David Moffitt, and Klaus Alt. 1998. Nutrients Available from Livestock
Manure Relative to Crop Growth Requirements. U.S. Department of Agriculture. Natural
Resources Conservation Service. Washington, DC. Available at: www.nhq.nrcs.usda.gov/
1 and/pub s/nl web. html.
Lassek, P.J. 1997. Lake Eucha drowning in algae. The Tulsa World. August 17.
Lassek, P.J. 1998. Tulsa Officials Seek Long-Term Solution to Protect Watershed. The Tulsa
World, March 16.
Leavenworth, Stuart. 1995a. "Airborne threats rain down on Neuse." The New sand Observer May
8. Available at: www.nando.net.
Leavenworth, Stuart. 1995b. "Half-million fish die in Neuse River." The News and Observer
September 22. Available at: www.nando.net.
Meadows, R. 1995. Livestock Legacy. Environmental Health Perspectives 103(12): 1096-1100.
Meyer, M. October 1994. How Common isMethemoglobinemia from Nitrate Contaminated Wells?
A South Dakota Perspective. Paper presented at the 39th Annual Midwest Ground Water
Conference, Bismarck, ND. Cited in: Michel et al., 1996.
2-26
-------
Michel, Kristin, J.R. Bacon, C.M. Gempesaw II, and J.H. Martin, Jr. August 1996. Nutrient
Management by Delmarva Poultry Growers: A Survey of Attitudes and Practices. University
of Delaware, College of Agricultural Sciences, Department of Food and Resource
Economics.
Morris, J. Glenn Jr., et al. 1998. University of Maryland School of Medicine, Baltimore, Maryland.
As reported by David Brown: "Pfiesteria Linked to Thinking Problems," The Washington
Post, August 14.
Mulla, David; A. Sekely, A. Birr, J. Perry, B. Vondracek, E. Bean, E. Macbeth, S. Goyal, B.
Wheeler, C. Alexander, G. Randall, G. Sands, and J. Linn. 1999. Generic Environmental
Impact Statement on Animal Agriculture: A Summary of the Literature Related to the Effects
of Animal Agriculture on Water Resources (G). Prepared for the Environmental Quality
Board by the University of Minnesota, College of Agriculture, Food, and Environmental
Sciences.
NAS (National Academy of Sciences), 2000. Proceedings of the National Academy of Sciences
April.
NCAES (North Carolina Agricultural Extension Service). 1982. Best Management Practices for
Agricultural Nonpoint Source Control: Animal Waste. North Carolina State University.
Biological and Agricultural Engineering Department. Raleigh, NC.
NCSU (North Carolina State University). 1998. Aquatic Botany Laboratory Pfiesteriapiscicida
homepage. Available at: www.pfiesteria.org/pfiestr/pfiestr.html.
New York Times (editorial). 1997. Why the Fish Are Dying. The New York Times September 22.
Available at: www.nytimes.com.
NOAA (National Oceanic and Atmospheric Administration). August 1997. The 1995 National
Shellfish Register of Classified Growing Waters. Office of Ocean Resources Conservation
and Assessment. National Ocean Service, U.S. Department of Commerce.
NRC (National Research Council). 1993. Soil and Water Quality: An Agenda for Agriculture.
National Academy Press: Washington, D.C.
ODNR (Ohio Department of Natural Resources). 1997. Division of Wildlife Pollution Investigation
Report - Manure Related Spills and Fish and Wildlife Kills. June 9.
Phillips, J. M., H. D. Scott, and D. C. Wolf. 1992. Environmental Implications of Animal Waste
Application to Pastures. In: Proceedings from the South Pasture Forage Crop Improvement
Conference. USD A/Agricultural Research Service (September), pp. 30-38.
2-27
-------
Puckett, LJ. 1994. Nonpoint and point sources of nitrogen in major watersheds of the United
States. U.S. Geological Survey, National Water Quality Assessment.
Ritter, W.F., A.E.M. Chirnside, andR.W. Lake. 1989. Influence of Best Management Practices on
Water Quality in the Appoquinimink Watershed. J. Environ. Sci. Health., A24(8):897-924.
Ritter, W. F., and A. E. M. Chirnside. 1990. Impact of Animal Waste Lagoons on Groundwater
Quality. Biological Wastes. 34:39-54.
Schiffman, S.S., E.A. Sattely Miller, M.S. Suggs, and E.G. Graham. 1995. The Effect of
Environmental Odors Emanating from Commercial Swine Operations on the Mood of
Nearby Residents. Brain Research Bulletin, 37:369-375.
Scholin, Christopher A., et al.. 2000. "Mortality of sea lions along the central California coast
linked to a toxic diatom bloom." Nature 403, 80 - 84.
Shields, T. 1997. Fish Kills Seen as 'Alarm Bell' For Chesapeake, Tributaries. The Washington Post
17 August. Bl.
Shields, T. and E.L. Meyer. 1997. New Fish Kills Found Miles from Pocomoke. The Washington
Post., September 11. Al.
Sims, J. Thomas. 1995. "Characteristics of Animal Wastes and Waste-Amended Soils: An
Overview of the Agricultural and Environmental Issues" in Animal Waste and the Land-
Water Interface. Kenneth Steele, ed., CRC Press/Lewis Publishers, Boca Raton.
Smith, R., G. Schwarz, and R. Alexander. 1997. Regional Interpretation of Water-Quality
Monitoring Data. Water Resources Research. Vol. 33(12): 2781-2798. December.
Stehman, Susan M. 2000. Ag-Related water borne pathogens. In: Managing Nutrients and
Pathogens from Animal Agriculture, Proceedings of a Conference for Nutrient Management
Consultants, Extension Educators, and Producer Advisors. Natural Resource, Agriculture,
and Engineering Service (NRAES). pp. 93-107. March 28-30.
Tetra Tech. 2000. Literature Review and Assessment of Pathogens, Heavy Metals, and Antibiotic
Content of Waste and Wastewater Generated by CAFOS. Unpublished report prepared at
the request of U.S. Environmental Protection Agency, Office of Water, EPA contract 68-C-
99-263.
Thomann , R. V., and J.A. Muller. 1987. Principles of Surface Water Quality Modeling and
Control. Harper Collins Publishers.
Thu, Kendall. 1998. Information presented at the Center for Disease Control Workshop on Public
Health Issues Related to Concentrated Animal Feeding Operations, June 23-24.
2-28
-------
USDA (U.S. Department of Agriculture). 1992. Agricultural Waste Management Field Handbook.
210-AWMFH. April.
USDA (U.S. Department of Agriculture). 1998. Agricultural Uses of Municipal Animal, and
Industrial Byproducts. Agricultural Research Service. January.
USDA/ARS (U.S. Department of Agriculture/Agricultural Research Service). 1998. Agricultural
Uses of Municipal, Animal, and Industrial Byproducts. Conservation Research Report
Number 44. January.
USDA/NRCS (U.S. Department of Agriculture/Natural Resources Conservation Service). 1992 and
1996 revisions. Agricultural Waste Management Field Handbook. 210-AWMFH.
USEPA (U.S. Environmental Protection Agency). 1990. National Survey of Pesticides in Drinking
Water Wells, Phase I Report. EPA-570990015.
USEPA (U. S. Environmental Protection Agency). 1991. Nitrogen Action Plan. Draft (March).
PM-221. 400/90/003. Nitrogen Work Group.
USEPA (U. S. Environmental Protection Agency). 1992b. ManagingNonpoint Source Pollution:
Final Report to Congress. Office of Water.
USEPA (U. S. Environmental Protection Agency). 1993. The Report of the EPA/State Feedlot
Workgroup. Office of Wastewater Enforcement and Compliance.
USEPA (U. S. Environmental Protection Agency). 1995. National Water Quality Inventory: 1994
Report to Congress. Office of Water. EPA 841-R-95-005. December.
USEPA (U.S. Environmental Protection Agency). 1997. Section 319 Success Stories: Volume II.
Highlights of State and Tribal Nonpoint Source Programs (Online). Office of Water.
Washington, D.C. EPA-R-97-001. October. Available at: www.epa.gov/owow/NPS/
Sections 1911.
USEPA (U.S. Environmental Protection Agency). 2000. National Water Quality Inventory: 1998
Report to Congress. EPA 841-R-00-001.
USFWS (U.S. Fish and Wildlife Service). 1991. Contaminants in Buffalo Lake National Wildlife
Refuge, Texas. Report by the Arlington Field Office. October.
USFWS (U.S. Fish and Wildlife Service). 1992. "An Overview of Irrigation Drainwater
Techniques, Impacts on Fish and Wildlife Resources, and Management Options" prepared
for USEPA, OPPE by U.S. Fish and Wildlife Service May, 1992. Cited in: Industrial
Economics, Incorporated, 1993. "Irrigation Return Flow Fee Feasibility Study," unpublished
memorandum prepared for USEPA, Office of Policy Analysis, October.
2-29
-------
USGAO (U.S. General Accounting Office). 1997. Drinking Water: Information on the Quality of
Water Found at Community Water Systems and Private Wells. Report to Congressional
Requesters. GAO/RCED-97-123. Available at: www.gao.gov/AIndexFY97/abstracts/
rc97123.htm. June.
U.S. Senate. 1997. Animal Waste Pollution in America: An emerging national problem. Report
Compiled by the Minority Staff of the United States Senate Committee on Agriculture,
Nutrition and Forestry for Senator Tom Harkin.
Vanderholm, D.H. 1975. Nutrient Losses from Livestock Waste During Storage, Treatment, and
Handling. Managing Livestock Wastes, p. 282.
Warrick, Joby. 1995b. Hog farm is fined $110,000 for spill. The News and Observer. August 23,
1995. Available at: www.nando.net.
Warrick, Joby. 1995c. Contaminated wells linked to Robeson hog operation. The News and
Observer. October 14, 1995. Available at: www.nando.net.
Warrick, Joby. 1995d. Tainted wells, poisoned relations The New sand Observer. October 15, 1995.
Available at: www.nando.net.
Warrick, Joby and Pat Stith. 1995a. "Boss Hog: North Carolina's Pork Revolution." The Raleigh,
NC News and Observer. Five part series published from February 19-26. Available at:
www.nando.net/sproject/hogs/hoghome.html
Warrick, Joby and Pat Stith. 1995b. Metals present new concern for livestock producers. The News
and Observer. September 17. Available at: www.nando.net.
Wetzel, R.G. 1983. Limnology. 2nd Edition. Saunders College Publishing.
2-30
-------
CONCEPTUAL FRAMEWORK AND OVERVIEW OF METHODS CHAPTER 3
Pollutants associated with AFOs can have a range of harmful impacts on water quality, on
aquatic and shoreline ecosystems, and on the range of uses (or services) that water resources provide.
While some pollutants pose a direct threat to human health (e.g., pathogens that prevent drinking
or contact with contaminated water), AFO-related pollutants can also contribute to the decline of
recreational and commercial activities, injury to species that live in or depend on contaminated
waters (e.g., aquatic shorebirds), and even a reduction in the intrinsic "existence" value that people
place on a pristine or well-protected ecosystem.
The benefits of a regulation that reduces AFO pollution are reflected by identifiable changes
in environmental quality that result from the regulation, and by the related improvements in the range
of potential uses of the resource. The value of the regulation is then measured according to the value
that people place on the changes in these potential uses. EPA characterizes these changes by
considering the use and non-use benefits that water resources provide under baseline conditions, and
contrasting these benefits with the enhanced benefits realized under each of the regulatory scenarios.
This chapter describes the general approach that EPA uses to value environmental quality
improvements associated with reduced AFO pollution. The first section describes the types of
environmental improvements and benefits to humans that would likely result from changes in water
quality due to the regulation of CAFOs. The chapter then identifies the key environmental changes
and benefits that are the focus of the evaluation of EPA's proposed regulations, and describes EPA's
approaches to measuring and valuing the selected benefits. The broad methods outlined in this
chapter form the basis of the specific benefits analyses described in Chapters 4 through 7.
3.1 POSSIBLE ENVIRONMENTAL IMPROVEMENTS
AND RESULTING BENEFITS
Groundwater and surface water resources (including rivers, lakes, estuaries, and oceans)
provide a range of benefits to humans and other species that reflect the actual and potential "uses"
that they support. Potential uses can include active consumption or diversion of water for industry,
agriculture, or drinking water, and can also include a range of active and passive "in-place" uses such
as swimming, fishing, and aesthetic enjoyment.
3-1
-------
Water resources also provide intrinsic (or non-use) benefits that reflect the importance of
protecting environmental quality regardless of any specific use that humans may enjoy or intend.
Intrinsic benefits include "existence value," i.e., the sense of well-being that people derive from the
existence of pristine water resources, even when they do not expect to see or use these resources.19
The protection of resources for future generations (intergenerational equity) or for non-human
species (ecological benefits) are other key intrinsic benefits.
Degradation of a water resource may restrict its use or the intrinsic benefits it provides, and
therefore reduce its value. Conversely, improvement in environmental quality provides benefits
associated with an increase in the range of potential uses and intrinsic benefits that a resource can
support. Exhibit 3-1 provides a summary of the potential benefits associated with an improvement
in the quality of aquatic resources.
Exhibit 3-1
POTENTIAL BENEFITS OF WATER QUALITY IMPROVEMENTS
Use Benefits
In-Stream
Near Stream
Option Value
Diversionary
Aesthetic
Commercial fisheries, shell fisheries, and aquaculture; navigation
• Recreation (fishing, boating, swimming, etc.)
Subsistence fishing
• Human health risk reductions
Water-enhanced non-contact recreation (picnicking, photography, jogging, camping, etc.)
• Nonconsumptive use (e.g., wildlife observation)
• Premium for uncertain future demand
• Premium for uncertain future supply
Industry /commercial (process and cooling waters)
Agriculture/irrigation
• Municipal/private drinking water (treatment cost savings and/or human health risk
reductions)
• Residing, working, traveling and/or owning property near water, etc.
Intrinsic (Non-Use) Benefits
Bequest
Existence
Ecological
Intergenerational equity
Stewardship/preservation
Vicarious consumption
• Reduced mortality /morbidity for aquatic and other species
Improved reproductive success for aquatic and other species
Increased diversity of aquatic and other species
Improved habitat, etc.
19 A common example of intrinsic value is the broad public support for the preservation of
National Parks, even by people who do not expect to visit them.
3-2
-------
AFO pollutants have impacts on a broad range of water resource services. Pollution by
nutrients, for example, can reduce the value of both groundwater and surface water as a drinking
water source, and algae in eutrophied surface water can reduce recreational and aesthetic uses (due
to foul odor and appearance), as well as clog municipal and industrial intakes. Acute nitrogen
loadings and decaying algae cause fish kills, which affect commercial and recreational fishing, and
indicate injury to natural resources; some of these injuries may require restoration in order to achieve
full recovery of the ecosystem. Both chronic and acute nutrient loadings can reduce aquatic
populations and the shoreline species that depend on them; this affects both opportunities to view
wildlife and ecological "existence" values. Finally, nutrient-related red tide and Pfiesteria events
can restrict access to shellfish and beaches, affecting shellfishing and recreational opportunities.
Other AFO pollutants have similar impacts or can cause additional effects (e.g., turbidity
from solids, human health effects from pathogens). In addition, any pollutant that reduces the quality
of an environmental resource may adversely affect intrinsic values, such as bequest values (i.e.,
preserving environmental quality for future generations). While the beneficial impacts of improved
control of any one pollutant can be difficult to isolate, AFO-related pollution generally involves a
broad range of impacts that, taken together, affect to some degree most of the potential uses and
intrinsic benefits of water resources.
3.2 SPECIFIC BENEFITS ANALYZED
The benefits of water quality improvements are a function of the specific pollutants reduced,
the water resources affected, and the improvements in the potential uses of these resources. The key
challenge of a benefits calculation is to establish a clear link between the implementation of a
regulation, the reduction of a pollutant, the resulting improvement in environmental quality, and the
value of that improvement.
While AFO-related pollutants can affect most potential uses of surface and groundwater,
EPA has identified a set of environmental quality changes that meet three criteria: 1) they represent
identifiable and measurable changes in water quality; 2) they can be linked with the proposed CAFO
regulations; and 3) together, they represent a broad range of potential human uses and benefits and
are likely to capture important environmental changes that result from the rule. Specifically, EPA
implements four analyses:
• Improvements in Water Quality and Suitability for Recreational
Activities: this analysis addresses increased opportunities for recreational
fishing and swimming, as well as the potential increase in non-use values
associated with improvements in inland surface water quality;
Reduced Incidence of Fish Kills: this analysis assesses the value of
reducing the incidence offish kills attributable to pollution from AFOs;
3-3
-------
Improved Commercial Shell Fishing: this analysis characterizes the impact
of pollution from AFOs on access to commercial shellfish growing waters,
and values the potential increase in commercial shellfish harvests that may
result from improved control of that pollution; and
Reduced Contamination of Private Wells: this analysis values the impact
of the revised regulations in reducing the concentration of nitrates in water
drawn from private wells.
EPA's analysis does not attempt to comprehensively identify and value all potential
environmental changes associated with proposed revisions to the CAFO regulations. For example,
the analysis of the suitability of water resources for recreational use excludes the Great Lakes, as well
as all estuarine or marine waters. In addition, the analysis examines only a limited set of potential
improvements in water quality; it does not attempt to value the improvement of non-boatable waters
to boatable condition, nor does it attempt to value improvements in the quality of water resources
that are already suitable for swimming. Furthermore, the analysis does not value the potential impact
of improvements in water quality on near-stream activities, such as birdwatching or camping, nor
does it consider non-water related benefits, such as potential reductions in odor from waste
management areas. EPA also does not evaluate potential impacts on certain diversionary uses (e.g.,
improvements in the quality of reservoirs and other sources of public water supplies, and the
associated reduction in the cost of treating water to remove AFO-related pollutants).
While changes in water quality resulting from CAFO regulations may have real impacts on
these types of uses, and may even be associated with significant benefits, several factors make it
difficult to measure the specific impacts of the regulation and identify related changes in value. For
example, analysis of potential changes in estuarine or marine water quality nationwide is currently
beyond the capabilities of the water quality model employed in this study. In addition, while EPA's
proposed CAFO regulations will contribute to improvements in environmental quality beyond
surface waters, it is difficult to establish clear relationships between regulation of AFOs and certain
environmental quality changes, such as reductions in odor or improvements in the health of
shorebirds. Although these benefits are not specifically addressed by the analysis, they likely
represent additional benefits of the regulation.
3.3 PREDICTING CHANGE IN ENVIRONMENTAL
QUALITY AND RESULTING BENEFICIAL USE
To calculate the benefits associated with a proposed regulations, an analysis must explore
the difference between present conditions (i.e., the baseline scenario) and the likely future conditions
that would result from the regulation. The baseline scenario is typically assessed using the best and
most recently collected data that characterize existing environmental quality. Because likely future
conditions are theoretical, the characterization of environmental quality for each of the regulatory
scenarios must be evaluated through environmental modeling or other approaches designed to
3-4
-------
simulate possible future conditions. The difference in environmental quality between the present
and future conditions thus represents the marginal environmental quality gains or human benefits
that would be produced under each scenario.
EPA's analysis of the proposed CAFO regulations examines the difference between the
baseline and each of the regulatory scenarios for each of the four benefits categories identified above.
Ideally, the baseline scenarios would be constant across benefit categories and analyses; however,
data limitations forced EPA to define baseline conditions based on the most up to date record of
existing conditions for each analysis. For instance, the analysis of increased commercial shellfish
supply benefits relies upon 1995 data on shellfish bed closures to define baseline conditions, whereas
the analysis of fish kill events relies upon data collected between 1980 and 1999. Detailed
information on the time frame used to define baseline scenarios for each of the selected
environmental benefit categories is provided for each of the analyses addressed in Chapters 4
through 7.
For each of the benefit categories analyzed, post-regulatory conditions are assessed using
modeling approaches most applicable to the specific analysis. For each of the selected benefit
categories, EPA models post-regulatory conditions as follows:
• Improvements in Water Quality and Suitability for Recreational
Activities: EPA relies on a national water quality model to predict changes
in the ambient concentration of pollutants attributable to changes in pollutant
loadings from CAFOs. Under each regulatory scenario, the model determines
whether estimated changes in pollutant concentrations would improve the
suitability of water resources for recreational uses such as fishing and
swimming.
• Reduced Incidence of Fish Kills: Through modeling of nitrogen and
phosphorus loading reductions, the analysis estimates changes in the
frequency offish kill events under each regulatory scenario.
• Improved Commercial Shell Fishing: EPA employs data on the impact of
agricultural pollution on commercial shellfish harvesting, combined with
modeled estimates of the change in pathogen loadings from CAFOs, to
estimate the potential increase in annual shellfish harvests under each
regulatory scenario.
• Reduced Contamination of Private Wells: EPA employs data from the
U.S. Geological Survey (USGS), EPA, and the Bureau of Census to model
the relationship between nitrate concentrations in private domestic wells and
sources of nitrogen to aquifers. EPA uses this model, combined with
estimates of the change in nitrogen loadings under alternate regulatory
scenarios, to predict changes in well nitrate concentrations nationally.
3-5
-------
3.4 VALUING BENEFITS
The final step of the benefits analyses is to estimate the economic value of the modeled
physical changes in environmental quality. This section provides a brief overview of economic
valuation concepts and discusses the valuation approach applied in the studies performed for the
proposed CAFO rule.
3.4.1 Overview of Economic Valuation
Economists define benefits by focusing on measures of individual satisfaction or well-being,
referred to as measures of welfare or utility. A fundamental assumption in economic theory is that
individuals can maintain the same level of utility while trading-off different "bundles" of goods,
services, and money. The tradeoffs individuals make reveal information about the value they place
on these goods and services.
The willingness to trade-off compensation for goods or services can be measured by an
individuals' willingness to pay. While these measures can be expressed in terms of goods, services,
or money, economists generally express willingness to pay in monetary terms. In the case of an
environmental policy, willingness to pay represents the amount of money an individual would give
up to receive an improvement (or avoid a decrement) in environmental quality.20
The use of willingness to pay to measure benefits is closely related to the concept of
consumer surplus. Resource economists generally rely on consumer surplus as a measure of overall
economic welfare for benefits to individuals. The concept of consumer surplus is based on the
principle that some consumers benefit at current prices because they are able to purchase goods (or
services) at a price that is less than their total willingness to pay for the good. For example, if a
consumer is willing to pay $4 for an additional gallon of clean drinking water that costs the
consumer only $1.50, then the marginal consumer surplus is $2.50.
20 Economists also sometimes consider a similar concept of "willingness to accept
compensation"; i.e., the amount of monetary compensation that would make the individual
indifferent between having an environmental improvement and foregoing the improvement.
3-6
-------
3.4.2 Primary Approaches for Measuring Benefits
Economists generally define the economic benefits provided by a natural resource as the sum
of individuals' willingness to pay for the goods and services the resource provides, net of any costs
associated with enjoying these services.21 In some cases (e.g., commercial fishing), natural resource
products are traded in the marketplace, and willingness to pay information can be directly obtained
from demand for these commodities. In other cases, when natural resource goods or services are not
traded in the market, economists use a variety of analytic techniques to value them, or to estimate
the economic benefits of improvements in environmental quality.22 These non-market methods,
which are grounded in the theory of consumer choice, utility maximization, and welfare economics,
attempt to determine individuals' willingness to pay for natural resource services directly, through
survey research, or indirectly, through the examination of behavior in related markets. Descriptions
of market and non-market methods for analyzing benefits follow below.
• Market Methods: To measure the economic value of environmental
improvements, market methods rely upon the direct link between the quality
or stock of an environmental good or service and the supply or demand for
that market commodity. Market methods can be used, for example, to
characterize the effect of an increase in commercial fish and shellfish harvests
on market prices. In turn, these market changes affect the welfare of
consumers and producers in quantifiable ways.
• Revealed Preference: Revealed preference approaches are premised on the
assumption that the value of natural resource services to users of those
services can be inferred by indirect economic measures. For example,
willingness to pay for recreational beach services can be estimated by
observing how the number of visits individuals make to a beach varies with
the cost of traveling to the beach. Similarly, property values can be
influenced by proximity to an environmental amenity or disamenity;
econometric analysis can estimate the nature and magnitude of such effects,
providing a basis for valuing natural resource services.
21 In the case of goods and services traded in the marketplace, net benefits also include
producer surplus: the excess of producer revenues over costs. For simplicity, we leave aside for now
any discussion of producer surplus in assessing the benefits associated with enjoyment of natural
resource services.
22 These same techniques can be applied to estimate the economic damages attributable to
a decline in environmental quality.
3-7
-------
Stated Preference: Stated preference models involve the direct elicitation
of economic values from individuals through the use of carefully designed
and administered surveys. Contingent valuation techniques are the most
widely used stated preference approach, and rely on surveys designed to
derive people's willingness to pay for an amenity (e.g., improved water
quality) described in the study. This method can be used to estimate both
use and non-use values.
Averted Cost: Changes in environmental quality can impose additional
costs on the users of an affected resource. For example, contamination of
drinking water supplies might lead homeowners to purchase in-home water
filters. A potential proxy measure of the benefits of preventing pollution of
the resource is the averted cost of these expenditures.
3.4.3 Valuation of CAFO Regulatory Benefits Based on Previous Studies
Because of their high resource demands, the use of primary approaches is beyond the scope
of this analysis. Instead, the analysis draws on previous studies that evaluated similar water quality
benefits issues. This approach—typically referred to as "benefits transfer"—involves the application
of values, functions, or data from existing studies to estimate the benefits of the resource changes
currently being considered, and is commonly used in analyzing the benefits of proposed
environmental regulations. The primary research material and analytic approach used for the
valuation of each benefit category are summarized below; more detailed descriptions of the methods
applied are provided in subsequent chapters of this report.
• Improvements in Water Quality and Suitability for Recreational
Activities: To determine how people value improvements in the suitability
of water resources for recreational activities (e.g., fishing, swimming), the
analysis relies on the results of a contingent valuation survey conducted by
Carson and Mitchell (1993). Based on this study, economic benefits are
determined from people's willingness to pay for achievement of water quality
levels that restore affected waters to fishable or swimmable conditions.
• Reduced Incidence of Fish Kills: The valuation of benefits from the
reduced incidence offish kills is based on fish replacement costs, as reflected
in an American Fisheries Society (1990) report. Because this value
represents only a portion of the economic damages associated with fish kills,
it likely provides a conservative estimate of the benefits of reducing the
frequency of such events.
3-8
-------
Improved Commercial Shell Fishing: To value the economic benefit of
increased shellfish harvests, the analysis relies on available literature that
models consumers' demand for shellfish. Based on the demand equations
from these primary sources, EPA determines the increase in consumer surplus
that would result from increased harvests.
Reduced Contamination of Private Wells: The analysis surveys the
literature concerning the values people place on avoiding or reducing nitrate
contamination in private domestic wells. Based on this review, it develops
estimates of people's willingness-to-pay to reduce nitrate concentrations to
certain levels, and applies these estimates to value predicted changes in the
quality of water that supplies private wells.
3.4.4 Aggregating Benefits
The final step in determining the benefits of the proposed CAFO regulatory scenarios is
aggregation of the benefits calculated for each of the benefit categories. To avoid over-estimation,
this requires consideration of the extent to which underlying analyses may double-count certain
benefits. For this analysis, however, the benefits that each of the underlying studies explore are
relatively distinct. As a result, the potential for double-counting appears to be small.
Another consideration in aggregating benefits is ensuring that all values are reported on a
comparable basis, taking into account the effects of inflation on real dollar values. For purposes of
this analysis, all values are reported in 1999 dollars. The price indices employed in converting
source data to 1999 dollars vary, depending on which index is most appropriate. Further information
on these adjustments is provided in the detailed discussion of each analysis.
The detailed analyses presented in Chapters 4 through 7 report benefits on an annual basis.
To determine the present value of these benefits, EPA employs three alternative discount rates: a
7 percent real discount rate, which is representative of the real rate of return on private investments
and consistent with the rate mandated by the Office of Management and Budget for analysis of
proposed regulations; a 3 percent real discount rate, which is representative of the social rate of time
preference for consumption of goods and services, and consistent with the rate recommended by
many economists for analysis of environmental benefits; and a 5 percent real discount rate, which
represents the mid-point of the 3 and 7 percent range.
In calculating the present value of benefits at the time new regulations are implemented, EPA
assumes an infinite time frame; i.e., as long as the regulations remain in effect, the associated
benefits will be enjoyed in perpetuity. EPA further assumes that its estimates of beneficial impacts
on surface water resources will be fully realized in the year immediately following implementation
of the revised regulations. This assumption reflects EPA's judgment that reductions in the loadings
of pollutants from CAFOs will quickly yield improvements in surface water quality. With respect
to reduced contamination of private wells, however, EPA assumes that several years will pass before
the full benefits of the regulation are realized. To permit consistent comparison of these benefits to
the annual benefits estimated for surface water resources, EPA presents the benefits of reduced
3-9
-------
contamination of private wells on an annualized basis, as well as on a present value basis. The
calculation of an annualized value for this benefits category indicates the constant flow of benefits
over time that would generate the same present value as the anticipated, uneven, flow of benefits.
Additional information on the calculation of present values and the aggregation of benefits
is presented in Chapter 8.
3.5 SUMMARY
Exhibit 3-2 summarizes EPA's approach to measuring and valuing the anticipated benefits
of the revised CAFO regulations. Additional information on the methods employed is provided in
the detailed discussion of each analysis that follows.
Exhibit 3-2
SUMMARY OF APPROACH TO ESTIMATING REGULATORY BENEFITS
Benefit Category
Improvements in
Water Quality and
Suitability for
Recreational
Activities
Reduced Incidence
of Fish Kills
Improved
Commercial
Shellfishing
Reduced
Contamination of
Private Wells
Human Use
Recreational fishing,
swimming, and non-use
benefits associated with
surface water resources.
Recreational fishing, near-
stream use and non-use
benefits.
Commercial shellfishing.
Drinking water.
Measurement Approach
Model potential changes in
water quality based on
estimated changes in
loadings of CAFO-related
pollutants.
Estimate changes in the
frequency of fish kill events
based on estimated
reductions in nutrient
loadings.
Estimate increased access to
shellfish growing waters
and resulting increase in
annual shellfish harvests,
based on modeled changes
in fecal coliform
concentrations.
Model potential changes in
private domestic well water
quality based on estimated
changes in loadings of
CAFO-related pollutants.
Valuation Approach
Stated preference approach
assessing willingness-to-
pay for water quality that
supports recreation.
Avoided damages based on
fish replacement costs.
Market estimate of
increased consumer surplus.
Stated preference approach
assessing willingness-to-
pay to reduce the
concentration of nitrates in
water drawn from private
domestic wells.
3-10
-------
3.6 REFERENCES
AFS. 1990. American Fisheries Society Socioeconomics Section, A Handbook of Monetary Values
of Fishes and Fish-Kill Counting Guidelines, Draft, July 1990.
Carson, Richard T. and Robert Cameron Mitchell. 1993. "The Value of Clean Water: The Public's
Willingness to Pay for Boatable, Fishable, and Swimmable Water Quality." Water
Resources Research, Vol. 29, No. 7.
3-11
-------
MODELING OF IMPROVEMENTS IN
SURFACE WATER QUALITY AND BENEFITS
OF ACHIEVING RECREATIONAL USE LEVELS CHAPTER 4
4.1 INTRODUCTION AND OVERVIEW
A major component of EPA's CAFO benefits analysis is an assessment of how water quality
in freshwater rivers and lakes would be influenced by reduced CAFO pollution, accompanied by an
evaluation of the economic value of these changes to society. EPA has developed a comprehensive
analysis of these benefits using the methodology summarized in Exhibit 4-1. As shown, key
components of the analysis include:
Development of model facilities that typify conditions across different production sectors,
facility sizes, and geographic regions;
• Modeling of "edge-of-field" pollutant releases that take into account manure management
practices, manure constituents, and physical conditions (e.g., soil characteristics);
Calculation of the number of AFOs in the various production sectors/size categories to allow
extrapolation of the model facility loadings estimates;
• Modeling of the change in surface water pollutant concentrations as determined by changes
in loadings; and
Valuation of the water quality changes through a benefits transfer analysis focused primarily
on the public's willingness to pay for improved water conditions necessary to support
recreation.
EPA implements this set of analyses for baseline conditions as well as the various regulatory
scenarios under consideration to allow estimation of overall water quality benefits. The following
sections summarize the five analytic components and the resulting estimates.
4-1
-------
Exhibit 4-1
OVERVIEW OF RECREATIONAL BENEFITS ANALYSIS
Model Facility
Analysis
Edge-of-Field
Loadings
Analysis
Surface
Water
Modeling
t
Analysis of
AFO/CAFO
Distribution
Valuation of
Water Quality
Changes
4.2 MODEL FACILITY ANALYSIS
Assessing the impacts of CAPO regulatory scenarios requires that EPA recognize the
diversity of animal feeding operations across the country. Exhibit 4-2 provides an overview of the
analysis used to define model facilities and their associated pollution potential.23 For detailed
information regarding the development of model facilities, see Chapters 4 and 11 of the Technical
Development Document of Proposed Effluent Limitations Guidelines for Animal Feeding Operations
(EPA, 2000a), hereafter referred to as the "TDD".
Exhibit 4-2
MODEL FACILITY ANALYSIS
Analyses:
Determine Application
Rates and Nutrient
Removal Capabilities
for Model Facilirv
Define
Model
Facility
Calculate
Pollutant
Production
Calculate
Excess Nutrients
Applied
Input Data:
1997 Census
of Agriculture
• Manure Production
• Nutrient Content
• Pathogen & Metal Content
• Number of Operations
• Available Land for Manure Applications
• Typical Cropping Systems and Yields
• Crop Removal Rates
Agronomic Manure
Application Rates
23 Note that for this analysis, the term agriculture facility, facility, or operation includes the
feedlot and the land application area under the control of the feedlot operator.
4-2
-------
First, EPA disaggregates the universe of AFOs according to a suite of characteristics directly
affecting manure generation, manure management, and pollutant loadings. AFOs are grouped into
five geographic regions, as shown in Exhibit 4-3. To establish geographic regions, EPA developed
algorithms to estimate the number of facilities by size (number of animals), using a combination of
inventory and sales data. NASS applied the algorithms to 1997 Census of Agriculture data to
generate the output by which EPA estimated facility counts. Due to disclosure criteria established
by NASS to protect respondent-level census data, the regions were aggregated into broader
production regions.
Exhibit 4-3
GEOGRAPHIC REGIONS FOR GROUPING AFOS
BiWicomico
Poultry
Imperial/Bee
y
^—
Historical Climate Station
County with Highest
Production by Sector
State Boundaries
AFO Regions
Okeechobee
Beef-Daiiy
Within each geographic region, EPA defines model facilities by production sector, subsector,
and size (number of animals). Based on these various dimensions, an example of a model facility
would be a large beef facility with more than 8,000 head in the Midwest region. Exhibit 4-4
4-3
-------
summarizes the key dimensions on which model facilities are defined. In all, EPA considered 200
different model facilities. The key model facilities are those that reflect the majority of production,
resulting in approximately 76 different model facilities used for further analysis.
Exhibit 4-4
SUMMARY OF MODEL FACILITY DIMENSIONS
Production Sector
Facility Size
Regions
Beef, cattle
Beef, veal
Dairy, milk
Dairy, heifers
Swine, farrow-finish
Swine, grower-finish
Layer, wet manure system
Layer, dry manure system
Broiler
Turkey
> 1,000 Animal Units
500-1,000 Animal Units
300-500 Animal Units
Pacific
Central
Midwest
South
Mid-Atlantic
To guide the selection of modeling parameters related to fields and soils, EPA must identify
a specific location for each model facility in a given geographic region. For these purposes, the
analysis assumes that the model facility is located in the highest animal-production county of the
region's highest production state for a given animal type.
EPA calculates manure production and the associated production of pollutants for each model
facility using a process developed by Lander, et al. (1998), and refined by Kellogg, et al. (2000). The
number of animals per operation is converted to USDA animal units24 using conversion factors
standardized to a 1,000-pound beef cow. EPA multiplies the number of animal units per model
facility by the manure production per animal unit to determine total manure production. Manure
production is adjusted to reflect the fraction that is recoverable, i.e., the portion of manure that is
collected, stored, or otherwise managed so as to be available for land application. Finally, EPA
calculates total generation of nutrients based on the typical nitrogen and phosphorus concentrations
per unit of recoverable manure for each animal type, e.g., pounds of nitrogen per ton of manure from
finishing pigs in the swine sector.25
24 The USDA animal unit is based on average liveweight of the animal, and is markedly
different from the animal unit definition in EPA's regulations at 40 CFR 122 and 412.
25 Metal production (zinc, copper, cadmium, nickel, lead) is calculated in terms of pounds
of metals excreted per animal unit, while pathogen production (fecal coliform and fecal
streptococcus) is calculated in terms of colonies per animal unit.
4-4
-------
Next, EPA defines land application practices for each model facility and the capacity for soil
and crop removal of nutrients applied to the land. This analysis entails several steps. The analysis
first considers the total nitrogen and phosphorus generated in manure at the model facility. EPA
divides these figures by the average total acreage available for land application of manure for an
operation in the given region, size class, and production sector; this average acreage is drawn from
a recent NRCS study (Kellogg et al., 2000).
EPA then considers the likely cropping systems at the model facilities and relates the quantity
of nutrients applied annually to cropland and pastureland nutrient requirements. For example, typical
cropping systems for the Mid-Atlantic AFO Region are corn, soybean, and wheat in two-year
rotation. The ratio of nutrients applied to crop nutrient requirements provides a measure of the
excess nutrients applied in the manure.26 This in turn forms the foundation for loadings analyses of
regulatory scenarios that call for adherence to agronomic rates of nutrient application. To
characterize land application practices, the analysis considers three categories of facilities:
Category 1 facilities include CAFOs with sufficient crop- or pastureland on-
site to apply the manure they generate at agronomic rates. The analysis
assumes that these facilities apply all manure on-site (i.e., no manure is
shipped off-site) under both baseline and post-regulatory conditions.
• Category 2 facilities include those with insufficient crop- or pastureland on-
site to apply the manure they generate at agronomic rates. For the baseline
scenario, the analysis assumes that these facilities apply all the manure they
generate on-site, regardless of the degree to which agronomic application
rates are exceeded. For the post-regulatory scenarios, the analysis assumes
that on-site manure application is limited to the agronomic rate, and that the
remaining manure is shipped off-site for application to crop- or pastureland
at agronomic rates. EPA's model captures the pollutant loadings associated
with both on-site and off-site application of the manure generated by
Category 2 facilities.27 (The sole exception to this approach occurs in
modeling loadings from Category 2 broiler operations. The baseline analysis
caps on-site application of this manure at five times the agronomic rate; any
excess manure is assumed to be shipped off-site. The analysis of post-
regulatory conditions assumes that Category 2 broiler facilities apply manure
26 EPA assumes that 30 percent of the animal waste's nitrogen content volatilizes during and
shortly after land application. The analysis also assumes that facilities use no fertilizers other than
manure.
27 For consistency, pollutant loadings from the off-site cropland to which these facilities are
assumed to ship manure are also captured in the baseline analysis. The modeling of baseline
conditions assumes the application of commercial fertilizer to this land.
4-5
-------
on-site only up to the agronomic rate, beyond which the excess is shipped off-
site. In both cases, the manure shipped off-site is not captured in the loadings
modeling.)
Category 3 facilities include CAFOs without cropland. EPA assumes that
these facilities transfer all manure off-site for use or disposal. The pollutant
loadings associated with this manure are not captured in modeling either
baseline or post-regulatory conditions.
4.3 EDGE-OF-FIELD LOADINGS ANALYSIS
The second major component of the water quality analysis is estimation of pollutant loadings
leaving the model facility, i.e., edge-of-field loadings. EPA estimates the loadings associated with:
(1) application of manure; (2) lagoons and other storage structures; and (3) feedlots. The sections
below review the methods applied for each of these analyses.
4.3.1 Loadings from Manure Application
EPA's loadings analysis first examines loadings from manure application to cropland and
pastureland. The analysis combines information on manure generation and land application practices
(see above) with data on the timing of application, hydrological conditions, geological conditions,
and weather patterns (see Exhibit 4-5). EPA integrates these data using the Groundwater Loading
Effects of Agricultural Management Systems (GLEAMS) model. This field-scale model simulates
hydrologic transport, erosion, and biochemical processes such as chemical transformation and plant
uptake. The model uses information on soil characteristics and climate, along with nutrient
production data, to model losses of nutrients in surface runoff, sediment, and groundwater leachate.
Loadings are modeled for the pre- and post-regulatory scenarios to estimate changes in loadings
attributable to the proposed standards.
The data used in the GLEAMS model runs include the following:
• Soils Data: GLEAMS uses data from the State Soil Geographic (STATSGO) data base
maintained by USDA's Natural Resources Conservation Service. Key soil parameters drawn
or estimated from the data base include permeability, soil porosity, baseline organic matter
content, percent clay, and percent silt. EPA employs data on these parameters, in
combination with data on other factors (see below), to characterize soil erosion, surface
runoff, and groundwater leaching at model facilities.
4-6
-------
Exhibit 4-5
EDGE-OF-FIELD LOADINGS ANALYSIS FOR MODEL FACILITIES
Input Data Analyses
Excess Nutrients Applied
(from model facility analysis)
Farm Practice Data/Assumptions:
• Planting Date
• Harvest Data
• Manure Application Date
Soils Information
(permeability, porosity, etc.)
Climate Data
Manure Characteristics
GLEAMS Modeling to
Calculate Pollutant Runoff
from Manure Application
Calculate Seepage Losses
from Lagoons and Other
Storage Structures
Total Loadings for
Each Model Facility
(pre-and post-regulation)
Climate Data: EPA prepared climate data using CLIGEN, a synthetic climate generator
commonly used in conjunction with a variety of agricultural runoff models. CLIGEN
simulates weather patterns based on 25 or more years of precipitation and temperature data.
Crop Planting and Harvest Dates: EPA developed assumptions for crop planting and
harvesting using USDA reports and determined likely manure application dates for model
facilities based on contacts with USDA Extension Agents in relevant locations. The
application dates are a function of the crops grown. Some single-cycle crops (e.g., corn)
allow only one application per year, while other crops (e.g., alfalfa) allow multiple
applications.
4.3.2 Loadings from Lagoons and Other Storage Structures
Lagoons and other manure storage structures at animal feedlots are also potential pollution
sources, posing risks primarily through seepage to groundwater and subsequent discharge to surface
water. For the purposes of this analysis, EPA assumes that all lagoons and other storage structures
leak. Storage structure seepage estimates were obtained from Ham and DeSutter (1999), who
measured nitrogen that leaked from three established swine-waste lagoons in Kansas. From these
results, it was assumed that 2,000 pounds per acre per year leaked from manure storage structures
4-7
-------
lined with silt loam soils. EPA scales seepage estimates for clay and sandy soils from these
estimates as described in the TDD.
For most storage structures, EPA models transport of pollution through groundwater and
estimates the associated attenuation of pollutants. However, conditions in some cases (as defined by
Sobecki and Clipper, 1999) suggest that leaks from lagoons or other storage structures may seep
directly to surface water, i.e., hydrologic conditions are such that pollutant concentrations are not
attenuated by dilution in groundwater. This might occur, for example, in the presence of sandy soils
or karst-like terrain. To characterize the potential for leaks from lagoons or other storage structures
to seep directly to surface water, EPA evaluated soil and hydrological conditions in each AFO
region. Based on this evaluation, EPA determined the percentage of the region's area in which the
potential for direct contamination of surface water is high. EPA's analysis assumes that this
percentage of storage leaks in each region would result in direct contamination of surface water.
4.3.3 Loadings from Feedlots
Another pollution source that EPA analyzes is runoff from feedlots. These loadings can be
particularly significant in the beef sector because the animals are typically housed in open lots.
To estimate feedlot runoff loadings, EPA first calculates the volume of runoff from the
feedlot at the model facility. The annual depth of runoff from the feedlot is calculated for each of
the five AFO regions using average precipitation from the National Climatic Data Center. The
volume of runoff is calculated using this depth of runoff and the estimated area of the dry lot and
feedlot handling areas for each model facility.28
To characterize the loadings of pollutants in feedlot runoff, EPA assumes a solids content
of 1.5 percent. The composition of these solids is estimated based on the characteristics of dry
manure, which varies across production sectors. Annual loadings of specific pollutants are then
determined, based on the estimated composition of solids, the assumed percentage of solids in
feedlot runoff, and the estimated annual volume of runoff from the feedlot.
4.3.4 Model Loadings Under Regulatory Scenarios
EPA applies the data and methods described above to analyze loadings associated with
baseline conditions and each of the various regulatory scenarios. Under all of the options, the
analysis assumes that regulated facilities modify current activities to comply with feedlot best
28 EPA assumes that only surface runoff occurs from the feedlot.
4-8
-------
management practices, mortality handling requirements, nutrient management planning/
recordkeeping, and elimination of manure application within 100 feet of surface water. Factors that
vary among regulatory scenarios that may impact loadings include:
• reduction of manure application to agronomic nitrogen rates; and
• reduction of manure application to agronomic phosphorous rates.
4.4 ANALYSIS OF AFO/CAFO DISTRIBUTION
To develop a national estimate of baseline pollutant loadings from AFOs, as well as estimates
of the change in loadings under alternate regulatory scenarios, EPA must determine the number of
operations that would be governed by the proposed regulations, i.e., the number of facilities
considered to be AFOs and the number of AFOs considered to be CAFOs, and therefore subject to
the proposed rule. These operations represent the universe to which model facility results are
extrapolated.
The sections below discuss EPA's approach and the resulting characterization of the
population of AFOs and CAFOs. More detailed information on the procedure used by EPA to
estimate the number of operations that may be subject to the proposed regulations can be found in
the TDD (USEPA, 2000a).
4.4.1 Approach
EPA estimates the number of operations that may be affected by alternative requirements
using a two-step procedure. First, EPA determines the number of operations that raise animals under
confinement by using available data on the total number of livestock and poultry facilities (see
below). Next, the number of CAFOs is determined based on operations that are defined as CAFOs
and smaller operations that are designated as CAFOs based on site-specific conditions, as
determined by the permitting authority. For purposes of this discussion, the affected CAFO
population includes those facilities that discharge or have the potential to discharge to U.S. waters.
This definition does not include those smaller operations that are not defined or designated as
CAFOs.
The USD A Census of Agriculture is a complete accounting of United States agricultural
production and is the only source of uniform, comprehensive agricultural data for every county in
the nation. The Census is conducted every five years by USDA's National Agricultural Statistics
Service (NASS).29 The Census is implemented through a mail questionnaire that is sent to a list of
known U.S. agriculture operations from which $1,000 or more of agricultural products were
produced and sold or normally would have been sold during the census year.
29 In prior years, the Census was conducted by the Department of Commerce's Bureau of the
Census.
4-9
-------
Aggregated 1997 Census data are readily available from USDA. In general, the published
compendium provides summary inventory and sales data for the nation and for states. The Census
database itself, however, contains respondent-level information that can be aggregated into more
precise agriculture facility size groupings. The requested data summaries used for EPA's analysis
were compiled with the assistance of staff at USDA's NASS, who performed special tabulations of
the data to obtain information on the characteristics of facilities at specific size thresholds for each
sector. All data provided to EPA were aggregated to ensure the confidentiality of an individual
operation. EPA supplemented the available data with information from other sources, including
other USDA data sets and industry publications. The following discussion briefly notes the nature
of key gaps in the Census data and EPA's approach to addressing them.
• All USDA data are reported across all animal agriculture operations and do not distinguish
between confinement and nonconfinement production types (e.g., pasture or rangeland
animals). However, only operations that raise animals under confinement (as defined under
40 CFR 122 Appendix B) would be subject to the proposed regulations. For analytical
purposes, EPA has assumed that all animals at larger dairy and poultry operations are grown
under confinement, which may overstate EPA's estimate of the number of operations subject
to the regulation. For the beef and hog sectors, the USDA has limited data on the number
of operations that are feedlot operations only. NASS Statistical Bulletin Number 953,
Cattle: Final Estimates 1994-1998, was used to estimate the number of beef feedlots with
more than 1,000 head; 1997 Census data on "Cattle fattened on grains and concentrates
(sold)" are used to distinguish confinement from non-confinement operations with less than
1,000 head (USDA/NASS, 1999a; USDA/NASS, 1999b). Available information from USDA
and industry feedback was used to adjust the total number of hog operations to exclude those
that are pasture operations (USDA/Animal and Plant Health Inspection Service, 1995;
NPPC, 1997).30
• Available Census data on the number of animal facilities by inventory size distribution do
not always correspond with the facility size definitions examined by EPA. Where data were
not available in the desired size ranges, EPA interpolated estimates from available data by
assuming that, for a given size group, the largest 40 percent of the facilities account for 60
percent of the animal inventory.
30 Available information from USDA indicates that few large hog operations are non-
confinement facilities (USDA/APHIS, 1995); therefore EPA assumes that all hog operations with
more than 2,500 swine are AFOs.
4-10
-------
USDA data are also not available for the number of poultry operations with wet manure
management systems. EPA estimated these figures using available data from USDA and
supplemental information from industry experts and agricultural extension agency personnel.
Information on the number of animal facilities that raise more than a single animal type is
also not available. EPA algorithms written for use with the 1992 Census data included an
estimate of CAFOs that maintained more than one animal type. The analysis revealed that,
for facilities with more than 1000 animal units, 21 percent raise more than one animal type;
for facilities with less than 1000 animal units, about 25 percent raise more than one animal
type. To the extent that combinations of animal types are located at facilities, facility counts
may be overstated.
Finally, USDA Census data report the number and size of livestock and poultry facilities as
of year-end (December 31) and may not adequately reflect seasonal fluctuations in beef,
dairy, and layer inventory, or the year-to-year fluctuations in number of animals sold. EPA
algorithms reflect average herd sizes at larger confinement facilities over the year. The
outputs are based on both reported inventory and sales, adjusted by expected turnovers. This
approach is consistent with that developed by USDA to estimate potential manure nutrient
loadings from animal agriculture (Lander, et al., 1998; Kellogg, et al., 2000).
4.4.2 Estimated Number of AFOs and CAFOs
Based on the USDA data sources described above, there were 1.2 million livestock and
poultry facilities in the United States in 1997. This number includes all operations in the beef, dairy,
pork, broiler, layer, and turkey production sectors, and includes both confinement and non-
confinement (grazing and range fed) production.
Of all these operations, EPA estimates that there are about 376,000 AFOs that raise or house
animals in confinement, as defined by the existing regulations. Exhibit 4-6 summarizes the number
of AFOs by production sector and facility size.
4-11
-------
Exhibit 4-6
TOTAL NUMBER OF AFOs BY PRODUCTION SECTOR AND FACILITY SIZE
Production Sector
Beef: cattle
Beef: veal
Dairy: milk
Dairy: heifers
Hogs: FF2
Hogs: GF2
Broilers
Layers: wet3
Layers: dry3
Turkeys
Sum Total
Total AFOs4
Total AFOs
106,080
850
116,880
1,250
64,240
53,620
34,860
3,110
72,060
13,720
466,670
375,740
^OOOAU1
2,080
10
1,450
400
2,420
1,670
3,940
360
360
370
13,060
12,850
300 AU-1000 AU
2,000
200
5,690
750
9,240
3,250
10,200
800
1,330
1,730
35,190
28,150
<300 AU
102,000
640
109,740
100
52,580
48,700
20,720
1,950
70,370
11,620
418,420
334,740
Source: Values presented in the table are EPA estimates, derived from published USDA/NASS data, including 1 997 Census
of Agriculture. For more information, see Technical Development Document of Proposed Effluent Limitations
Guidelines for Animal Feeding Operations.
1 As defined by the existing regulation, one animal unit (AU) is equivalent to one slaughter or feeder cattle; 0.7 mature dairy
cattle; 2.5 hogs (over 55 pounds); 0.5 horses; 10 sheep or lambs; 55 turkeys; 100 laying hens or broilers (with continuous
overflow watering); 30 laying hens or broilers (with liquid manure system); or 5 ducks.
2 FF = farrow-finish (includes breeder and nursery pigs); GF=grower-finish. Data from USDA's NAHMS indicate that
roughly 40 percent of hog farms are grower- finish and 60 percent are farrowing operations.
3 The "Layers: wet" category covers operations with liquid manure systems. Such AFOs are currently defined as CAFOs
for operations with 30,000 birds (1,000 AUs). No layer operations use continuous watering systems. "Layers: dry" are
defined at 1,000 AUs for operations with 100,000 birds.
4 "Total AFOs" eliminates double counting of operations with mixed animal types. Operations with mixed animal types
account for roughly 20 percent of total AFOs.
Exhibit 4-7 summarizes the total number of CAFOs defined or designated under each of the
regulatory scenarios. Under the proposed scenarios, between about 29,000 and 33,500 AFOs will
be defined or designated as CAFOs, and are therefore subject to the proposed rule.31
31 This number is likely the upper bound estimate of the total number of operations that will
be subject to the proposed revisions.
4-12
-------
Exhibit 4-7
TOTAL NUMBER OF CAFOs BY
REGULATORY SCENARIO
Regulatory Scenario
Option 1 -Scenario 1
Option 1 -Scenario 2/3
Option 1 -Scenario 4a
Option 1 -Scenario 4b
Option 2-Scenario 1
Option 2-Scenario 2/3*
Option 2-Scenario 4a*
Option 2-Scenario 4b
Total CAFOs
17,700
33,500
28,980
45,140
17,700
33,500
28,980
45,140
Source: Values presented in the table are EPA estimates, derived from
published USDA/NASS data, including 1997 Census of
Agriculture. For more information, see Technical Development
Document of Proposed Effluent Limitations Guidelines for Animal
Feeding Operations.
* Proposed scenarios.
4.4.3 Geographic Placement of Facilities
Finally, AFOs and CAFOs by region are placed into counties (and eventually watersheds)
using the published county level Census data. Where county level data was not presented, the
facilities in the undisclosed counties were imputed from state- and region-level data.
4.5 SURFACE WATER MODELING
EPA develops estimates of changes in surface water quality by building on the analysis
of edge-of-field pollutant loadings for model facilities and the analysis of the distribution of
AFOs and CAFOs. These data are integrated into the National Water Pollution Control
Assessment Model (NWPCAM), a national-scale model designed to translate pollutant loadings
into water quality changes and associated economic benefits to support policy-level regulatory
decisionmaking.
4-13
-------
NWPCAM covers virtually all inland waters in the U.S., allowing EPA to examine how
changes in loadings under various regulatory scenarios would influence key water quality
parameters.32 The model incorporates routines that simulate overland transport of pollutants,
discharge of pollutants to nearby surface waters, discharges to surface water from other (non-
AFO/CAFO) sources, and the fate and transport of pollutants in the interconnected network of
surface waters. Specifically, the modeling involves the following steps:
Developing the network of rivers and streams that serves as the geographic foundation for
the modeling;
• Distributing AFO/CAFOs and associated facility-level edge-of-field loadings to
agricultural lands within a defined watershed or county;
Simulating transport of nutrients/pollutants and subsequent discharge to nearby
waterbodies;
• Delivering nutrient/pollutant loadings from point sources (e.g., municipal wastewater
treatment plants, industrial facilities) and non-point sources (e.g., non-AFO/CAFO
agricultural run-off, municipal run-off) to waterbodies; and
• Simulating dilution, transport, and kinetics of the nutrients/pollutants loaded to the
waterbody as the nutrients/pollutants are transported along the waterbody.
Exhibit 4-8 summarizes these steps and the primary data used in the analysis. The sections below
discuss the modeling in more detail and provide an overview of the estimated changes in pollutant
loadings under the various regulatory scenarios.33
32 NWPCAM does not address water quality benefits in bays, estuarine waters, or the Great
Lakes.
33 Both the water quality modeling and the economic benefits analysis are presented in greater
detail in Estimation of National Surface Water Quality Benefits of Regulating Concentrated Animal
Feeding Operations (CAFOs) Using the National Water Pollution Control Assessment Model
(NWPCAM) (USEPA, 2000b). This report is provided under separate cover as Attachment A.
4-14
-------
Analyses: [Develop Hydrologi.
Network
Input Data:
1
L
•ReachFileDataon
Surface Waters
•Land Use/Land
Cover Data
•Watershed Data
Exhibit 4-8
WATER QUALITY MODELING ANALYSIS
Denver Loadings from
Other Point and Non-Point
Sources to Waterbodies
•Loadings Data for Point Sources
•Loadings Data for Non-Point
Sources (other than AFOs/CAFOs)
4.5.1 Defining the Hydrologic Network
In the initial step of the analysis, EPA prepares the hydrological network of rivers and
streams that serves as the geographic backdrop to the modeling. The hydrological network is
developed from EPA's Reach Files, a series of hydrologic databases describing the inland surface
waters of the U.S. Each "reach" in the database represents a segment of a river or stream; these
segments are linked together to characterize complete systems of rivers and streams. EPA's
Reach File 3 (RF3) forms the geographic foundation for NWPCAM, allowing the model to
simulate the flow of water and pollutants from a point of origin to major rivers, and ultimately
to ocean discharge.34
Once the hydrologic network is established, EPA uses a geographic information system
(GIS) approach to overlay information on land-cover, characterizing land across the U.S. at a
square-kilometer degree of resolution. From these data, EPA can identify areas classified as
"agricultural" land. Each land section, or "cell", is associated with a river reach in the hydrologic
network.
34 RF3 includes numerous small tributaries and headwaters. To simplify the modeling, EPA
focuses on loadings to a subset of larger rivers and streams (referred to as "RF3 Lite"). Unless
otherwise noted, the water quality and economic modeling results reported below reflect changes in
loadings to the RF3 Lite subset of RF3.
4-15
-------
4.5.2 Distributing AFOs and CAFOs to Agricultural Land
Once the hydrologic network is established, NWPCAM integrates data on the location of
AFOs and CAFOs to spatially orient the facilities relative to surface waters. This analytic step
links directly to the analyses discussed above wherein EPA determined the numbers of AFOs and
CAFOs by county and, through analysis of model facilities, estimated the edge-of-field loadings
associated with each facility. Here, AFOs/CAFOs and their associated edge-of-field loadings are
randomly distributed to agricultural land in the appropriate county. By placing each facility in
a land use cell, the facility can be linked to a river reach in the hydrologic network.35
4.5.3 Calculating AFO/CAFO-Related Loadings to Waterbodies
Once facility pollutant loadings are linked to a geographic area and river reach, these
loadings are delivered from the agriculture cells to the river reaches using a routine to simulate
an overland transport process. Overland travel times and associated nutrient decay are based on
flow in a natural ditch or channel, as may typically be found on agricultural lands. A unit runoff
(ft3/sec/km2) is derived for each watershed based on data compiled by the U.S. Geological Survey.
The unit runoff therefore represents runoff from each agricultural cell within the watershed and
can be used to derive time-of-travel estimates necessary to route pollutants from the facility to a
river reach. NWPCAM also calculates nutrient/pollutant decay associated with overland
transport. Total loadings to any given river reach are the total loadings discharged from all land-
use cells draining to the reach (as well as discharges from upstream river reaches).
4.5.4 Loadings from Other Sources
In addition to loadings from AFOs/CAFOs, NWPCAM integrates data on loadings from
other pollutant sources. This complete inventory of loadings is needed to assess the cumulative
changes in water quality (i.e., the attainment of beneficial use levels) in surface waters.
Specifically, the model integrates data on discharges from municipal and industrial point sources
as well as loadings from (non-AFO) non-point sources. These loadings are constant across
regulatory scenarios. To model nutrient loads for non-point sources, EPA uses SPARROW
(STMtially Referenced Regression On Watershed attributes) (Smith, et al., 1997), a statistical
modeling approach for estimating major nutrient source loadings at a detailed geographic scale
based on watershed characteristics.36
35 EPA did not model facilities below 300 animal units. Discharges from these small
operations were included in the agriculture nonpoint source component.
36 Non-point source data for fecal coliform, fecal streptococci, and sediments were not
available at the national level; therefore, only nutrients are covered in the analysis of non-AFO non-
point sources.
4-16
-------
4.5.5 Fate and Transport Modeling
Once all loadings to surface waters have been estimated, NWPCAM routes pollutants
through the hydrologic network from upstream to downstream reaches. The model simulates
pollutant decay during this routing process. The resulting pollutant concentrations are then
compared to beneficial use criteria to determine how potential recreational uses would change
with improved water quality (see below).
4.5.6 Estimated Changes in Loadings
Exhibit 4-9a summarizes the NWPCAM estimates of baseline loadings from AFOs and
CAFOs and shows loadings under the various regulatory scenarios. Similarly, Exhibit 4-9b
presents the resulting removals under each of the regulatory scenarios. As shown, reduction of
nutrient and pollutant loadings is greater under the phosphorus-based standards (Option 2),
particularly under Option 2-Scenario 4b. Significant reductions are also realized under the
proposed scenarios; for example, nitrogen loadings are reduced from 67 million kilograms per
year in the baseline to between 37 and 38 million kilograms per year.
Exhibits 4-10a and 4-1 Ob present additional modeling results for metals. As shown,
significant reductions in loadings of metals could be realized under the regulatory scenarios.
However, these results differ from those presented above for several reasons. First, the estimates
reflect changes in loadings to the larger set of RF3 rivers and streams as opposed to the smaller
set of RF3 Lite rivers and streams. Second, EPA does not model metals independently; instead,
overland transport of metals is assumed to be similar to the overland transport of phosphorus,
with the assumption that metals adhere to soil in a manner similar to phosphorus. For these
reasons, the metals loading reductions cannot be compared directly to the reported reductions for
nutrients and other pollutants. The metals reductions are reported here for descriptive purposes
and do not play a role in the calculation of economic benefits (see below).
4.6 VALUATION OF WATER QUALITY CHANGES
To value predicted reductions in the pollution of rivers and streams by CAFOs,
NWPCAM applies estimates of Americans' willingness to pay for improvements in water quality.
This approach first entails relating changes in water quality parameters - e.g., concentrations of
chlorophyll • - to the ability of a body of water to support fishing or swimming. Once the
potential improvement in the ability of modeled rivers and streams to support these uses is
determined, the analysis relies upon estimates of willingness to pay for such improvements
obtained from the results of a contingent valuation survey developed by Richard Carson and
Robert Mitchell. This survey, which is national in scope, characterizes households' annual
willingness to pay to improve freshwater resources from baseline conditions to fishable or
swimmable quality.
4-17
-------
Exhibit 4-9a
ESTIMATED ANNUAL AFO/CAFO NUTRIENT/POLLUTANT LOADINGS
UNDER BASELINE AND REGULATORY SCENARIOS
Regulatory Scenario
Baseline Conditions
Option 1-Scenario 1
Option 1-Scenario 2/3
Option 1-Scenario 4a
Option 1-Scenario 4b
Option 2-Scenario 1
Option 2-Scenario 2/3*
Option 2-Scenario 4a*
Option 2-Scenario 4b
Nitrogen
(million kg)
67
53
52
53
50
43
37
38
34
Phosphorus
(million kg)
76
41
31
34
28
34
22
24
17
Fecal
Coliforms
(billion
colonies)
50
24
19
21
15
21
15
18
12
Fecal
Streptococci
(billion
colonies)
117
80
72
73
70
67
57
59
52
Sediments
(billion kg)
118
118
118
118
118
92
83
84
80
Source: Estimation of National Surface Water Quality Benefits of Regulating Concentrated Animal Feeding
Operations (CAFOs) Using the National Water Pollution Control Assessment Model (NWPCAM),
prepared for U. S. EPA Office of Wastewater Management, prepared by Research Triangle Institute,
August 2000.
* Proposed scenarios.
4-18
-------
Exhibit 4-9b
ESTIMATED ANNUAL REMOVALS UNDER REGULATORY SCENARIOS
Regulatory Scenario
Option 1-Scenario 1
Option 1-Scenario
2/3
Option 1-Scenario 4a
Option 1-Scenario 4b
Option 2-Scenario 1
Option 2-Scenario
2/3*
Option 2-Scenario
4a*
Option 2-Scenario 4b
Nitrogen
(million kg)
14
15
14
17
24
30
29
33
Phosphorus
(million kg)
35
45
42
48
42
54
52
59
Fecal
Conforms
(billion
colonies)
26
31
29
35
29
35
32
38
Fecal
Streptococci
(billion
colonies)
37
45
44
47
50
60
58
65
Sediments
(billion
kg)
0
0
0
0
26
35
34
38
Source: Estimation of National Surface Water Quality Benefits of Regulating Concentrated Animal Feeding Operations
(CAFOs) Using the National Water Pollution Control Assessment Model (NWPCAM), prepared for U.S. EPA Office
of Wastewater Management, prepared by Research Triangle Institute, August 2000.
* Proposed scenarios.
4-19
-------
Exhibit 4-10a
ESTIMATED ANNUAL AFO/CAFO METALS LOADINGS
UNDER BASELINE AND REGULATORY SCENARIOS*
Regulatory Scenario
Baseline Conditions
Option 1-Scenario 1
Option 1-Scenario 2/3
Option 1-Scenario 4a
Option 1-Scenario 4b
Option 2-Scenario 1
Option 2-Scenario 2/3**
Option 2-Scenario 4a**
Option 2-Scenario 4b
Zinc
(million
kg)
27
17
12
13
12
14
9
10
8
Copper
(thousand
kg)
1,495
949
727
796
677
796
516
600
444
Cadmium
(thousand
kg)
42
19
9
11
8
16
5
7
3
Nickel
(thousand
kg)
572
353
277
306
251
286
192
226
154
Lead
(thousand
kg)
1,143
748
570
613
545
634
405
457
366
Source : National AFO/CAFO Metals Edge-of-Field Loadings and Reductions from Agricultural Landuse
Cells to RF3 Reaches for AFO/CAFO Rulemaking Scenarios, table prepared for EPA by Research
Triangle Institute, August 2000.
* Metals loadings are estimated based on modeling of the overland transport of phosphorus. These
estimates do not play a role in the subsequent modeling of economic benefits.
** Proposed scenarios.
4-20
-------
Exhibit 4-10b
ESTIMATED ANNUAL METALS REMOVALS
UNDER REGULATORY SCENARIOS*
Regulatory Scenario
Option 1-Scenario 1
Option 1-Scenario 2/3
Option 1-Scenario 4a
Option 1-Scenario 4b
Option 2-Scenario 1
Option 2-Scenario 2/3**
Option 2-Scenario 4a* *
Option 2-Scenario 4b
Zinc
(million
kg)
10
15
14
15
13
18
17
19
Copper
(thousand
kg)
546
768
699
818
699
979
895
1,051
Cadmium
(thousand
kg)
23
33
31
34
26
37
35
39
Nickel
(thousand
kg)
219
295
266
321
286
380
346
418
Lead
(thousand
kg)
395
573
530
598
509
738
686
777
Source: National AFO/CAFO Metals Edge-of-Field Loadings and Reductions from Agricultural Landuse Cells to RF3
Reaches for AFO/CAFO Rulemaking Scenarios, table prepared for EPA by Research Triangle Institute, August
2000.
* Metals loadings are estimated based on modeling of the overland transport of phosphorus. These estimates do not
play a role in the subsequent modeling of economic benefits.
** Proposed scenarios.
4-21
-------
The discussion that follows summarizes the valuation approach. It begins by explaining
the process by which EPA relates the results of the surface water modeling effort to the ability
of a body of water to support a particular use. It then describes Carson and Mitchell's contingent
valuation study and its use in analyzing the economic benefits of alternate CAFO regulations.
4.6.1 Support of Designated Uses
EPA's approach to relating
surface water conditions to the ability
of a body of water to support a
particular designated use is based on
a water quality index - often referred
to as a water quality "ladder" - that
Resources for the Future initially
developed to support Carson and
Mitchell's contingent valuation
survey. As Exhibit 4-11 shows, the
ladder uses a scale that ranges from 0
to 10, with 0 representing the worst
possible water quality and 10
representing the best possible quality.
The low end of the scale represents
water quality so poor that it supports
no plant or animal life, and human
contact with it would be unsafe; the
high end of the scale represents water
safe enough to drink. Between these
extremes, the ladder depicts levels of
water quality sufficient to support
boating, fishing, or swimming.
Each step of the water quality
ladder is defined by measures of the
following parameters:
Exhibit 4-11
WATER QUALITY LADDER
Best Possible
Water Quality
10
Worst Possible
Water Quality
SWIMMABLE:
i Safe for swimming
FISHABLE:
Game fish like
bass can live in it
BOATABLE:
Okay for boating
dissolved oxygen content;
biological oxygen demand;
suspended sediment concentrations;
4-22
-------
pathogen counts; and
chlorophyll • concentrations.
In order for a body of water to be considered boatable, fishable or swimmable, it must satisfy the
minimum conditions consistent with that use for all modeled parameters. With the exception of
chlorophyll •, these minimum conditions are the same for all areas. The maximum chlorophyll
• concentration considered consistent with a particular use varies across four geographic regions,
which EPA defined specifically for this analysis. This approach takes into account the impact of
regional variation in factors like climate on the relationship between chlorophyll • concentrations
and the trophic state of a lake or reservoir.
Based on the framework described above, NWPCAM classifies each segment of each
modeled river or stream as swimmable, fishable, boatable, or non-supportive of any of these uses.
For each scenario analyzed, the model calculates the total stream-miles that support each
designated use.
4.6.2 Application of CV Study
The contingent valuation survey upon which this analysis relies examined households'
willingness to pay to maintain or achieve specified levels of water quality in freshwater lakes,
rivers and streams throughout the United States (Carson and Mitchell, 1993).37 The survey was
conducted in 1983 via in-person interviews at 61 sampling points nationwide, and employed a
national probability sample based on the 1980 Census. Respondents were presented with the
water quality ladder depicted in Exhibit 4-11 and asked to state how much they would be willing
to pay to maintain or achieve various levels of water quality throughout the country. In eliciting
responses, the survey used a payment card showing the amounts average households were then
currently paying in taxes or higher prices for certain publicly provided goods (e.g., national
defense); respondents were then asked their willingness to pay for a given water quality change.
The survey respondents were told that improvements in water quality would be paid for in higher
product prices and higher taxes.
37 The scope of the survey excluded the Great Lakes.
4-23
-------
Exhibit 4-12 presents the results of the survey, adjusted to account for inflation and
changes in real income between 1983 and 1999.38 These values represent "best estimates" of
mean annual household willingness to pay (WTP) for the specified water quality improvement.
Exhibit 4-12
INDIVIDUAL HOUSEHOLD WILLINGNESS TO PAY
FOR WATER QUALITY IMPROVEMENTS
(1999 $)
Water Quality Improvement
Swimmable: WTP to raise all sub-swimmable water quality to
swimmable
Fishable: WTP to raise all sub-fishable water quality to fishable
Beatable: WTP to maintain beatable water quality
Total WTP
$634
$429
$245
Incremental
WTP
$205
$184
$245
Source: Carson and Mitchell, 1993. The values originally reported have been adjusted to account for inflation
and changes in real income between 1983 and 1999.
Applying the willingness to pay estimates presented above to analyze the benefits of
alternate CAFO regulations requires consideration of how households' willingness to pay for
water quality improvements is likely to vary with the extent and location of the resources affected.
All else equal, people are likely to value an action that improves water quality along a ten-mile
stretch of river more highly than they would value an action that improves only a one-mile stretch.
Similarly, people are likely to place greater value on improving the quality of water resources that
are nearer to them. This is simply because less time and expense is typically required to reach
nearer resources; as a result, these resources generally provide lower cost and more frequent
opportunities for recreation and enjoyment. This assumption is supported by the results of the
Carson and Mitchell survey, which asked respondents to apportion their willingness to pay values
between improving the quality of local waters - where local waters were defined as those in each
respondent's own state - and improving the quality of non-local waters (i.e., those located out-of-
state). On average, respondents allocated two-thirds of their values to achieving water quality
goals in-state, and one-third to achieving those goals in the remainder of the nation.
38 EPA employed the Consumer Price Index to adjust 1983 values to 1999 values. In
addition, the adjustment to 1999 values takes into account the increase in real per capita disposable
income over the period of interest. The adjustment for changes in real income is consistent with the
survey's results, which found that respondents' willingness to pay for water quality improvements
increased in almost direct proportion to household income.
4-24
-------
To reflect the considerations noted above, the analysis of the benefits of alternate CAFO
regulations examines water quality improvements on a state-by-state basis and separately
calculates the benefits of in-state and out-of-state improvements, assuming that households will
allocate two-thirds of their willingness to pay values to the improvement of in-state waters. In
addition, the analysis takes into account the extent of each scenario's estimated impacts (i.e., the
number of stream-miles that improve from non-supportive or boatable to fishable, or from non-
supportive, boatable or fishable to swimmable) by scaling household willingness to pay for a
given improvement in the quality of the nation's waters by the proportion of total stream-miles
in-state or out-of-state that are projected to make the improvement. For purposes of these
calculations, the analysis relies on estimates of the 1999 population and the average number of
individuals per household (2.62) as reported in the Statistical Abstract of the United States.
Appendix 4-A provides a detailed summary of the calculations employed.
It is important to note that the valuation process described above does not address a
number of potential improvements in water quality. For example, the analysis assigns no
economic value to improvements in the quality of waters already classified as swimmable under
the baseline scenario; the Carson and Mitchell contingent valuation survey did not examine this
area, and thus provides no basis for valuing such changes. In addition, the analysis does not value
the improvement of non-boatable waters to boatable condition. Although the survey asked
respondents how much they would be willing to pay to avoid a drop in water quality to non-
boatable conditions, the definition of non-boatable water quality - "water containing raw sewage,
with strong odors, floating garbage and pathogens that would cause illness through human
contact" - was so extreme that the resulting willingness to pay values may be overstated.
Consequently, EPA has chosen not to employ Carson and Mitchell's estimates of willingness to
pay to maintain boatable water quality in this analysis. EPA is reassessing the boatable level in
the water quality ladder to be more consistent with the language of the Carson and Mitchell study.
Overall, the Agency believes that the benefits calculated here are conservative because they
reflect only swimming and fishing improvements, not boating improvements or improvements
in the quality of waters already considered swimmable.
4.6.3 Estimated Benefits
Exhibit 4-13 presents NWPCAM's estimates of the annual economic benefits associated
with each regulatory scenario. As the table indicates, the estimates range from approximately $5
million per year under Option 1-Scenario 1 to $145 million per year under Option 2-Scenario 4b.
EPA estimates that the annual benefits of the proposed scenarios range from approximately
$108.5 million per year to $127.1 million per year. Roughly 80 percent of these benefits are
derived from improving previously non-fishable waters to fishable status. The remaining benefits
are attributed to improving a smaller number of stream miles to swimmable condition.
4-25
-------
Exhibit 4-13
ANNUAL ECONOMIC BENEFIT OF ESTIMATED
IMPROVEMENTS IN SURFACE WATER QUALITY
(1999 $, millions)
Regulatory Scenario
Option 1 -Scenario 1
Option 1 -Scenario 2/3
Option 1 -Scenario 4a
Option 1 -Scenario 4b
Option 2-Scenario 1
Option 2-Scenario 2/3*
Option 2-Scenario 4a*
Option 2-Scenario 4b
Waters
Improved to
Fishable
$2.8
$3.2
$3.1
$3.8
$71.0
$102.4
$84.0
$115.5
Waters
Improved to
Swimmable
$2.1
$3.1
$2.4
$3.4
$16.7
$24.7
$24.5
$29.5
Total Benefits
$4.9
$6.3
$5.5
$7.2
$87.6
$127.1
$108.5
$145.0
* Proposed scenarios.
4.7 REFERENCES
Carson, Richard T. and Robert Cameron Mitchell. 1993. "The Value of Clean Water: The
Public's Willingness to Pay for Boatable, Fishable, and Swimmable Water Quality."
Water Resources Research, Vol. 29, No. 7.
Ham, J.M. and T.M. DeSutter. 1999. "Seepage Losses and Nitrogen Export from Swine-Waste
Lagoons: A Water Balance Study." Journal of Environmental Quality. 28:1090-1099.
Kellogg, Robert L., Charles Lander, David Moffitt, and Noel Gollehon. 2000. Manure Nutrients
Relative to the Capacity of'Cropland andPastureland to Assimilate Nutrients: Spatial
and Temporal Trends for the U.S. Forthcoming. U.S. Department of Agriculture,
National Resources Conservation Service. Washington, DC.
Lander, C.H., D. Moffitt, and K. Alt. 1998. Nutrients Available from Livestock Manure
Relative to Crop Growth Requirements. U.S. Department of Agriculture. Natural
Resources Conservation Service. Washington, DC. February.
NPPC (National Pork Producers Council). 1998. Pork facts 1998/1999. Des Moines, IA:
National Pork Producers Council. http://www.nppc.org/PorkFacts/pfindex.html
4-26
-------
Smith, Richard A., Gregory E. Schwarz, and Richard B. Alexander. 1997. "Regional
Interpretation of Water-Quality Monitoring Data." Water Resources Research, Vol. 33,
no. 12. December.
Sobecki, T.M., and M. Clipper. 1999. Identification of Acreage of U.S. Agricultural Land with
a Significant Potential for Siting of Animal Waste Facilities and Associated Limitations
from Potential of Groundwater Contamination, Draft. U.S. Environmental Protection
Agency, Office of Water. December.
USDA/APfflS (Animal and Plant Health Inspection Service). 1995. Swine'95. Part III and Part
I: Reference of 1995 Swine Management Practices. October.
USDA/NASS (U.S. Department of Agriculture, National Agricultural Statistics Service). 1999a.
7997 Census of Agriculture. Volume I, Geographic Area Series Part 51. Washington,
DC. March.
USDA/NASS (U.S. Department of Agriculture, National Agricultural Statistics Service). 1999b.
Cattle Final Estimates 1994-1998. January.
USEPA(U.S. Environmental Protection Agency). 2000a. Technical Development Document for
Proposed Effluent Limitations Guidelines for Animal Feeding Operations. Office of
Water.
USEPA (U.S. Environmental Protection Agency). 2000b. Estimation of National Surface Water
Quality Benefits of Regulating Concentrated Animal Feeding Operations (CAFOs) Using
the National Water Pollution Control Assessment Model (NWPCAM). Office of Water.
July.
4-27
-------
Appendix 4-A
NWPCAM CALCULATION OF THE ECONOMIC BENEFITS
OF IMPROVED SURFACE WATER QUALITY
Definitions
N = national benefits of estimated improvements in water quality
Sj = total benefits of estimated improvements in water quality for residents of state "j"
B(1 j) = benefits of in-state improvements in water quality for residents of state "j"
B(nJ) = benefits of out-of-state improvements in water quality for residents of state "j"
Mj = total stream-miles in state "j"
Mn = total stream-miles outside state "j"
M (xj} = stream-miles in state "j" that achieve water quality improvement "x"
M (x n) = stream-miles outside state "j" that achieve water quality improvement "x"
Hj = total households in state "j"
WTPX = average household willingness to pay for water quality improvement "x"
Calculations
N=« S:
SJ~B(U)+B(nJ)
B(1J)=' (M(xJ)/Mj)(Hj)(WTPx)(2/3)
x
B(nJ)= ' (M(Xin)/Mn)(Hj)(WTPx)(l/3)
4A-1
-------
REDUCED INCIDENCE OF FISH KILLS CHAPTER 5
5.1 INTRODUCTION
Episodic fish kills resulting from manure runoff, spills, and other discharges from AFOs
remain a serious problem in the United States. As described in Chapter 2, large releases of
nutrients, pathogens, and solids from AFOs can cause sudden, extensive kill events.39 In less
dramatic cases, nutrients contained in runoff from AFOs can trigger increases in algae
growth—often called algae blooms—that reduce concentrations of dissolved oxygen in water and
can eventually cause fish to die.40
In addition to killing and harming fish directly, pollution from AFOs can affect other
aquatic organisms that in turn harm fish. In particular, the Eastern Shore of the United States has
been plagued with problems related to Pfiesteria, a dinoflagellate algae that, under certain
circumstances, can transform into a toxin that attacks fish, breaking down their skin tissue and
leaving lesions or large, gaping holes that often result in death. The transformation ofPfiesteria
to its toxic form is believed to be the result of high levels of nutrients in water. Fish kills related
to Pftesteria in North Carolina's Neuse River have been blamed on waste spills and runoff from
the state's booming hog industry.
39 For example, in 1998, the release of manure into the West Branch of Wisconsin's
Pecatonica River resulted in a complete kill of smallmouth bass, catfish, forage fish, and all but the
hardiest insects in a 13-mile reach.
40 For example, in 1996, the gradual runoff of manure into Atkins Lake, a shallow lake in
Arkansas, resulted in a heavy algae bloom that depleted the lake of oxygen, killing many fish.
5-1
-------
This chapter examines the damages attributable to AFO-related fish kills and estimates
the economic benefits that revised CAFO standards would provide in reducing such incidents.
As explained below, the analysis employs state data on historical fish kill events, combined with
predicted reductions in the frequency of such events under alternate regulatory scenarios, to
estimate the decrease that would occur in the number of fish killed annually in AFO-induced
incidents. It then employs data on average fish replacement costs to develop a conservative
estimate of the economic benefits associated with the predicted reduction in fish kill incidents.
5.2 ANALYTIC APPROACH
5.2.1 Data Sources and Limitations
Data on fish kill incidents are limited, with no consistent collection and reporting
requirements and no national repository offish kill data. States are not required to report fish
kills to EPA. As a result, EPA does not maintain a comprehensive database detailing the
frequency or severity offish kill events.
Despite the lack of EPA reporting requirements, many states do record information on fish
kills. For purposes of this analysis, EPA has compiled a database offish kill events in 19 states.
This database incorporates a range of information on each incident. Exhibit 5-1 lists the 19 states
included in the database, the years for which data from each state were obtained, and the average
number offish kill events reported by each state annually.41
As Exhibit 5-1 indicates, the data upon which this analysis relies are not comprehensive.
The fish kill database excludes 31 states, including several, such as Oklahoma, that host a
relatively large number of AFOs. The period of time for which data were obtained also varies
from state to state; the information collected from some states, such as Missouri, covers nearly
two decades, while that collected from others, such as West Virginia, covers only a few years.
In addition, even in the states and years for which data were collected, it is likely that some fish
kill events remain unreported, particularly if they occurred in remote areas.42 These data gaps
introduce considerable uncertainty into the analysis.
41 EPA's database incorporates records on fish kills obtained from the Natural Resources
Defense Council (NRDC) and the Izaak Walton League (IWL).
42 For instance, in 1995 the Raleigh News & Observer reported a 1991 manure spill incident
in the North Carolina town of Magnolia that neither the town nor the responsible farm reported to
state water quality officials.
5-2
-------
Exhibit 5-1
FISH KILL EVENT DATA OBTAINED BY EPA
State
Arkansas
Illinois
Indiana
Iowa
Kansas
Kentucky
Minnesota
Mississippi
Missouri
Montana
Nebraska
New Mexico
New York
North Carolina
Ohio
South Carolina
Texas
West Virginia
Wisconsin
Years for which
Data
were Obtained
1995-1999
1987-1999
1994-1999
1981-1998
1990-1999
1995-1998
1981-1991
1990-1998
1980-1999
1994-1998
1991-1998
1995-1998
1984-1996
1994-1998
1995-1998
1995-1998
1990-1998
1995-1997
1988-1998
Total
Average Annual
Number of Recorded
Events
8.6
14.0
27.2
26.3
15.7
15.5
23.9
18.6
125.3
1.8
22.1
4.8
18.0
41.2
20.3
5.5
114.7
6.0
6.4
515.9
In addition to the data gaps cited above, the analysis is limited by inconsistencies in the
information collected in state fish kill reports. Some states appear to have established consistent
guidelines for investigating a kill, which often include reporting the number of stream miles or
lake acres affected, estimating the number offish killed, describing the exact location of the kill,
identifying the source of the pollutants suspected to have caused the kill, and obtaining water
5-3
-------
quality samples for testing. Other states appear to gather information on an ad hoc basis. In
addition, the data present a number of anomalies or other limitations. For example, 25 percent
of the records included in EPA's database give no estimate of the number offish killed or provide
only a qualitative description of the incident's magnitude. Another 13 percent of the records
indicate that the number offish killed in the event was zero.43 In addition, most reports do not
indicate the type(s) offish killed.
Despite the apparent limitations of these data, they are useful for purposes of this analysis.
EPA's database is the most comprehensive source of information on fish kill events currently
available, and in most instances characterizes the source of the pollutants that caused individual
fish kill events. Thus, EPA can apply these data to characterize a baseline of kill events
potentially attributable to pollution from AFOs.
5.2.2 Predicted Change in Fish Kills Under Alternate CAFO Regulations
To estimate the potential benefits of alternate CAFO regulations in reducing fish kill
incidents, EPA's analysis must first assess the current — or baseline — number of AFO-related
fish kills. It must then determine the impact of alternate regulatory scenarios in reducing these
incidents. EPA's approach to this analysis is described below.
5.2.2.1 Baseline Scenario
The EPA database records fish kill events attributable to a wide range of sources and
causes, the classification of which varies from state to state. Exhibit 5-2 lists the sources and
causes that are potentially associated with AFOs, and indicates EPA's assessment of the strength
of the association; i.e., whether the Agency considers the link to AFOs to be definite or possible.
Based on this assessment of sources and causes, EPA reviewed the data on each reported event,
identifying fish kills positively or possibly attributable to pollution from AFOs. Exhibit 5-3
presents the results of this evaluation. As the exhibit indicates, EPA's database lists 589 fish kill
events that are positively or possibly attributable to pollution from AFOs. These incidents killed
a reported total of approximately 4.2 million fish. Based on these data, EPA estimates that in the
states evaluated, incidents positively attributable to pollution from AFOs kill an average of 351
43 This may be due to a variety of circumstances. In some cases, the report may accurately
indicate an event in which contamination occurred (such as a manure spill or municipal waste
release) but no fish were killed. In other cases, a record may indicate zero fish killed simply because
investigators were unable to develop a count (e.g., because the number killed was too great to count,
or because the investigation was conducted too late to determine the number killed).
5-4
-------
thousand fish per year. In addition, incidents possibly attributable to pollution from AFOs kill,
on average, another 40 thousand fish annually. In total, the database indicates that in the 19
states analyzed, pollution from AFOs may kill an average of 391 thousand fish each year.44
Exhibit 5-2
SOURCES AND CAUSES OF FISH KILL EVENTS POTENTIALLY
RELATED TO ANIMAL FEEDING OPERATIONS
Source
animal feeding/waste
operations
agriculture
agriculture point source
algae related
Relation to
AFOs
Definite
Possible
Possible
Possible
Cause
ammonia toxicity
lagoon breaks
manure
nutrients
algae blooms
dissolved oxygen
employee error
equipment failures
fertilizer
nonpoint source runoff
spills
weather
Relation to
AFOs
Definite
Definite
Definite
Definite
Possible
Possible
Possible
Possible
Possible
Possible
Possible
Possible
44 EPA estimates the average number offish killed annually in the 19 states of record by
dividing the total number offish killed in each state by the number of years for which data from the
state are reported. EPA then sums the state averages to obtain the annual average for all 19 states.
5-5
-------
Exhibit 5-3
NUMBER OF FISH KILL EVENTS POTENTIALLY
RELATED TO ANIMAL FEEDING OPERATIONS
Cause
Ammonia Toxicity
Ammonia Toxicity/
Dissolved Oxygen
Dissolved Oxygen
Employee Error/Manure
Equipment Failure
Equipment Failure/Manure
Equipment
Failure/Fertilizer
Fertilizer
Lagoon Breaks
Manure
Manure/Dissolved Oxygen
Manure/Weather
Algal Toxins/ Algal
Oxygen Deficiency
Nonpoint Source Runoff
Nutrient
Spills
Weather
Other / Unknown
Total:
Number of Occurrences for Each Source
Animal
Feeding
and Waste
Operations
8
96
-
3
-
-
-
-
5
74
-
-
-
1
-
3
-
1
191
Agricultu
re
6
3
26
2
1
3
1
24
-
158
-
1
6
5
1
-
-
2
239
Agri-
cultur
e
Point
Source
3
4
1
-
-
-
-
-
-
1
2
-
-
-
-
1
3
-
15
Algae-
Relate
d
-
_
33
-
-
-
-
-
-
-
-
-
1
-
-
-
-
-
34
Unknow
n Source
-
-
-
-
-
-
-
-
-
110
-
-
-
-
-
-
-
-
110
Total
17
103
60
5
1
o
6
i
24
5
343
2
1
7
6
1
4
3
3
589
Note: Numbers in bold indicate fish kill events considered to be definitely induced by AFOs.
5-6
-------
5.2.2.2 Regulatory Scenarios
Due to time and resource constraints, EPA has not conducted a detailed analysis of the
impact of alternate CAFO standards on the frequency or severity offish kill events. It is likely,
however, that the implementation of new regulations would have a number of beneficial effects.
For example, because more AFOs would be subject to regulation as CAFOs, the number offish
kill incidents caused by lagoon breaks and similar catastrophic events would likely diminish. In
addition, the improvements in manure management practices required under the new regulations
would likely reduce the chronic discharge of nutrients to the nation's waters, and thus reduce the
number offish killed as a result of severe eutrophi cation.
In lieu of more detailed modeling, EPA has attempted to develop a reasonable estimate of
the impact of alternate CAFO standards on fish kills. The analysis begins with EPA's estimate of
the number offish killed annually by releases from AFOs.45 EPA multiplies this figure by the
anticipated percentage reduction in nutrient loadings from the animal feeding operations modeled
by NWPCAM (see Chapter 4). The resulting value, for each regulatory scenario, represents an
estimate of the reduction in the number offish killed annually by releases from AFOs.
Because the relationship between nutrient loadings and fish kill events is complex, this
approach provides only a rough approximation of the beneficial impacts of alternate regulations.
To reflect the underlying uncertainty, the analysis employs two different scaling factors:
the percentage reduction in phosphorus loadings; and
the percentage reduction in nitrogen loadings.
Exhibit 5-4 summarizes the estimated percentage reduction in nitrogen and phosphorus loadings
under each regulatory scenario. The values reported are those estimated by NWPCAM for the
full RF3 set of rivers and streams. The analysis uses these values, rather than those reported for
the RF3 Lite subset, in order to reflect changes in loadings to small as well as large rivers and
streams.46
45 For purposes of this analysis, EPA considers all fish kill events identified as definitely or
possibly attributable to AFOs.
46 Chapter 4 provides additional detail on the RF3 and RF3 Lite datasets.
5-7
-------
Exhibit 5-4
REGULATORY SCENARIO SCALING FACTORS
Regulatory Scenario
Option I/Scenario 1
Option I/Scenario 2/3
Option I/Scenario 4a
Option I/Scenario 4b
Option 2/Scenario 1
Option 2/Scenario 2/3*
Option 2/Scenario 4a*
Option 2/Scenario 4b
Percent Nitrogen
Reduction
18.10
23.02
21.82
25.66
34.55
43.00
41.48
47.24
Percent Phosphorus
Reduction
45.53
59.79
56.51
63.00
55.22
72.06
68.81
77.40
* Proposed scenarios
Based on the methods described above, EPA estimates the anticipated reduction in fish kills
under each of the regulatory scenarios. Exhibit 5-5 presents the results. As the exhibit shows, EPA
estimates that under the proposed scenarios, the reduction in fish killed annually would range from
162 thousand to 282 thousand.
Exhibit 5-5
ESTIMATED REDUCTION IN THE NUMBER OF FISH KILLED
ANNUALLY DUE TO RELEASE OF POLLUTANTS FROM AFOs
(thousands)
Regulatory Scenario
Option I/Scenario 1
Option I/Scenario 2/3
Option I/Scenario 4a
Option I/Scenario 4b
Option 2/Scenario 1
Option 2/Scenario 2/3*
Option 2/Scenario 4a*
Option 2/Scenario 4b
Scaling Factor
Nitrogen Reduction
71
90
85
100
135
168
162
185
Phosphorus Reduction
178
234
221
247
216
282
269
303
*Proposed scenarios
5.2.3 Valuation of Predicted Reduction in Fish Kills
The economic damages that stem from natural resource injuries like fish kills include the
costs of restoring the resource to its prior state, any interim lost use values (e.g., the economic value
of lost fishing days from the time the damage occurs until fish stocks are restored), and any interim
5-8
-------
lost non-use values. Unfortunately, estimating these values for a large number of heterogeneous
fish kill events nationwide is infeasible without a significant investment of analytic resources.
Determining full habitat restoration costs requires a case-by-case assessment of the nature of the
injury and the restoration options available, while estimating interim lost non-use values requires
the use of stated preference techniques to explore people's willingness to pay to avoid temporary
depletions of fish stocks and associated damage to fish habitat. The economics literature does
provide estimates of potential lost use values— e.g., willingness to pay for another day of fishing
or willingness to pay for an additional fish caught—that could, theoretically, be applied to the
analysis using a benefits transfer approach. This assessment, however, would need to make general
assumptions about a number of highly variable site-specific factors, such as the duration of the
reduction in fish stocks, the effect of this reduction on recreational fishing activity in the affected
areas, and the availability and characteristics of alternative fishing areas. Thus, an evaluation of
interim lost use values would be subject to considerable uncertainty.
Because of the difficulties cited above, this analysis takes a simplified approach, estimating
the economic benefits of reducing the frequency offish kills based on one component of resource
restoration costs: the replacement cost of the fish. Specifically, the analysis employs replacement
cost estimates presented in a report developed by the American Fisheries Society (AFS, 1990).
These replacement values incorporate the cost of raising fish at a hatchery, transporting them, and
placing them in the water. As such, they provide a conservative estimate of the economic benefits
of reducing the incidence offish kills.47
The American Fisheries Society report provides replacement cost estimates for a variety
of fish species and size categories. Unfortunately, the available data on fish kills do not always
indicate the species offish affected, and generally do not report mortality by size offish. In light
of these limitations, EPA applies a general fish replacement cost estimate, derived by selecting
species known to have been killed in incidents related to AFOs and averaging reported replacement
costs for these species across all size classes. The resulting average replacement cost employed
in the analysis equals $1.31 per fish ($1999).48 To value the benefits of each regulatory scenario,
the analysis simply multiplies this average replacement cost by the corresponding estimated
reduction in the number offish killed each year.49
47 The analysis employs fish replacement costs as a proxy measure for valuing anticipated
reductions in fish kill incidents. The approach does not presume that all fish killed would necessarily
be restocked.
48 To adjust replacement costs to 1999 dollars, EPA applies the Gross Domestic Product
deflator.
49 To bound the potential uncertainty associated with fish replacement costs, the analysis also
develops benefits estimates based on the averages of the minimum and maximum replacement costs
reported for each species. Appendix 5-A presents the results of this analysis.
5-9
-------
5.3 RESULTS
Exhibit 5-6 presents estimates of the annual benefits attributable to the reduced incidence
offish kills under each of the regulatory scenarios. As the exhibit indicates, these benefits range
from $93 thousand to $397 thousand, depending upon the regulatory option considered and the
scaling factor employed. Annual benefits under the proposed scenarios are estimated to range from
$213 thousand to $369 thousand.
Exhibit 5-6
ESTIMATED ANNUAL BENEFITS
ATTRIBUTED TO REDUCTION IN FISH KILLS
($1999, thousands)
Regulatory Scenario
Option I/Scenario 1
Option I/Scenario 2/3
Option I/Scenario 4a
Option I/Scenario 4b
Option 2/Scenario 1
Option 2/Scenario 2/3*
Option 2/Scenario 4a*
Option 2/Scenario 4b
Scaling Factor
Nitrogen Reduction
$93
$118
$112
$132
$177
$220
$213
$242
Phosphorus Reduction
$233
$306
$290
$323
$283
$369
$353
$397
*Proposed scenarios
5.4 LIMITATIONS AND CAVEATS
EPA's analysis of the benefits of alternate CAFO regulations in reducing fish kills is subject
to numerous data gaps and uncertainties. In the face of these uncertainties, the analysis employs
a number of simplifying assumptions. The major limitations of the analysis are summarized below.
• The scope of the analysis is limited to 19 states. The data available from
these states may not include all fish kill events, and the data on reported
incidents often fail to include estimates of the number of fish killed.
Therefore, EPA's baseline estimate is likely to understate the number offish
kill events and the total number of fish killed nationwide each year in
incidents related to pollution from AFOs.
5-10
-------
EPA has not undertaken a detailed analysis of the impact of alternate
regulatory scenarios on the incidence of fish kills. In lieu of a detailed
analysis, EPA assumes that fish kills attributable to releases of pollution
from AFOs will be reduced in proportion to estimated reductions in
loadings of nutrients from AFOs. The direction and magnitude of bias
associated with these assumptions is unknown.
To value estimated reductions in fish kill incidents, the analysis applies an
estimate of average fish replacement costs. Because this proxy measure
ignores other aspects of the economic damages associated with fish kills
(i.e., habitat restoration costs, interim lost use values, and interim lost non-
use values), it likely understates the economic benefit of reducing fish kill
incidents.
5.5 REFERENCES
AFS. 1993. American Fisheries Society Socioeconomics Section, Sourcebookfor Investigation
and Valuation of Fish Kills.
Griffiths, Charles and Cynthia Morgan. 2000. "Benefits of Avoiding Fish Kills by Regulating
Livestock Waste," National Center for Environmental Economics, U.S. Environmental
Protection Agency.
Warrick, Joby and Pat Smith. 2000. The News & Observer, "New studies show that lagoons are
leaking: Groundwater, rivers affected by waste," Sunday, February 19,1995, obtained from:
http://www.nando.net/sproject/hogs/lwater.html, June 28.
5-11
-------
Appendix 5-A
CALCULATION OF ANNUAL BENEFITS USING
MINIMUM AND MAXIMUM FISH REPLACEMENT VALUES
Replacement values for a fish species can vary significantly depending on the size class of
the fish. This appendix presents the results of the benefits assessment using averages of the
minimum and maximum replacement values reported for each fish species. The average minimum
and maximum replacement costs equal $0.28 and $2.37 ($1999), respectively. Exhibit 5A-1
presents annual benefit estimates based upon the minimum fish replacement cost; annual benefits
for the proposed scenarios in this case range between $45 and $79 thousand. Exhibit 5 A-2 presents
annual benefit estimates based on the maximum replacement cost; in this case, annual benefits
under the proposed scenarios range from $385 to $668 thousand.
Exhibit 5A-1
ESTIMATED ANNUAL BENEFITS
ATTRIBUTED TO REDUCTION IN FISH KILLS:
MINIMUM REPLACEMENT COST
($1999, thousands)
Regulatory Scenario
Option I/Scenario 1
Option I/Scenario 2/3
Option I/Scenario 4a
Option I/Scenario 4b
Option 2/Scenario 1
Option 2/Scenario 2/3*
Option 2/Scenario 4a*
Option 2/Scenario 4b
Scaling Factor
Nitrogen Reduction
$20
$25
$24
$28
$38
$47
$45
$52
Phosphorus Reduction
$50
$66
$62
$69
$60
$79
$75
$85
*Proposed scenarios
5A-1
-------
Exhibit 5A-2
ESTIMATED ANNUAL BENEFITS
ATTRIBUTED TO REDUCTION IN FISH KILLS:
MAXIMUM REPLACEMENT COST
($1999, thousands)
Regulatory Scenario
Option I/Scenario 1
Option I/Scenario 2/3
Option I/Scenario 4a
Option I/Scenario 4b
Option 2/Scenario 1
Option 2/Scenario 2/3*
Option 2/Scenario 4a*
Option 2/Scenario 4b
Scaling Factor
Nitrogen Reduction
$168
$213
$202
$238
$320
$399
$385
$438
Phosphorus Reduction
$422
$554
$524
$584
$512
$668
$638
$718
*Proposed scenarios
5A-2
-------
IMPROVED COMMERCIAL SHELLFISHING CHAPTER 6
6.1 INTRODUCTION
The National Oceanic and Atmospheric Administration (NOAA) has identified pathogen
contamination of U.S. coastal waters as a leading cause of government restrictions on commercial
shellfish harvesting. Among the sources of pollution that contribute to such contamination are
animal feeding operations (AFOs) and runoff from agricultural lands. This chapter estimates the
impact of pollution from AFOs on commercial access to shellfish growing waters, the resulting
impact on commercial shellfish harvests, and the potential increase in harvests that would result
under alternate approaches to regulating the discharge of pollutants from CAFOs. It then uses
available estimates of consumer demand for shellfish to calculate the economic benefits associated
with the predicted increase in commercial shellfish harvests under each regulatory scenario.
6.2 ANALYTIC APPROACH
6.2.1 Data on Shellfish Harvest Restrictions Attributed to AFOs
EPA's analysis of the impact of pollution from AFOs on shellfish harvests is based on
information from The 1995 National Shellfish Register of Classified Growing Waters (NOAA,
1997) and related databases. NOAA produces the Register, which is published every five years,
in cooperation with the nation's shellfish-producing states, federal agencies such as the U.S. Food
and Drug Administration (FDA), and the Interstate Shellfish Sanitation Conference (ISSC). Its
purpose is to summarize the status of shellfish-growing waters under the National Shellfish
Sanitation Program (NSSP), which ISSC administers. The NSSP establishes comprehensive
guidelines to regulate the commercial harvesting, processing, and shipment of shellfish. These
guidelines include the measurement of fecal coliform concentrations as an indicator of pollution
in shellfish-growing waters. Based in large part upon these measurements, shellfish-growing areas
are designated as approved, conditionally approved, restricted, conditionally restricted, prohibited,
or unclassified, and subjected to appropriate harvest and processing standards. Exhibit 6-1
describes these standards for each designation.
6-1
-------
Exhibit 6-1
NSSP STANDARDS FOR CLASSIFIED SHELLFISH GROWING WATERS
Classifi
cation
Description
Standard1
Approv
ed Waters
Growing waters from which shellfish may be
harvested for direct marketing.
Conditi
onally Approved
Waters
Growing waters meeting the approved
classification standards under predictable conditions.
These waters are open to harvest when water quality
standards are met. At all other times these waters are
closed.
MPN may not
exceed 14 per 100 ml, and
not more than 10 percent
of the samples may
exceed an MPN of 43 per
100 ml for a 5-tube
decimal dilution test.
Restrict
ed Waters
Growing waters from which shellfish may be
harvested only if they are relayed or depurated before
direct marketing.2
Conditi
onally Restricted
Waters
Growing waters that do not meet the criteria for
restricted waters if subjected to intermittent
microbiological pollution, but may be harvested if
shellfish are subjected to a suitable purification process.
MPN may not
exceed 88 per 100 ml, and
not more than 10 percent
of the samples may
exceed an MPN of 260
per 100 ml for a 5-tube
decimal dilution test.
Prohibit
ed Waters
Growing waters from which shellfish may not be
harvested for marketing under any conditions.
NA
Unclassi
fied Waters
Growing waters that are part of a state's shellfish
program but are inactive (i.e., there is no harvesting) and
unmonitored.
NA
Source: National Oceanic and Atmospheric Administration, The 1995 National Shellfish Register of
Classified Growing Waters, obtained from: http://seaserver.nos.noaa.gov/projects/95register/,
11 June 2000.
Notes:
MPN = fecal conform most probable number (median or geometric mean).
The process of relaying shellfish refers to the transfer of shellfish from restricted waters to
approved waters for natural biological cleansing using the ambient environment as a treatment
system, usually for a minimum of 14 days before harvest. Depuration is the process of removing
impurities by placing the contaminated shellfish in clean water for a period of time.
6-2
-------
The 1995 Shellfish Register provides information on 21.4 million acres of estuarine and non-
estuarine commercial shellfish-growing waters as of January 1, 1995. A companion CD contains
a GIS-based database of the location of all 4,320 shellfish growing areas in 21 coastal states, the
acreage of each growing area, and the species harvested.50 These species are classified into 13
categories of clams, four categories of oysters, six categories of mussels, and two categories of
scallops. In most cases, each category represents a unique species (e.g., Blue Mussel (Mytilus
edulis)), but in some instances a category may include two or more species (e.g., Other Mussels
(Mytilus galloprovincialis and Mytilus edulisj). The types of species harvested vary geographically,
with large differences between the East and West Coasts.
In addition to the data described above, the shellfish database notes for each growing area
any harvest limitations imposed and the known or possible source(s) of pollutants causing any
impairment. The list of pollutant sources includes both "Animal Feedlots" and "Agriculture
Runoff." Sources of impairment are further classified as actual or potential contributors. If a source
is listed as an actual contributor, its significance as a cause of impairment is rated as high, medium,
or low. Exhibit 6-2 shows the acreage of shellfish-growing waters that are potentially or known to
be impaired by pollution from AFOs and/or agricultural runoff. As the exhibit indicates, AFOs
and/or agricultural runoff are known or potential contributors to the impairment of more than 1.6
million acres of shellfish-growing waters.
50 The Shellfish Register includes data for the following states: Alabama, California,
Connecticut, Delaware, Florida, Georgia, Louisiana, Massachusetts, Maryland, Maine, Mississippi,
North Carolina, New Hampshire, New Jersey, New York, Oregon, Rhode Island, South Carolina,
Texas, Virginia, and Washington.
6-3
-------
Exhibit 6-2
SHELLFISH HARVEST LIMITATIONS BY REGION
Region
North Atlantic (MA, ME, NH)
Middle Atlantic (CT, DE, MD, NJ, NY, RI, VA)
South Atlantic (FL, GA, NC, SC)
Gulf of Mexico (AL, LA, MS, TX)
Pacific (CA, OR, WA)
Total
Approved
Acres
2,920,575
4,969,680
3,505,729
3,238,431
206,574
14,840,989
Harvest-
Limited
Acres
714,191
973,715
1,751,844
3,067,730
214,494
6,721,975
Harvest-Limited
Acres with
Impacts from
AFOs and/or
Agricultural
Runoff
33,626
100,284
660,679
718,828
96,296
1,609,713
Discrepancies between reported totals and sum of regional totals are due to rounding.
Source: U.S. National Oceanic and Atmospheric Administration, The 1995 National Shellfish Register of
Classified Growing Waters, U.S. Department of Commerce, Silver Spring, MD, August 1997.
6.2.2 Estimated Impact on Shellfish Harvests
As a causal factor in the imposition of government restrictions or prohibitions on shellfish
harvesting, pollution from AFOs likely serves to reduce shellfish landings below levels that would
otherwise be realized. To evaluate the potential beneficial effects of new CAFO regulations, EPA's
analysis begins by estimating the adverse impacts currently attributable to pollution from AFOs. The
approach to this analysis involves the following steps.
• Step 1: characterize current, or baseline, annual shellfish landings.
• Step 2: estimate the area of shellfish-growing waters from which current
landings are harvested.
Step 3: calculate the average annual per-acre yield of shellfish from
harvested waters.
• Step 4: estimate the area of shellfish-growing waters that are currently
unharvested as a result of pollution from AFOs.
6-4
-------
Step 5: estimate the foregone harvest, i.e., the potential annual harvest of
shellfish from waters that are currently unharvested as a result of pollution
from AFOs.
Each of these steps is described in greater detail below.
6.2.2.1 Baseline Annual Shellfish Landings
To characterize the baseline quantity (Q0) of shellfish harvested in each coastal state, the
analysis relies on data collected by NOAA's National Marine Fisheries Service (NMFS), which
reports commercial fishing harvests by state, year, and species (NMFS, 2000). NMFS maintains
complete commercial harvest data on various species of clams, mussels, oysters and scallops for each
state. The data consist of total pounds harvested and total ex-vessel revenues for harvested species.
The data are provided as state-wide totals only and do not disaggregate harvest quantities between
shellfish growing areas within each state. For the purpose of this analysis, EPA obtained shellfish
harvest data by species and state for the five most recent years available: 1994 through 1998. The
analysis employs the mean of the reported annual values for each species and state to characterize
shellfish harvests under baseline conditions.51
6.2.2.2 Estimated Acreage of Harvested Waters
The available data do not indicate the distribution of shellfish landings from waters that the
1995 Shellfish Register identifies as approved, conditionally approved, restricted, or conditionally
restricted. For purposes of this analysis, EPA assumes that baseline landings are harvested primarily
from approved or conditionally approved waters. Thus, in a given state (j), the area of shellfish
growing waters assumed to be harvested is determined by the following calculation:
Acres Harvested^ = Acres Approved^ + Acres Conditionally Approved^
51 The calculation of the mean ignores years for which harvest data for a particular species
are unavailable. If landings in these years were actually zero, this approach will overstate average
annual landings.
6-5
-------
6.2.2.3 Average Annual Yield of Harvested Waters
To calculate the average annual yield (Y) of harvested waters for a given species (n) in a
given state (j), the analysis simply divides the annual baseline harvest (Q0) for that species and state
by the acres assumed to be harvested:
Y(no)= Qo(n,j/Acres Harvesteda)
This calculation provides an estimate of the pounds of shellfish landed per year from harvested
waters.
6.2.2.4 Characterization of Waters that are
Unharvested due to Pollution from AFOs
The next step in the analysis is to estimate the area of shellfish-growing waters that are
currently unharvested due, at least in part, to pollution from AFOs. Consistent with the approach
outlined thus far, EPA assumes that waters classified in the 1995 Shellfish Register as restricted,
conditionally restricted, or prohibited are essentially unharvested. Thus, in a given state (j), the area
of shellfish growing waters assumed to be unharvested is determined by the following calculation:
Acres Unharvested^ = Acres Restricted^ + Acres Conditionally Restricted^ + Acres Prohibited^
This calculation, however, includes all impaired waters. To identify areas impaired, in whole or in
part, by pollution from AFOs, EPA's analysis considers two cases. Under Case 1, EPA evaluates
only those shellfish-growing waters for which AFOs are specifically identified as a contributing
source of impairment. Under Case 2, EPA expands the analysis to include shellfish-growing waters
that the Register identifies as impaired, in whole or in part, by AFOs and/or agricultural runoff. The
inclusion of Case 2 is justified by the classification of shellfish-growing waters on the basis of fecal
coliform levels. To the extent that agricultural runoff causes elevated fecal coliform counts, animal
manure, potentially from AFOs, is the likely contributing factor.52
52 In addition, NOAA staff who maintain the Register suggest that difficulty in pinpointing
the source of pollution often results in classifying impacts from AFOs under the more general
heading of "Agriculture Runoff." Personal communication with Jamison Higgins, NOAA, April 12,
1999.
6-6
-------
6.2.2.5 Estimated Impact of Pollution from
AFOs on Commercial Shellfish Landings
To characterize the impact of pollution from AFOs on commercial shellfish landings, it is
first necessary to estimate the potential yield of impaired shellfish growing areas. For purposes of
this analysis, EPA assumes that the average annual yield from harvested waters, as calculated above,
is representative of the potential annual yield from impaired waters. Thus, the foregone harvest (QF)
from an area of any size for a given species (n) in a given state (j) is calculated as follows:
= Y(nj) x Acres Unharvesteda)
It would be inappropriate, however, to assume that pollution from AFOs accounts for the
entire foregone harvest. The Shellfish Register identifies AFOs and/or agricultural runoff as a source
of impairment for only a subset of impaired growing waters. Moreover, even in situations where
AFOs and/or agricultural runoff contribute to impairment, they are not necessarily the sole source
of impairment. Under these circumstances, it is difficult to characterize the marginal impact of
pollution from AFOs and/or agricultural runoff on annual shellfish harvests.
To address this concern, EPA relies on the Shellfish Register's characterization of the
significance of pollution from AFOs and/or agricultural runoff as a source of impairment. As noted
above, the Register indicates whether a source is an actual or potential contributor to impairment,
and when identified as an actual contributor notes the overall importance of the source (i.e., high,
medium, or low). EPA uses this information to characterize the marginal impact of pollution from
AFOs and/or agricultural runoff on shellfish harvests. It does so by subjectively assigning a
weighting factor that varies with the contribution of AFOs and/or agricultural runoff to impairment
of a given shellfish growing area; Exhibit 6-3 summarizes the weighting factors employed. EPA
then applies these weighting factors to estimates of the foregone harvest (QF) from each shellfish-
growing area to estimate the marginal impact of pollution from AFOs and/or agricultural runoff on
annual shellfish harvests. Mathematically, the harvest of a given species that EPA estimates as lost
due to pollution from AFOs and/or agricultural runoff (QA) in an impaired area is simply the product
of the foregone harvest of the species from the area (QF) and the appropriate weighting factor(W):
QA = QFxW
Summing QA across all impaired shellfish growing areas in a given state provides EPA's estimate
of the marginal impact of pollution from AFOs and/or agricultural runoff on annual commercial
harvests of each shellfish species.
6-7
-------
Exhibit 6-3
WEIGHTING FACTORS EMPLOYED TO CHARACTERIZE
THE IMPACT OF AFOs AND AGRICULTURAL RUNOFF ON HARVESTS
FROM IMPAIRED SHELLFISH-GROWING WATERS
Significance of Source
Not a contributor
Potential contributor
Actual contributor (low)
Actual contributor (medium)
Actual contributor (high)
Weighting Factor
.00
.25
.50
.75
1.0
6.2.3 Estimated Impact of Alternate Regulations
on Commercial Shellfish Harvests
The next step in EPA's analysis is to estimate the impact of alternate CAFO regulatory
scenarios on commercial shellfish harvests. To do so, EPA employs information obtained from the
surface water quality modeling effort described in Chapter 4. The modeling exercise does not extend
to estuaries or near-coastal waters, where most commercial shellfish-growing areas are located;
however, it does consider the impact of each regulatory scenario on fecal coliform counts associated
with pollution from AFOs. Since classification of shellfish-growing areas is based in part on
measurement of fecal coliform counts, the estimated impact of alternate regulatory scenarios on this
measure of water quality in upstream areas provides a reasonable proxy for characterizing potential
changes downstream.
EPA's approach to estimating the beneficial impacts of new CAFO regulations on
commercial shellfish harvests assumes that the adverse impact of pollution from AFOs will be
reduced in proportion to modeled reductions in fecal coliform pollution attributable to AFOs. The
details of this approach are described below.
• First, EPA assigns each coastal state included in the shellfish analysis to one
of the 18 hydroregions that EPA's water quality model evaluates. In all but
one case, the entire coastline of each state falls within a single hydroregion.
The sole exception, Virginia, was assigned to the hydroregion that
encompasses the majority of its coastline.
6-8
-------
Next, for each hydroregion/state, EPA calculates the percentage reduction in
fecal coliform counts predicted under each regulatory scenario.53
Third, EPA multiplies its estimates of the percentage reduction in fecal
coliform counts attributable to AFOs by its previously developed estimates
of the impact of pollution from AFOs and/or agricultural runoff on shellfish
harvests (QA). This calculation was performed separately for each species
and state. The result, QR, represents the incremental increase in harvest
associated with each regulatory alternative.
Adding QR to baseline harvests (Q0) yields an estimate of annual shellfish harvests following
implementation of revised CAFO regulations (Qx). This calculation is performed for each state and
species. Thus:
Ql(nj) = Qo(nj) +QR(n,j)
6.2.4 Valuation of Predicted Change in Shellfish Harvests
The appropriate measure of the economic benefits of an increase in commercial shellfish
harvests is the welfare gain (i.e., the change in producer and consumer surplus) associated with the
increased harvest. For purposes of this analysis, EPA focuses solely on changes in consumer
surplus.54 This focus is necessary because the information required to evaluate any changes in
producer surplus that might result from an increase in shellfish harvests (i.e., a long-run supply curve
for each species harvested) is difficult to obtain. In addition, the shellfish harvesting industry is to
a significant extent characterized by regulated harvest levels and unregulated harvester effort (i.e.,
open access fisheries).55 Generally accepted natural resource economics theory suggests that, in open
access fisheries, overcapitalization leads to zero producer surplus. Thus, although shellfish
harvesting is not entirely open access, any producer surplus in the industry is likely to be small, and
any changes in producer surplus brought about by new CAFO regulations is likely to be minor.
53 Regional estimates of changes in fecal coliform counts are only available for NPDES
Scenarios 1, 2/3, and 4b. For purposes of characterizing changes in fecal coliform counts under
NPDES Scenario 4a, EPA employs the estimated percentage reduction nationwide.
54 As discussed in Chapter 3, the concept of consumer surplus is based on the principle that
some consumers benefit at current prices because they are able to purchase a good at a price that is
less than the amount they are willing to pay.
55 Anecdotal evidence suggests that some shellfishing areas are leased by municipalities to
individual enterprises with sole rights to harvest the area. In these cases, the limits on competition
could lead to positive producer surplus. The extent of this practice, however, is unclear.
6-9
-------
To calculate the change in consumer surplus associated with an increase in commercial
shellfish harvests, the analysis makes use of information on consumer demand. Exhibit 6-4
illustrates a simple demand curve. The demand curve is the downward sloping solid line labeled D,
and the initial quantity sold is the dashed, vertical line at Q0. The intersection of these two lines
gives the price at which quantity Q0 is sold. This price is marked as P0 and represented by the dashed
horizontal line. The consumer surplus for quantity Q0 is the area below the demand curve and above
the horizontal line at P0. That is, the consumer surplus for Q0 is the area labeled "C" in Exhibit 6-4.
Exhibit 6-4
CONSUMER DEMAND AND CONSUMER SURPLUS
Price
PO
PI
00 Q1
Quantity
The measurement of the benefits of revised CAFO regulations relies on the assumption that
a decrease in the contamination of shellfish-growing waters would increase commercial access to
shellfish beds, and thus increase the quantity of shellfish supplied to consumers (i.e., an increase
from Q0 to C^). This in turn would result in a lower market price for shellfish (i.e., P^. The benefit
to consumers can be determined based on the old and new prices and quantities. Before the change,
the area labeled "C" in Exhibit 6-4 measures consumer surplus. After the change, consumer surplus
is measured by the area of A+B+C. Thus, the difference in consumer surplus between these
scenarios (i.e., Area A + Area B) is the additional consumer surplus attributable to the proposed rule
and the appropriate economic measure of benefits to consumers.
6-10
-------
6.2.4.1
Characterization of Consumer Demand for Shellfish
Analysis of the changes in consumer surplus that might result from an increase in shellfish
harvests requires an understanding of the effect of an increased harvest on market prices. To gather
the necessary information, EPA reviewed the economics literature. This review identified a number
of relevant studies: Lipton and Strand (1992), which estimates a demand equation for surf clams and
ocean quahogs on the East Coast; Wessells et al. (1995), which estimates a demand equation for U.S.
harvested mussels in Montreal; Cheng and Capps, Jr. (1988), which estimates demand equations for
oysters and total shellfish in the U.S.; and Capps, Jr. and Lambregts (1991), which estimates demand
equations for scallops and oysters in Houston, Texas. Exhibit 6-5 lists the demand elasticities
obtained from each of these studies.56 These demand elasticities provide the means to determine the
consumer benefits associated with changes in shellfish harvests.
Exhibit 6-5
SHELLFISH DEMAND ELASTICITIES
Citation
Cheng and Capps
Cheng and Capps
Capps and Lambregts
Capps and Lambregts
Wessells et al.
Lipton and Strand
Lipton and Strand
Species
oysters
total shellfish
oysters
scallops
mussels
surf clams
ocean auahoss
Elasticity
-1.132
-0.885
not significant
-1.84
-1.98
-2
-0.87
6.2.4.2 Determining the Change in Consumer
Surplus Associated with Increased Harvests
EPA's analysis of the benefits of an increase in shellfish harvests begins by estimating prices
and quantities (i.e., P0 and Q0) under baseline conditions, as well as the quantity of shellfish EPA
estimates would be harvested following the implementation of new CAFO regulations (Qj).
Consistent with the analysis of shellfish harvests described above, Q0 for each state and species is
based on NMFS data, and specified as the mean annual harvest for the years 1994 through 1998. P0
is calculated by dividing the total reported revenues from 1994 through 1998 for each species and
state, adjusted to 1999 dollars, by the total quantity harvested.57 Ql is determined as described above,
56 The price elasticity of demand represents the percentage change in demand for a good
brought about by a one percent change in its price; thus, a price elasticity of -2 implies that a one
percent increase in price will result in a two percent decrease in demand.
57 EPA adjusts reported revenues to 1999 dollars using the Consumer Price Index. For
purposes of calculating P0, EPA considers only those years for which harvest and revenue data are
available.
6-11
-------
adding to Q0 the increase in shellfish harvests estimated to occur under each regulatory scenario (QR).
EPA determined the value of these factors for each broad category of shellfish for which NMFS data
are available: scallops, oysters, mussels, and clams. When the data allow, EPA developed separate
values for quahogs, surf clams, and other clams. This approach enables the analysis to take
advantage, whenever possible, of the demand equations identified for the quahog and surf clam
subcategories.58
Once P0, Q0, and Ql are estimated, the appropriate price elasticities of demand are applied
to determine the new price (Pj) associated with an increase in shellfish harvests. For purposes of this
analysis, the percentage change in price is determined by dividing the percentage increase in the
quantity of shellfish supplied in each case by the appropriate price elasticity. This percentage change
is then applied to the initial price (P0) to calculate the new price (Pj) for each species harvested.59
58 The analysis employs the Wessells et al. demand elasticity for mussels and the Capps and
Lambregts demand elasticity for scallops for all states in which these species are harvested. When
disaggregated data on surf clam or quahog harvests are available, the analysis relies on the demand
elasticities for these species developed by Lipton and Strand; in all other instances, demand for clams
is analyzed using the total shellfish price elasticity estimated by Cheng and Capps. For oysters, the
analysis relies upon the demand elasticity estimated by Cheng and Capps; this value was selected
because it was based on evaluation of a broader market than that considered by Capps and
Lambregts.
59 Mathematically, the price elasticity of demand (•) is calculated as:
. =.Q/.p
where:
•Q = (Qi-Q0)/Qo
•P = (P1-P0)/P0
therefore:
.p = .Q/.
Pi = (Qi-Qo)(Po)/[('XQo)] + PO
6-12
-------
EPA employs the estimated values for P0, Pl3 Q0 and Qx to measure the increase in consumer
surplus associated with the projected increase in shellfish harvested and resulting reduction in market
price under each regulatory scenario. This calculation is conducted for every state and species
category. The estimated annual benefit under each regulatory scenario is simply the sum of the
estimated increase in consumer surplus across states and species.60
6.3 RESULTS
Exhibit 6-6 summarizes the estimated economic benefits associated with increased shellfish
harvests under each regulatory scenario evaluated. The exhibit presents two cases: Case 1, which
considers beneficial impacts on shellfish growing waters that the Shellfish Register specifically
identifies as impaired by pollution from AFOs; and Case 2, which expands the analysis to consider
beneficial impacts on shellfish growing waters identified as impaired by pollution from AFOs and/or
agricultural runoff. As the exhibit indicates, EPA's estimates of annual benefits under Case 2 are
more than an order of magnitude greater than under Case 1; this range reflects the significant
increase in the number and area of shellfish growing waters considered to be impaired by AFOs
when runoff from agricultural land, as opposed to pollution specifically attributed to AFOs, is
included in the analysis. For the proposed scenarios, the estimate of annual benefits ranges from
approximately $195 thousand under Case 1 to $2.7 million under Case 2. Within each case, the
range of annual benefits across regulatory scenarios is relatively small: from $146 thousand to $245
thousand under Case 1, and from $1.8 million to $3.0 million under Case 2. In both cases, Option
I/Scenario 1 provides the lowest benefits, while Option 2/Scenario 4b provides the highest.
Exhibit 6-6
ESTIMATED ANNUAL BENEFITS OF INCREASED COMMERCIAL SHELLFISH HARVESTS
(1999 $, thousands)
Regulatory Scenario
Option I/Scenario 1
Option I/Scenario 2/3
Option I/Scenario 4a
Option I/Scenario 4b
Option 2/Scenario 1
Option 2/Scenario 2/3*
Option 2/Scenario 4a*
Option 2/Scenario 4b
Casel: AFOs
$146
$198
$178
$221
$165
$219
$195
$245
Case 2: AFOs and
Agricultural Runoff
$1,816
$2,357
$2,204
$2,564
$2,141
$2,717
$2,417
$2,960
* Proposed scenarios
60 The calculation of increased consumer surplus is based on a simple geometric
approximation of the change in areas under the demand curve, rather than formal integration using
calculus. As a result, the estimated increase in consumer surplus may be slightly overstated.
6-13
-------
6.4 LIMITATIONS AND CAVEATS
The analysis set forth above is subject to a number of uncertainties and relies upon several
simplifying assumptions. These factors may lead to a potential under- or over-estimation of the
benefits of decreasing AFO-related contamination of commercial shellfish growing waters. The
most significant of these limitations are described below.
• The analysis assumes that a reduction in pollution from AFOs will result in
an increase in commercial shellfish harvests. While this assumption appears
reasonable in light of the extent to which AFOs contribute to current
restrictions or prohibitions on shellfish harvesting, the actual impact of these
restrictions or prohibitions on annual shellfish landings is unknown.
• To estimate the potential impact of pollution on annual shellfish landings, the
analysis calculates an average annual yield (pounds per acre) for shellfish
growing waters. The calculation of this figure assumes that current harvests
are obtained from waters classified as approved or conditionally approved.
To the extent that this approach over- or understates the increase in annual
yields that might be realized from waters currently subject to harvest
restrictions or prohibitions, the analysis may either over- or understate the
impact of pollution on annual shellfish landings.
• The actual contribution of AFOs to the impairment of shellfish growing
waters is unclear. In light of ambiguities in the data and uncertainties
associated with the impact of pollution from other sources, the analysis
considers two cases and employs subjectively developed weighting factors
to characterize the impact of pollution from AFOs on shellfish harvests. The
degree of judgment required in the analysis and the broad range of results
across the cases analyzed suggests considerable uncertainty concerning the
impact of pollution from AFOs.
• Similarly, in characterizing the impact of alternate regulatory scenarios, the
analysis assumes that the adverse impact of pollution from AFOs (i.e., the
foregone harvest) will be reduced in proportion to modeled reductions in
fecal coliform pollution. While this approach may provide a reasonable
approximation of the relative impacts of alternate CAFO regulations, it is less
reliable than detailed modeling of pathogen concentrations in shellfish-
growing areas. The nature and magnitude of any bias introduced by reliance
on this approach is unclear.
6-14
-------
The analysis relies on estimates of the price elasticity of demand for shellfish
that are not necessarily representative of current conditions or of conditions
nationwide. The nature and magnitude of any bias introduced by reliance on
these estimates, however, is unclear.
6.5 REFERENCES
Capps, Oral Jr. and Johannes Adrianus Lambregts. 1991. "Assessing Effects of Prices and
Advertising on Purchases of Finfish and Shellfish in a Local Market in Texas," Southern
Journal of Agricultural Economics, Vol. 23, July, pp. 181-194.
Cheng, Hsiang-tai and Oral Capps, Jr. 1998. "Demand Analysis of Fresh and Frozen Finfish and
Shellfish in the United States," American Journal of Agricultural Economics, Vol. 70,
August, pp. 533-542.
Griffiths, Charles and Sabrina Lovell. 2000. "Benefits of Proposed CAFO Rules to Shellfish Beds,"
Office of Economy and Environment, U.S. Environmental Protection Agency, July 28.
Lipton, D.W. and I. Strand. 1992. "Effect of Stock Size and Regulations on Fishing Industry Cost
and Structure: The Surf Clam Industry," American Journal of Agricultural Economics, Vol.
64, February, pp. 197-208.
U.S. National Oceanic and Atmospheric Administration. 1997. The 1995 National Shellfish Register
of Classified Growing Waters, Office of Ocean Resources Conservation and Assessment,
Silver Spring, MD. Available from: http://seaserver.nos.noaa.gov/projects/95register/.
U.S. National Oceanic and Atmospheric Administration, Fisheries Statistics and Economics
Division, "Commercial Fisheries," U.S. Department of Commerce, obtained from
http://www.st.nmfs.gov/stl/commercial/index.html.
Wessells, Cathy R., Christopher J. Miller and Priscilla M. Brooks. 1995. "Toxic Algae
Contamination and Demand for Shellfish: A Case Study of Demand for Mussels in
Montreal," Marine Resource Economics, Vol. 10, pp. 143-159.
6-15
-------
REDUCED CONTAMINATION OF PRIVATE WELLS CHAPTER 7
7.1 INTRODUCTION
CAFOs can contaminate aquifers and thus impose health risks and welfare losses on those
who rely on groundwater for drinking water or other uses. Of particular concern are nitrogen and
other animal waste-related contaminants (which come from manure and liquid wastes) that leach
through soils and ultimately reach groundwater. Nitrogen loadings convert to elevated nitrate
concentrations at household and community system wells, and elevated nitrate levels in turn pose
a risk to human health.
The federal health-based National Primary Drinking Water Standard for nitrate is 10 mg/L.
This Maximum Contaminant Level (MCL) applies to all Community Water Supply systems, but not
to households that rely on private wells. As a result, households served by private wells are at risk
to exposure to nitrate concentrations above 10 mg/L, which EPA considers unsafe for sensitive
subpopulations (e.g., infants). Nitrate above concentrations of 10 mg/L can cause
methemoglobinemia ("blue baby syndrome") in bottle-fed infants (National Research Council,
1997), which causes a blue-gray skin color, irritableness or lethargy, and potentially long-term
developmental or neurological effects. Generally, once nitrate intake levels are reduced, symptoms
abate. If the condition is untreated, however, methemoglobinemia can be fatal.61
The most recent U.S. Census data show that approximately 13.5 million households located
in counties with AFOs are served by domestic wells. A number of sources provide information on
the percentage of such wells with nitrate concentrations in excess of 10 mg/L. As indicated in
Exhibit 7-1, the values reported vary widely, depending on the location studied, local hydrology, and
other factors. According to the nationwide USGS Retrospective Database, however, the
concentration of nitrate in 9.5 percent of domestic wells in the U.S. exceeds the 10 mg/L threshold.
Thus, EPA estimates that approximately 1.3 million households in counties with AFOs are served
by domestic wells with nitrate concentrations above 10 mg/L.
61 No other health impacts are consistently attributed to elevated nitrate concentrations in
drinking water. As discussed in Chapter 2, however, other health effects are suspected.
7-1
-------
Exhibit 7-1
PERCENTAGE OF WELLS EXCEEDING THE MCL FOR NITRATE
Study
CDC, 1998
Agriculture
Canada, 1991 (as
cited by Giraldez
and Fox, 1995)
Krossetal., 1993
Retrospective
Database (USGS)
Richards et al.,
1996
Spalding and
Exner, 1993
Swistock et al., 1993
U.S. EPA, 1990
USGS, 1985
USGS, 1998
Vitosh, 1985 (cited in
Walker and Hoehn,
1990)
Location
Illinois, Iowa, Kansas, Minnesota,
Missouri, Nebraska, North Dakota,
South Dakota, Wisconsin
Ontario
Iowa
National
Indiana, Kentucky, Ohio, West
Virginia
Iowa, Kansas, Nebraska, North
Carolina, Ohio, Texas
Pennsylvania
National
Upper Conestoga River Basin
Nemaha Natural Resources District, Nebraska
Southern Michigan
Type of
Well
Domestic
Domestic
farm
Rural
Domestic
Rural
Rural
Private
Rural
domestic
Rural
Rural
Rural
Percent Exceeding
lOmg/L
13.4%
13%
18%
9.5%
3.4%
20%, 20%, 20%,
3.2%, 2.7%, 8.2%,
respectively
9%
2.4%
40+%
10%
34%
EPA's proposed revisions to the NPDES regulation and effluent guidelines would affect the
number and type of facilities subject to regulation as CAFOs, and would also introduce new
requirements governing the land application of manure. As a result, EPA anticipates that its
regulatory proposal will reduce nitrate levels in household wells. In light of clear empirical evidence
from the economics literature that households are willing to pay to reduce nitrate concentrations in
their water supplies — especially to reduce concentrations below the MCL — the anticipated
improvement in the quality of water drawn from private domestic wells represents a clear economic
benefit. This chapter estimates these benefits for each of the eight regulatory scenarios evaluated.
7-2
-------
7.2 ANALYTIC APPROACH
Exhibit 7-2 provides an overview of EPA's approach to estimating the benefits of well nitrate
reductions. As the exhibit indicates, the analysis begins by developing a statistical model of the
relationship between nitrate concentrations in private domestic wells and a number of variables
found to affect nitrate levels, including nitrogen loadings from AFOs. It then applies this model, in
combination with the projected change in nitrogen loadings from CAFOs under each regulatory
scenario, to characterize the distribution of expected changes in well nitrate concentrations. Next,
the analysis applies this distribution to the number of households served by private domestic wells
to calculate (1) the increase in the number of households served by wells with nitrate concentrations
that are below the MCL and (2) the marginal change in nitrate concentrations for households
currently served by wells with nitrate concentrations below the MCL. Finally, the analysis employs
estimates of households' values for reducing well nitrate concentrations to develop a profile of the
economic benefits of anticipated improvements in well water quality. Additional detail on EPA's
analytic approach is provided below.
7.2.1 Relationship Between Well Nitrate
Concentrations and Nitrogen Loadings
EPA's approach begins with the use of regression analysis to develop a model characterizing
the empirical relationship between well nitrate concentrations and a number of variables that may
affect nitrate levels, including nitrogen loadings from AFOs. The variables included in the model
are based on a review of hydrogeological studies that have observed statistical relationships between
groundwater nitrate concentrations and various other hydrogeological and land use factors. The
following discussion describes the variables included in EPA's model and the sources of data for
each variable. It also notes potentially significant variables that the model does not include.
Appendix 7-A and Appendix 7-B provide additional detail on the model's development.
7.2.1.1 Included Variables and Data Sources
Although the groundwater monitoring and modeling studies that EPA reviewed covered
different geographic areas and focused on varying nitrate sources (e.g., septic systems, agricultural
fertilizers, animal feedlots), they often found similar variables to be significant. In particular,
nitrogen application or loadings rates, whether from animal wastes, private septic systems, or
agricultural fertilizers, were the most consistent and significant factor affecting well nitrate levels
(e.g., Burrow, 1998; CDC, 1998).62 EPA's model includes variables characterizing nitrogen loadings
from each of these sources:
62 Citations provided here and in Section 7.2.1.2 are suggestive of the literature on these
issues. A more comprehensive review of the literature is provided in Attachment B.
7-3
-------
Exhibit 7-2
OVERVIEW OF ANALYTIC APPROACH
DATA SOURCES
NPLA
Retrospective Database
U.S. Census
Ag Census
NPLA scenarios
U.S. Census
Benefits Transfer
ANALYSIS
Baseline model: Statistical model estimation
Nitrates = cc + po)Co+...+pn)Cn + .
Calculation of changes in well
nitrates under options/scenarios
T
• Change in number of households above
10 mg/L MCL
• Change in nitrates 1 < N < 10 mg/L
T
Annual benefit estimates for
CAFO regulatory options
7-4
-------
AFOs — Studies that addressed the effect of animal manure production on
groundwater nitrate concentrations found a positive correlation between these
variables (e.g., Ritter and Chirnside, 1990; Division of Water Quality,
Groundwater Section, 1998). EPA's model therefore includes a variable that
characterizes nitrogen loadings from AFOs. EPA obtained data on these
loadings, aggregated at the county level, from the National Pollutants
Loadings Analysis (Tetra Tech, 2000).
Septic Systems — Several studies found that the proximity of septic systems
to wells is a small, but significant, contributing factor to elevated nitrate
concentrations (e.g., Carleton, 1996; Richards et al., 1996). As a proxy
measure for loadings from septic systems, EPA's model includes a variable
characterizing the use of private septic systems in each county. Information
on septic system use was drawn from the 1990 U.S. Census.
Other Sources — Several studies found that the type of crop cultivated in the
vicinity of wells significantly influences well nitrate levels, reflecting
variation in the crops' nutrient and water needs and suggesting that
agricultural fertilizers are a significant source of nitrogen to groundwater
(e.g., Swistock et al., 1993; Lichtenberg and Shapiro, 1997). EPA obtained
data on nitrogen loadings associated with agricultural fertilizers and other
sources (e.g., atmospheric deposition) from the USGS Retrospective
Database (1996).
In addition to variables characterizing nitrogen loadings, EPA's model includes the following
variables describing well, soil, and land use characteristics found to significantly influence well
nitrate concentrations:
• Well Depth — Several studies found well depth to be a significant variable,
inversely correlated with well nitrate concentrations, regardless of nitrate
source (e.g., Detroy, 1988; Ham et al., 1998).
Soil Group: A number of studies identified at least one hydrogeological
characteristic, such as aquifer composition and soil type, as a significant
factor affecting well nitrate concentrations (e.g., Lichtenberg and Shapiro,
1997; Lindsey, 1997).
• Land Use: Agricultural land use in the vicinity of wells was found to be
associated with higher groundwater nitrate in several studies (e.g., Mueller
etal., 1995; Carleton, 1996).
7-5
-------
For purposes of model development, EPA obtained data on these variables from the USGS
Retrospective Database.
7.2.1.2 Omitted Variables
Because of incomplete or unreliable national data, EPA's model does not include all of the
potentially significant variables the literature identifies. For example, several studies cite well
construction and age as significant variables with respect to well nitrate concentrations (e.g.,
Spalding and Exner, 1993; Swistock et al., 1993). In general, older wells are more vulnerable to
nitrate contamination because their casings are more likely to be cracked, allowing surface
contaminants to enter the well. Different construction materials and methods also affect how easily
nitrate or other pollutants can reach groundwater via direct contamination at the wellhead. Data on
this variable, however, are often unreliable because they are generally obtained by surveying well
owners and relying on their subjective assessment of when and how a well was constructed; no
reliable, nationally comprehensive data on well construction are available.
Several studies also found the distance from a pollutant source to the well to be significantly
correlated with well nitrate concentrations (e.g., Swistock et al., 1993; Division of Water Quality,
Groundwater Section, 1998). Although spatial data for well locations are available, data on the
location of animal feedlots, cropland, and septic systems are not; therefore, the model excludes this
variable.
7.2.2 Modeling of Well Nitrate Concentrations
Under Alternate Regulatory Scenarios
To estimate the impact of selected variables on well nitrate concentrations, EPA compiled
a database of 2,928 records. Each record provides information characterizing a different well,
including the observed well nitrate concentration; well depth, soil, and land use information; data
on baseline nitrogen loadings from AFOs; and data characterizing nitrogen loadings from septic
systems, agricultural fertilizer, and other sources. EPA developed its regression model on the basis
of this database.
After estimating the regression model using baseline loading information, EPA estimated
expected values for well nitrate concentrations, both for the baseline and for each of the eight
alternative regulatory scenarios. The calculation of expected values under each scenario employed
data on AFO nitrogen loadings obtained from the National Pollutants Loadings Analysis (Tetra Tech,
2000); these loadings vary across the regulatory scenarios, reflecting different manure application
7-6
-------
rates, manure management practices, and other factors.63 To examine the impact of alternate
regulatory scenarios on well nitrate concentrations, the AFO loadings variable is the only
independent variable that changes value; the values for all other variables are held constant. Exhibit
7-3 summarizes the expected percentage changes in well nitrate concentrations under each scenario.64
Exhibit 7-3
PERCENT REDUCTION IN PROJECTED NITRATE CONCENTRATIONS
Regulatory Scenario
Option 1 — Scenario 1
Option 1 — Scenario 2/3
Option 1 — Scenario 4a
Option 1 — Scenario 4b
Option 2 — Scenario 1
Option 2 — Scenario 2/3*
Option 2 — Scenario 4a*
Option 2 — Scenario 4b
Projected Nitrate Concentration (mg/L)
Mean Percent Reduction
3.0
3.2
3.7
3.7
3.5
3.7
4.3
4.3
Median Percent Reduction
0.8
0.8
0.8
0.8
0.8
0.8
0.9
0.9
*Proposed scenarios
7.2.3 Discrete Changes from above the MCL to below the MCL
As noted above, the most recent U.S. Census data show that approximately 13.5 million
households located in counties with AFOs are served by domestic wells. The USGS Retrospective
Database indicates that the concentration of nitrate in 9.47 percent of U.S. domestic wells exceeds
10 mg/L. Thus, under the baseline scenario, EPA estimates that approximately 1.3 million
households in counties with AFOs are served by domestic wells with nitrate concentrations above
10 mg/L.
63 For additional information on the development of pollutant loadings estimates under
alternate regulatory scenarios, see Chapter 4.
64 Testing of EPA's model indicates that it underestimates well nitrate concentrations (see
Attachment B). As a result, comparing predicted values under alternate regulatory scenarios to
observed baseline values would bias the analysis. To avoid this bias, EPA compares the well nitrate
concentrations the model predicts under each regulatory scenario to the values it predicts under
baseline conditions. The benefits assessment is based on the resulting projected percentage changes
in expected well nitrate concentrations.
7-7
-------
To estimate the impact of alternative CAFO standards on the number of wells that would
exceed the nitrate MCL, EPA applied the mean percentage reduction in nitrate concentrations
predicted under each regulatory scenario to the nitrate concentration values that the USGS
Retrospective Database reports. Based on the resulting values, EPA calculated the percentage
reduction in the number of wells with nitrate concentrations exceeding 10 mg/L. As shown in
Exhibit 7-4, it then applied these values to EPA's baseline estimate of the number of households in
counties with AFOs that are served by domestic wells with nitrate concentrations above 10 mg/L.
Based on this analysis, EPA estimates that the regulatory scenarios evaluated would bring between
152,000 and 166,000 households under the 10 mg/L nitrate threshold.
Exhibit 7-4
EXPECTED REDUCTIONS IN NUMBER OF HOUSEHOLDS WITH WELL
NITRATE CONCENTRATIONS ABOVE 10 mg/L
Regulatory Scenario
Option 1 — Scenario 1
Option 1 — Scenario 2/3
Option 1 — Scenario 4a
Option 1 — Scenario 4b
Option 2 — Scenario 1
Option 2 — Scenario 2/3*
Option 2 — Scenario 4a*
Option 2 — Scenario 4b
Percentage of Wells above
MCL at Baseline Expected
to Achieve MCL under
Option/Scenario
11.9%
11.9%
12.6%
12.6%
12.6%
12.6%
13.0%
13.0%
Reduction in Number of
Households
above the MCL
152,204
152,204
161,384
161,384
161,384
161,384
165,974
165,974
* Proposed scenarios
7.2.4 Incremental Changes below the MCL
Households currently served by wells with nitrate concentrations below the 10 mg/L level
may also benefit from marginal reductions in nitrate concentrations. For purposes of this analysis,
EPA assumes that such incremental benefits would be realized only for wells with baseline nitrate
concentrations between 1 and 10 mg/L; presumably, an individual would not benefit if nitrate
concentrations were reduced to below background levels, which for purposes of this analysis are
assumed to be 1 mg/L.65 Exhibit 7-5 shows, for each regulatory scenario, EPA's estimate of the
65 EPA's analysis also ignores marginal reductions in nitrate concentrations for wells that
would remain above the MCL. The Agency's review of the economics literature failed to identify
studies that would provide an adequate basis for valuing such changes.
7-8
-------
mean and median reduction in nitrate concentrations for wells with baseline values between 1 and
10 mg/L. The exhibit also indicates for these wells the aggregate expected reduction in nitrate levels,
expressed in mg/L.66 EPA estimates that approximately 600,000 households would benefit from
these marginal reductions.
Exhibit 7-5
MEAN AND MEDIAN REDUCTIONS IN NITRATE CONCENTRATIONS FOR WELLS WITH
CONCENTRATIONS BETWEEN 1 AND 10 mg/L AT BASELINE
Regulatory Scenario
Option 1 — Scenario 1
Option 1 — Scenario 2/3
Option 1 — Scenario 4a
Option 1 — Scenario 4b
Option 2 — Scenario 1
Option 2 — Scenario 2/3*
Option 1 — Scenario 4a*
Option 2 — Scenario 4b
Mean Nitrate
Reduction
(mg/L)
0.16
0.16
0.19
0.19
0.18
0.19
0.22
0.22
Median Nitrate
Reduction
(mg/L)
0.12
0.12
0.15
0.15
0.14
0.15
0.18
0.18
Total Expected National
Nitrate Reduction
(mg/L)
961,741
1,007,611
1,186,423
1,186,423
1,103,166
1,159,907
1,374,990
1,374,990
* Proposed scenarios
7.2.5 Valuation of Predicted Reductions
in Well Nitrate Concentrations
EPA's analysis relies upon a benefits transfer approach to value predicted reductions in well
nitrate concentrations. EPA used three general steps to identify and apply values for benefits
transfer:
(1) A literature search to identify potentially applicable primary studies.
(2) Evaluation of the validity and reliability of the studies identified. Primary
evaluation criteria included:
66 The information reported in Exhibit 7-5 pertains only to wells with baseline nitrate
concentrations below the MCL. Information for wells with baseline nitrate concentrations above the
MCL is not included, since the benefits associated with reducing nitrate concentrations in these wells
to below the MCL are potentially captured in valuing the achievement of safe nitrate concentrations.
7-9
-------
• the relevance (applicability) of the commodity being valued in the
original studies to the policy options being considered for CAFOs;
and
• the robustness (quality) of the original study, evaluated on multiple
criteria such as sample size, response rates, significance of findings
in statistical analysis, etc.
(3) Selection and adjustment of values for application to CAFO impacts.
Appendix 7-C provides detailed information on EPA's literature search and the criteria applied to
evaluate and select the studies employed in the benefits assessment.
Through its review and evaluation of the relevant literature, EPA selected three studies to
provide the primary values used for the benefit transfer:
A study by Poe and Bishop (1992), which the analysis employs to value
changes in well nitrate concentrations from above the MCL to below the
MCL.
A study by Crutchfield et al. (1997), which EPA employs to value marginal
changes in nitrate concentrations below the MCL.
• A study by De Zoysa (1995), which EPA employs to value marginal changes
in nitrate concentrations below the MCL.
The discussion below briefly summarizes these studies. Additional information is provided in
Exhibit 7-6.
7-10
-------
Exhibit 7-6
SUMMARY INFORMATION ON STUDIES USED FOR BENEFITS TRANSFER
Study Reference
Year of Analysis
Place
Household Water
Supply/
Groundwater Use
Groundwater
Baseline Scenarios
Change in
Groundwater
Scenario
Source of
Contaminants
Types of Values
Estimated
Duration of Payment
Vehicle
MeanHHWTPin
1999 Dollars
Poe and Bishop
1992
Portage County, WI
100% on private wells
An increase in the number of
wells in Portage County with
nitrate contamination
Groundwater protection
program to keep nitrate levels
below EPA standards
Agricultural activities
Option price (use value)
Annually, for as long as
respondent lives in the county
$412.00 (25% reduction in
nitrates to safe level)
$484.00 (households with
100% probability of future
contamination)
Crutchfield et al.
1997
IN, Central ME, PA, WA
IN 73%; NE 3 1%; PA 47%;
WA 26% non-municipal
None given
If tap water has 50% greater
N levels than EPA's MCL,
how much to reduce to min.
safety standards; how much
to eliminate
Not specified
Total value
Monthly, in perpetuity
$21.72 to reduce from 10
mg/L to 0 mg/L ($2. 17 per
mg/L)
De Zoysa
1995
Maumee River Basin,
northwest Ohio
Not specified
Typical N concentrations
range from 0.5-3 mg/L,
although some are much
higher
Reduce levels to
0.5-1 mg/L
Agricultural fertilizer
Total value
One time
$59.33 (lower bound
mean)
$1.78 per mg/L (using 3%
discount rate)
7.2.5.1
Poe and Bishop (1992)
Poe and Bishop (1992, 1999) and Poe (1993) report on the results of a contingent valuation
study conducted in rural Portage County, Wisconsin, to estimate the conditional incremental benefits
of reducing nitrate levels in household wells. The area has had extensive nitrate problems, and
previous research suggested that 18 percent of private wells in the area exceeded the MCL. The
survey comprised two stages. In the first stage, individuals were asked to submit water samples from
their tap and to complete an initial questionnaire. In the second stage, individuals were provided
with their nitrate test results, general information about nitrates, and a graphical depiction of their
exposure levels relative to both natural levels and the MCL; they then were asked to respond to
contingent valuation questions (expost).
7-11
-------
The respondents' willingness-to-pay values varied, as expected, in accordance with the results
of their wells' nitrate tests and other information provided to them. Poe (1993) reports that
households' whose wells were considered certain at some point in the future to exceed the nitrate
MCL would be willing to pay, on average, $484 per year for a program to keep all wells in Portage
County at or below the MCL. Poe and Bishop (1999) expand on the results of the survey by
developing a non-linear valuation function that characterizes how household willingness to pay for
a 25 percent reduction in well nitrate concentrations varies with the initial extent of nitrate
contamination. Their analysis shows that household willingness to pay for such a program increases
as baseline well nitrate concentrations increase from 2 mg/L to 14.5 mg/L, then declines to zero at
a baseline concentration of approximately 22.5 mg/L. Based on their valuation function, Poe and
Bishop estimate that households would be willing to pay an average of $412 per year for a 25 percent
reduction from a baseline nitrate contamination level of 14.5 mg/L. Since such a change would
reduce nitrate concentrations to very near the MCL, EPA considers it representative of household
willingness to pay to reduce such concentrations to safe levels. Taking the mid-point of the $484
and $412 values reported by Poe (1993) and Poe and Bishop (1999), respectively, EPA estimates that
households whose wells exceed the nitrate MCL would be willing to pay $448 per year to reduce
nitrate concentrations to safe levels.
The reliability of these results appears to be reasonably high because the contingent valuation
(CV) instrument was developed and implemented with careful attention to detail and established CV
research protocol. A potential limitation is that the study is based on a relatively small sample size
(480 households); however, good response rates were obtained from this sample (approximately 80
percent for the first stage and 64 percent for the ex post stage). The Poe and Bishop study is the only
study EPA reviewed that elicited such informed ex post values. These value statements may be
considered more reliable than others because respondents knew more about the condition of their
own water supply and thus were able to make better informed decisions. Moreover, in comparison
to the other studies evaluated, the value estimates from this study seemed to represent a conservative
lower bound on households' values for reducing nitrates to the MCL.
7.2.5.2 Crutchfield et al. (1997)
Crutchfield et al. (1997) evaluated the potential benefits of reducing or eliminating nitrates
in drinking water by estimating average willingness to pay for safer drinking water. They surveyed
800 people in rural and non-rural areas in four regions of the United States (Indiana, Nebraska,
Pennsylvania, Washington), using the contingent valuation method (CVM) and posing questions in
a dichotomous choice format. Respondents were specifically asked what their willingness to pay
would be to have the nitrate levels in their drinking water (a) reduced to "safe levels," and
(b) completely eliminated. Respondents were told that this would be accomplished using a filter
installed at their tap, and the cost would be included in their monthly water bill. Respondents were
also asked questions regarding sociodemographic characteristics such as income, age, education, and
whether they currently use treated or bottled water. Across all regions, the resulting willingness to
pay, per household, to reduce nitrates to safe levels ranged from $45.42 per month to $60.76 per
7-12
-------
month, with a mean of $52.89. The willingness to pay to completely remove nitrates from drinking
water ranged from $48.26 per month to $65.11 per month, with a mean of $54.50. The study found
two variables to be significantly related to a respondent's willingness to pay: "years lived in ZIP
code," which was positively correlated with willingness to pay, and "age of respondent," which was
negatively correlated.
7.2.5.3 De Zoysa (1995)
De Zoysa (1995) applied the contingent valuation method to evaluate the benefits of a
number of programs to enhance environmental quality in Ohio's Maumee River basin, including a
program to stabilize and reduce groundwater nitrate levels. The study solicited willingness-to-pay
values from residents of both rural and urban areas in the river basin, as well as residents of one out-
of-basin urban area. A portion of respondents were asked whether they would pay different amounts,
via a one-time special tax, to reduce nitrate contamination from fertilizer applied to fields. Under
the hypothetical scenarios, nitrate concentrations would be reduced from the current range of 0.5-3.0
mg/L to a range of 0.5-1.0 mg/L. Individuals were also asked questions regarding sociodemographic
characteristics, preferences for priorities for public spending, and how they used the resource in
question. Based on the lower bound of the mean values reported, the study found an average one-
time household willingness to pay of $52.78 for a 1 mg/L reduction in groundwater nitrate
concentrations. The study also found that income, the level of priority placed on groundwater
protection, and interest in increasing government spending on education, healthcare, and vocational
training all were positively and significantly correlated with willingness to pay to improve
groundwater quality.
7 2.5.4 Adjustments to the Values
EPA employs the results of the Crutchfield et al. and De Zoysa reports to estimate annual
household willingness to pay to reduce well nitrate concentrations when those concentrations are
already below the nitrate MCL. EPA derives the appropriate value from Crutchfield by comparing
the reported monthly willingness-to-pay values for reducing nitrate concentrations from above the
MCL to the MCL and from above the MCL to zero. The difference between these values is $1.61
per month. For a change between the MCL of 10 mg/L and 0 mg/L, this represents a per mg/L
monthly willingness to pay of $0.16, or $1.92 annually (1997$). To derive a comparable annual
value from De Zoysa, EPA annualizes the willingness to pay value obtained from that study (an
average one-time household willingness to pay of $52.78 for a 1 mg/L reduction in groundwater
nitrate concentrations) at an annual discount rate of 3 percent. This calculation yields an estimated
annual household willingness to pay for a 1 mg/L reduction in nitrate concentrations of $1.58
7-13
-------
(1995$). EPA applied the Consumer Price Index (CPI) to convert these values to 1999 dollars.67
The Agency then applied the midpoint of the two values, $1.97 per mg/L per household per year, to
value changes in well nitrate concentrations between 10 mg/L and 1 mg/L. Reductions in well
nitrate concentrations below 1 mg/L are not valued, since EPA assumes a natural nitrate background
level of 1 mg/L.
As noted above, EPA relies on the findings of Poe and Bishop to estimate that households
whose wells exceed the nitrate MCL would be willing to pay $448 per year to reduce nitrate
concentrations to safe levels. These values are expressed as willingness to pay per year as long as
the individual lives in the county, and thus can be directly translated to the policy scenarios.
Exhibit 7-7 summarizes the point value estimates used for benefits transfer.
Exhibit 7-7
WILLINGNESS-TO-PAY VALUES APPLIED TO BENEFITS TRANSFER
Study
Poe and Bishop
Average of Crutchfield et al. and De Zoysa
Value
Annual WTP
Annual WTP per mg/L between 10 mg/L and 1 mg/L.
1999$
$448.00
$1.97
7.2.5.5
Timing of Benefits
It is unlikely that changes in CAFO regulations would immediately result in the changes in
well nitrate concentrations that EPA's statistical model predicts. While hydrogeological conditions
and other factors may vary significantly from case to case, considerable time may pass before most
wells reach the steady state nitrate concentrations the model forecasts. Therefore, it is necessary to
develop a time profile of the anticipated benefits of revised CAFO standards.
EPA estimates that approximately 75 percent of affected wells would realize the full benefits
of reduced nitrogen loadings within 20 years (Hall, 1996). Assuming that the number of wells
achieving new steady state conditions increases linearly over time, this translates to approximately
3.7 percent of wells achieving new steady state conditions each year. At this rate, all affected wells
would achieve new steady state conditions in approximately 27 years. For purposes of characterizing
the benefits of reduced contamination of private wells, EPA's analysis adopts these assumptions.
67 CPI-U Series ID CUUROOOOSAO, not seasonally adjusted, U.S. city average, all items.
7-14
-------
7.3 RESULTS
Exhibit 7-8 illustrates the time profile of benefits for EPA's proposed revisions to the CAFO
standards. As the exhibit shows, annual benefits increase from approximately $3 million per year
in the first year following implementation to between $70 million and $80 million annually in the
twenty-seventh and subsequent years. The profile of benefits over time for all regulatory scenarios
is similar, but in each case reaches a different steady-state level. Exhibit 7-9 summarizes the
estimated annual benefits, once steady state conditions are achieved, for each of the regulatory
scenarios evaluated. As the exhibit indicates, these benefits range from a low of approximately $70
million per year, the estimate arrived at for Option 1-Scenario 1, to a high of approximately $77
million per year, the estimate developed for both Option 2-Scenario 4a and Option 2 Scenario 4b.
7.4 LIMITATIONS AND CAVEATS
Omissions, biases, and uncertainties are inherent in any analysis relying on several different
data sources, particularly those that were not developed specifically for that analysis. Exhibit 7-10
summarizes key omissions, uncertainties, and potential biases for this analysis.
7-15
-------
Exhibit 7-8
ANNUAL BENEFITS OF REDUCING PRIVATE WELL CONTAMINATION
$90.00
$80
$70.00
$60
$50.00
$40
ON
ON
ON
0)
I
•3 $30.00
$20
$10.00
0 10 20 30 40 50 60 70
Years After Implementation
80 90
100
- Option 1-
Option 1-
Option 1-
- Option 1-
- Option 2-
Option 2-
- Option 2-
- Option 2-
Scenario 1
Scenario 2/3
Scenario 4b
Scenario 4a
Scenario 1
Scenario 2/3
Scenario 4a
Scenario 4b
7-16
-------
Exhibit 7-9
ESTIMATED ANNUAL BENEFITS OF REDUCED
CONTAMINATION OF PRIVATE WELLS UNDER STEADY
STATE CONDITIONS
($1999, millions)
Regulatory Scenario
Option 1 -Scenario 1
Option 1 -Scenario 2/3
Option 1 -Scenario 4a
Option 1 -Scenario 4b
Option 2-Scenario 1
Option 2-Scenario 2/3*
Option 2-Scenario 4a*
Option 2-Scenario 4b
Annual Benefits
$70.08
$70.17
$74.64
$74.64
$74.47
$74.59
$77.07
$77.07
* Proposed scenarios
7-17
-------
Exhibit 7-10
OMISSIONS, BIASES, AND UNCERTAINTIES IN THE NITRATE LOADINGS ANALYSIS
Variable
Likely
Impact on Net
Benefit
Comment
Well, land, and nitrate data
Geographic coverage
Well location selection
Year of sample
Nitrate loadings from AFOs with 0-
300AU
Percent of wells above 10 mg/L
Sampling methods
Unknown
Positive
Unknown
Positive
Unknown
Unknown
Data availability limited the well samples used in the statistical
modeling to those from 138 counties nationwide.
Wells sampled in the USGS Retrospective database may or
may not be random. Samples may be focused on areas with
problems with nitrate.
Samples taken over 23 years. Land use and other factors
influencing nitrate concentrations in the vicinity of the well
may have changed over time.
Data for the smallest AFOs were not included in this analysis
because they will not be affected by the proposed regulations.
This may subsequently underestimate total loadings, resulting
in an overestimate of the impact of nitrogen loadings on well
nitrate concentrations.
Based on the USGS Retrospective database, EPA assumes that
9.74 percent of wells currently exceed the MCL. If the true
national percent is lower (higher), EPA's analysis overstates
(understates) benefits.
Data set compiled from data collected by independent state
programs, whose individual methods for measuring nitrate
may differ.
Model Variables
Well construction and age
Spatial data
Unknown
Unknown
No reliable data available nationally.
No national data available on the distance from well to
pollutant source.
Benefit Calculations
Per household value for reducing well
nitrates to the MCL
Years until wells achieve steady state.
Values for marginal reductions below
the MCL
Baseline characterization
Exclusion of values for reduced nitrate
concentrations in wells that would
remain above the MCL after the
implementation of new regulations
Exclusion of values for marginal
reductions in nitrate concentrations
below the MCL, for wells with nitrate
concentrations above the MCL at
baseline and below the MCL after
implementation of new regulations
Negative
Negative
Positive
Negative
Negative
Negative
The Poe and Bishop values generally appear to be a lower
bound estimate of households' WTP for reducing nitrates to
the MCL.
The analysis assumes a linear path over 27 years until reduced
nitrogen loadings would result in most wells achieving
reduced nitrate concentrations. A large portion of wells
(especially shallower wells) may achieve this on a much faster
time path.
If most of the benefits from reductions in nitrate
concentrations below the MCL are related to a threshold effect
or removing all human induced nitrates, then the assumption
that benefits increase linearly with reductions in nitrate
concentrations from 10 mg/L to 1 mg/L will overstate the
benefits of marginal reductions.
Baseline well concentrations are based on observed levels that
are in some cases more than 20 years old. These reflect AFO
loadings from past decades that likely understate current
loadings and, hence, underestimate anticipated well
concentrations absent regulations.
Reductions in nitrate concentrations in wells that would
remain above the MCL after the implementation of new
regulations are not valued. The Agency's review of the
economics literature failed to identify studies that would
provide an adequate basis for valuing such changes.
The benefits of marginal changes in nitrate concentrations
between 10 mg/L to 1 mg/L for wells with nitrate levels above
the MCL at baseline and below the MCL after implementation
of new regulations are not calculated. These benefits are
potentially captured in valuing the achievement of safe nitrate
concentrations.
7-18
-------
7.5 REFERENCES
Agriculture Canada. 1991. Ontario Farm Groundwater Quality Survey. Agriculture Canada,
Ottawa.
Burrow, K.R. 1998. Occurrence of Nitrate and Pesticides in Groundwater beneath Three
Agricultural Land-Use Settings in the Eastern San Joaquin Valley, California. USGS,
Denver, CO.
Carleton, G.B. 1996. Nitrate in Groundwater and Surface Water in a Residential Subdivision, West
Mercer Township, New Jersey. USGS, Denver, CO.
CDC. 1998. A Survey of the Quality of Water Drawn from Domestic Wells in Nine Midwest
States. NCEH 97-0265. Centers for Disease Control and Analysis. National Center for
Environmental Health.
Crutchfield, S.R., J.C. Cooper, and D. Hellerstein. 1997. Benefits of Safer Drinking Water: The
Value of Nitrate Reduction. Agricultural Economic Report 752. U.S. Department of
Agriculture, Economic Research Service, Food and Consumer Economics Division.
De Zoysa, A.D.N. 1995. A Benefit Evaluation of Programs to Enhance Groundwater Quality,
Surface Water Quality and Wetland Habitat in Northwest Ohio. Dissertation, The Ohio State
University, Columbus.
Desvousges, W.H., M.C. Naughton, and G.P. Parsons. 1992. "Benefit Transfer: Conceptual
Problems in Estimating Water Quality Benefits Using Existing Studies." Water Resources
Research 28(3):675-683.
Detroy, M.G. 1988. Groundwater Quality Monitoring Program in Iowa: Nitrate and Pesticides in
Shallow Aquifers. USGS, Denver, CO.
Division of Water Quality, Groundwater Section. 1998. Impact of Animal Waste Lagoons on
Ground Water Quality. Draft Report. Department of Environment and Natural Resources,
Raleigh, NC.
Giraldez, C. and G. Fox. 1995. "An Economic Analysis of Groundwater Contamination from
Agricultural Nitrate Emissions in Southern Ontario." Canadian Journal of Agricultural
Economics 43:387-402.
Hall. 1996. Simulation of Nitrates in a Regional Subsurface System: Linking Surface Management
with Ground Water Quality. Ph.D. Thesis. Colorado State University.
7-19
-------
Ham, J.M., L.N. Reddi, C.W. Rice and J.P. Murphy. 1998. Evaluation of Lagoons for Containment
of Animal Waste. Manhattan: Department of Agronomy, Kansas State University.
Kross, B.C., G.R. Hallberg, and D.R. Bruner. 1993. "The Nitrate Contamination of Private Well
Water in Iowa." American Journal of Public Health 83:270.
Lichtenberg, E. and L.K. Shapiro. 1997. "Agriculture and Nitrate Concentrations in Maryland
Community Water System Wells." Journal of Environmental Quality 26:145-153.
Lindsey, B.D. 1997. Nitrate in Groundwater and Streambase Flow in the Lower Susquehanna River
Basin, Pennsylvania and Maryland. USGS, Denver, CO.
Mueller, O.K., P.A. Hamilton, D.R. Helsel, KJ. Hitt, and B.C. Ruddy. 1995. Nutrients in Ground
Water and Surface Water of the United States - An Analysis of Data through 1992. U.S.
Geological Survey Water-Resources Investigations Report 95-4031.
National Research Council. 1997. Valuing Ground Water: Economic Concepts and Approaches.
National Academy Press, Washington, DC.
Poe, G.L. 1993. Information, Risk Perceptions, and Contingent Value: The Case of Nitrates in
Groundwater. Ph.D. Dissertation, University of Wisconsin-Madison.
Poe, G.L. and R.C. Bishop. 1992. Measuring the Benefits of Groundwater Protection from
Agricultural Contamination: Results from a Two-Stage Contingent Valuation Study. Staff
Paper Series, University of Wisconsin-Madison.
Poe, G.L. and R.C. Bishop. 1999. "Valuing the Incremental Benefits of Groundwater Protection
When Exposure Levels Are Known." Environmental and Resource Economics 13:341-367.
Randall, A. and D. de Zoysa. 1996. Groundwater, Surface Water, and Wetlands Valuation for
Benefits Transfer: A Progress Report. W-133 Benefits and Costs Transfer in Natural
Resource Planning, Ninth Interim Report, Department of Economics, Iowa State University,
Ames, IA.
Richards, R.P. et al. 1996. "Well Water Quality, Well Vulnerability, and Agricultural
Contamination in the Midwestern United States." Journal of Environmental Quality 25:389-
402.
Ritter, W.F. and A.E.M. Chirnside. 1990. "Impact of Animal and Waste Lagoons on Groundwater
Quality." Biological Wastes 34:39-54.
Spalding, R.F. and M.E. Exner. 1993. "Occurrence of Nitrate in Groundwater — A Review."
Journal of Environmental Quality 22:392-402.
7-20
-------
Stratus. 2000. The Benefits of Reducing Nitrate Contamination in Private Domestic Wells. Report
prepared for U.S. Environmental Protection Agency, Office of Science and Technology.
Swistock, B.R., W.E. Sharpe, and P.O. Robillard. 1993. "A Survey of Lead, Nitrate, and Radon
Contamination of Individual Water Systems in Pennsylvania." Journal of Environmental
Health 55:6.
Tetra Tech. 2000. Development of Pollutant Loading Reductions from the Implementation of
Nutrient Management and Best Management Practices. Draft report prepared for
U.S. Environmental Protection Agency, Office of Science and Technology.
U.S. EPA. 1990. National Survey of Pesticides in Drinking Water Wells. U.S. Environmental
Protection Agency.
USGS. 1985. Occurrence of Nitrate and Herbicides in Groundwater in the Upper Conestoga River
Basin, Pennsylvania. Water Resources Investigations Report 85-4202.
USGS. 1996. Retrospective Database for Nutrients in Groundwater and Surface Water. USGS.
http://water.usgs.gov/nawqa/nutrients/datasets/retrodata.html. Accessed 6/1/00.
USGS. 1998. Nitrate Concentrations in Ground Water in the Paleovalley Alluvial Aquifiers of the
Nemaha Natural Resources District, Nebraska, 1989 and 1994-96. USGS Water-Resources
Investigations Report 98-4106. USGS, Lincoln, NE.
Vitosh, M.L. 1985. "Nitrogen Management Strategies for Corn Producers." Extension Bulletin
WQ06. Cooperative Extension Service, Michigan State University. August.
Walker, D.R. and J.P. Hoehn. 1990. "The Economic Damages of Groundwater Contamination in
Small Rural Communities: An Application to Nitrates." Northcentral Journal of
Agricultural Economics 12:47-56.
7-21
-------
Appendix 7-A
MODEL VARIABLES
EPA's statistical analysis of the relationship between nitrogen loadings and well nitrate
concentrations is based on the following linear model:
Nitrate (mg/L) = 30 + 3! Ag Dummy + B2 Soil Group + B3 Well Depth
+ B4 Septic Ratio + B5 Alt N Source + B6 Loadings Ratio + •;
where nitrate concentration (mg/L) is the dependent variable.
The variables used to explain nitrate concentrations in well water (i.e., the model's
independent variables) can be classified into two groups: well and land characteristics, and nitrogen
inputs. Definitions of these variables are provided below. Unless otherwise noted, EPA obtained
the data used in developing the model from the USGS Retrospective Database.
Well and Land Characteristics
Ag Dummy: This variable describes the predominant land use at the well's location (1 for
agricultural land; 0 otherwise). Other land uses identified in the database include woods, range,
urban, and other.
Soil Group: The soil group variable is an index that integrates several factors — including
runoff potential, permeability, depth to water table, depth to an impervious layer, water capacity, and
shrink-swell potential — to characterize hydrological conditions in the vicinity of the well. Possible
values range from a minimum of 1 to a maximum of 4.
Well Depth: The well depths reported in the USGS database range from 1 foot to 5,310 feet.
For observations used in the regression analysis, the maximum well depth is 1,996 feet.
Nitrogen Inputs
AFO Loadings Ratios and Scenarios: The AFO loadings ratio is a unique value for each
county. It is calculated by dividing estimated leached nitrate loadings for the county (pounds per
year) by the county's total area (acres). The analysis employs baseline loadings data to estimate the
7A-1
-------
coefficients for the independent variables. It applies these coefficients, combined with loadings data
representative of post-regulatory conditions, to estimate changes in well nitrate concentrations under
each regulatory scenario.
Septic Ratio: The septic ratio is a proxy measure of potential nitrogen loadings from septic
systems. The analysis develops a unique value for each county. This value is calculated by dividing
the number of housing units in the county that use septic systems by the county's total area (acres).
EPA obtained data on septic system use from the 1990 U.S. Census.
Alternate NSource: Alternate nitrogen sources include fertilizer and atmospheric deposition.
Values for this variable are reported in pounds per acre per year.
Summary Statistics
Exhibit 7A-1 reports summary statistics on the variables used in the analysis.
Exhibit 7A-1
SUMMARY STATISTICS
Variable
Nitrate Concentrations
Ag Dummy
Soil Group
Well Depth
Septic Ratio
Alternate N Source
Loadings Ratio (Baseline
Scenario)
N
2928
2928
2928
2928
2928
2928
2928
Mean
3.585
0.775
2.418
169.191
0.029
28.890
6.626
Standard
Deviation
6.552
0.418
0.658
133.468
0.028
18.981
14.022
Minimum
0.050
0.000
1.000
1.000
0.000
0.869
0.003
Maximum
84.300
1.000
4.000
1,996.000
0.151
99.631
63.354
7A-2
-------
Appendix 7-B
THE GAMMA MODEL
The analysis uses a gamma model to fit the right skew of observed values for well nitrate
concentrations as well as the nonnegative constraint on the dependent variable. Visual inspection
of the nitrate concentration distribution suggests a gamma distribution with density function:
f(y) = —-exP(-0y)yc
T(a)
Let For this distribution, the expected value of yi is
exp (• • j)
The use of the gamma distribution instead of the more commonly employed exponential distribution
is appropriate because • is assumed to equal 1 in the exponential distribution, but was estimated to
be significantly different than 1 in EPA's empirical work. The gamma distribution also offers the
advantages of making the density function more flexible and giving more curvature to the
distribution. The likelihood function is:
Exhibit 7B-1 provides statistical results from the gamma model. All coefficients are of the
expected sign and significant at the 1 percent level. In particular, the coefficient for the Loadings
Ratio variable is positive, as expected, indicating that an increase in nitrogen loadings leads to
increased well nitrate concentrations.
7B-1
-------
Exhibit 7B-1
GAMMA REGRESSION RESULTS
Variable
Intercept
Ag Dummy
Soil Group
Well Depth*
Septic Ratio
Alt N Source*
Loadings Ratio
Alpha
Parameter
Estimate
1.492
0.691
-0.335
-0.106
2.623
20.258
0.010
0.498
Standard Error
0.151
0.066
0.043
0.015
1.102
1.628
0.002
0.011
Asymptotic T-
Statistic
9.891
10.452
-7.725
-7.178
2.380
12.444
5.037
46.368
Significance
0.000
0.000
0.000
0.000
0.009
0.000
0.000
0.000
Mean log-likelihood = -1.854
N = 2,928
*The raw data were scaled by a factor of 100 for Well Depth and 1,000 for Alt N Source in order for the GAUSS
program to converge to a solution.
7B-2
-------
Appendix 7-C
LITERATURE SEARCH AND EVALUATION
Literature Search
The objective of EPA's literature search was to identify prior studies that had developed or
elicited values for changes in groundwater quality, focusing in particular on values for reduced
nitrates. The search drew in part upon two databases: the Colorado Association of Research
Libraries (CARL), which includes the holdings of several university libraries in Colorado and the
West; and the Environmental Valuation Resource Inventory (EVRI), a database compiled by
Environment Canada that includes empirical studies on the economic value of environmental
benefits and human health effects. In addition, EPA solicited suggestions for studies pertaining to
groundwater valuation and nitrate contamination through the ResEcon listserver, which reaches a
network of approximately 700 academics, professionals, and other individuals with interests in
natural resource and environmental economics. Through this extensive search and additional review
of selected bibliographies, EPA identified 11 potentially relevant studies. Since most households'
values for reducing nitrates in private domestic wells are primarily nonmarket values, most of the
identified studies involve stated preference value elicitation (e.g., contingent valuation).
Evaluating Studies for Benefits Transfer
The economics literature suggests several criteria in evaluating primary studies for undertaking
benefits transfer. Desvousges et al. (1992) develop five criteria to guide the selection of studies for
application to a surface water quality issue: that the studies to be transferred (1) be based on
adequate data, sound economic method, and correct empirical technique (i.e., "pass scientific
muster"); (2) evaluate a change in water quality similar to that expected at the policy site; (3) contain
regression results that describe willingness to pay as a function of socioeconomic characteristics; (4)
have a study site that is similar to the policy site (in terms of site characteristics and populations);
and (5) have a study site with a similar market as the policy site. NOAA condenses the five
Desvousges criteria into three considerations: (1) comparability of the users and of the resources
and/or services being valued and the changes resulting from the discharge of concern; (2)
comparability of the change in quality or quantity of resources and/or services; and (3) the quality
of the studies being used for transfer [59 FR 1183]. In a general sense, items (2), (4), and (5) of
Desvousges et al. and items (1) and (2) of NOAA are concerned with the applicability of an original
study to a policy site. Items (1) and (3) of Desvousges et al. and item (3) of NOAA are concerned
with the quality of the original study.
To assess original studies for use in valuing estimated changes in well nitrate levels under
alternate CAFO regulations, EPA evaluated the applicability and the quality of the original studies
on several criteria. To the extent feasible, EPA obtained or derived information from each of the
7C-1
-------
reports or papers for 28 categories of information used to characterize the studies. Because
applicability to CAFOs and quality of the value estimates are distinct concepts, EPA evaluated these
characteristics of the studies separately. Overall, the goal of the rating process was to identify studies
that elicited high-quality value estimates (reliable and valid) and which were most applicable to the
benefits assessment. There were three steps in the rating process:
(1) identify study characteristics upon which to judge applicability and quality;
(2) assign scores to the studies based on these characteristics;
(3) assign weights to these scores for aggregating scores into unidimensional
measures of applicability and quality.
Criteria for Ranking based on Applicability
Applicability refers to the relationship between values elicited in the primary groundwater
valuation studies and benefit estimates necessary for application to the analysis of alternate CAFO
regulations. EPA's criteria for evaluation of applicability included comparison of the following
characteristics of studies with likely CAFO situations:
location (urban, rural, etc.);
water supply/groundwater use (percentage on wells);
type of contaminants (scenario involved nitrate contamination of groundwater);
source of contaminants (scenario involved conditions similar to those relevant
for CAFOs);
• value estimates are for the correct theoretical construct (e.g., total willingness
to pay for reducing groundwater contamination from nitrates).
Criteria for Ranking based on Quality
Analysis of study quality was based on evaluation of the validity and reliability of the value
estimates derived in the primary groundwater valuation research. Most of the 11 identified studies
involved stated preference elicitation using survey methods. Based on professional experience as
to what constitutes a valid and reliable stated preference valuation study, EPA identified
characteristics of these studies that indicate reliability and validity. Criteria for evaluation of study
quality included:
7C-2
-------
whether the study was published or peer reviewed;
whether the survey implementation met professional standards;
how many respondents there were and what the response rate was;
whether and how the groundwater baseline was characterized and what change
was presented in the groundwater scenario;
whether the credibility of scenario change was assessed;
what valuation methodology was used and whether it was appropriate for
eliciting the intended value measures;
the type and duration of payment vehicle;
whether appropriate empirical estimation was undertaken;
whether expected explanatory variables were found to be significant.
7C-3
-------
INTEGRATION OF RESULTS CHAPTER 8
8.1 INTRODUCTION
This chapter summarizes EPA's estimates of the benefits associated with potential revisions
to the NPDES provisions and Effluent Limitation Guidelines (ELGs) pertaining to CAFOs. It first
describes the Agency's approach to aggregating the results of the studies described in Chapters 4
through 7. It then describes EPA's approach to discounting future benefits and presents the
aggregated benefits of each regulatory scenario analyzed, both in a single present value and as an
annualized benefits stream. Finally, the chapter discusses the key limitations of the analysis and the
implications of these limitations in characterizing the benefits of revised CAFO standards.
8.2 INTEGRATION OF ANALYTIC RESULTS
To develop an integrated assessment of the benefits of each alternative regulatory scenario,
EPA simply adds the results of the analyses presented in Chapters 4 through 7. To the extent that
these analyses address similar benefits, this approach may lead to double-counting and
overestimation of benefits. In this case, however, EPA has determined that the potential for double-
counting is small. The analyses that generate the largest benefits estimates — the NWPCAM
analysis of the benefits of improved surface water quality, the evaluation of potential improvements
in commercial shell fishing opportunities, and the assessment of potential reductions in the
contamination of private wells — examine different water resources and different uses of those
resources.68 Thus, the benefits estimated in these analyses are clearly additive. The only possible
source of double-counting, therefore, lies in integrating the results of the NWPCAM analysis with
EPA's evaluation of the benefits attributable to reducing the frequency and magnitude offish kills.
68 The NWPCAM analysis addresses changes in the quality of lakes, rivers, and streams
above the head of tide, and considers recreational swimming, fishing, and non-use values. The
assessment of potential improvements in shell fishing opportunities focuses on marine and estuarine
waters and the commercial use of shellfish resources. The assessment of potential reductions in the
contamination of private wells considers the use of groundwater as a source of domestic water
supplies.
8-1
-------
The extent to which the NWPCAM analysis and the fish kills analysis may double-count
benefits is unclear, but unlikely to be significant. Both analyses address changes in the quality of
rivers, lakes, and streams.69 In addition, at least some of the benefits of reducing the incidence of
fish kills likely stem from the associated improvement in recreational fishing opportunities, a
beneficial use which the NWPCAM analysis considers. Thus, some double-counting is possible.
The NWPCAM analysis, however, is based upon modeling of surface water quality under steady
state conditions; the analysis is not likely to capture all of the impacts of revised CAFO standards
on circumstances (e.g., the overflow of a lagoon under severe storm conditions) that may lead to fish
kills. In addition, the approach employed in valuing the reduced incidence offish kills explicitly
excludes lost use values, such as those associated with recreational fishing. These considerations
suggest that at least some, if not all, of the benefits estimated in the fish kills analysis are incremental
to those estimated in the NWPCAM analysis.
From a practical standpoint, the implications of any double-counting between the NWPCAM
analysis and the fish kills analysis are minimal. At most, the estimated annual benefits of reducing
the incidence offish kills amount to no more than approximately 5 percent of the annual benefits
estimated in the NWPCAM analysis; this is the case for Option 1-Scenario 4a. Under EPA's
proposed regulatory scenarios, the estimated annual benefits of reducing the incidence offish kills
equal less than 0.4 percent of the annual benefits estimated in the NWPCAM analysis. Thus, EPA
has concluded that its approach to integrating the findings of the underlying analyses does not result
in any significant degree of double-counting.
8.3 PRESENT VALUE OF BENEFITS
The results of the analyses in Chapters 4 through 7 are expressed as annual benefits streams.
To calculate the present value of these benefits at the time new regulations are implemented, EPA
employs three alternative real discount rates: three, five, and seven percent. The seven percent
discount rate represents the real rate of return on private investments and is consistent with the rate
mandated by the Office of Management Budget for analysis of proposed regulations. The three
percent discount rate reflects the social rate of time preference for consumption of goods and
services, and is consistent with the rate recommended by many economists for analysis of
environmental benefits. The five percent discount rate represents the mid-point of the three to seven
percent range.
In calculating the present value of benefits, EPA assumes an infinite time frame; i.e., as long
as the regulations remain in effect the associated benefits will be enjoyed in perpetuity. As a
practical matter, this approach is equivalent to assuming that the regulations will remain in effect for
several generations, since the present value of benefits beyond this point approaches zero; however,
69 The data upon which the fish kills analysis is based include fish kill incidents below the
head of tide. The NWPCAM analysis extends only to freshwater resources.
8-2
-------
it avoids the need to arbitrarily specify a period of time over which the regulations are assumed to
remain in effect, and allows EPA to represent fully the present value of the benefits estimated. For
purposes of illustrating the sensitivity of EPA's results to this assumption, Appendix 8-A illustrates
the impact of alternative time frames on present value estimates developed for one of EPA's
proposed regulatory scenarios. Appendix 8-B provides additional detail on the calculation of present
values.
Exhibits 8-1 through 8-4 present the results of the present value calculations for each of the
benefit categories addressed in Chapters 4 through 7. As shown, the estimated benefits for
alternative regulatory scenarios vary significantly, with scenarios that employ Option 2 (which would
establish a phosphorus-based manure application limit) generally providing higher benefits than
scenarios employing Option 1 (a nitrogen-based manure application limit). In all cases, Option 2-
Scenario 4b yields the highest benefits, while Option 1-Scenario 1 yields the lowest. Within benefits
categories, benefits are lowest using the seven percent discount rate and highest using the three
percent discount rate, reflecting the impact of alternative discounting assumptions on the present
value of future benefits.
Exhibit 8-1
PRESENT VALUE OF ESTIMATED BENEFITS:
ACHIEVING RECREATIONAL USE LEVELS
(1999 dollars, millions)
Regulatory Scenario
Option 1 -Scenario 1
Option 1 -Scenario 2/3
Option 1 -Scenario 4a
Option 1 -Scenario 4b
Option 2-Scenario 1
Option 2-Scenario 2/3*
Option 2-Scenario 4a*
Option 2-Scenario 4b
Discount Rates
3 Percent
$163.3
$210.0
$183.3
$240.0
$2,920.0
$4,236.7
$3,616.7
$4,833.3
5 Percent
$98.0
$126.0
$110.0
$144.0
$1,752.0
$2,542.0
$2,170.0
$2,900.0
7 Percent
$70.0
$90.0
$78.6
$102.9
$1,251.4
$1,815.7
$1,550.0
$2,071.4
* Proposed scenarios
8-3
-------
Exhibit 8-5 presents, on a present value basis, EPA's aggregated estimate of the benefits of
alternative revisions to the CAFO standards. As the exhibit shows, the present value of benefits
under EPA's proposed regulatory scenarios ranges from approximately $2.1 billion dollars (assuming
a discount rate of seven percent and employing the low-end of the estimated range of benefits for
Option 2-Scenario 4a) to $6.1 billion dollars (assuming a discount rate of three percent and
employing the high-end of the estimated range of benefits for Option 2-Scenario 2/3).
Exhibit 8-2
PRESENT VALUE OF ESTIMATED BENEFITS:
REDUCED INCIDENCE OF FISH KILLS
(1999 dollars, millions)
Regulatory Scenario
Option 1 -Scenario 1
Option 1 -Scenario 2/3
Option 1 -Scenario 4a
Option 1 -Scenario 4b
Option 2-Scenario 1
Option 2-Scenario 2/3*
Option 2-Scenario 4a*
Option 2-Scenario 4b
Discount Rates
3 Percent
$3.1 -$7.8
$3.9 -$10.2
$3.7 - $9.7
$4.4 -$10.8
$5.9 - $9.4
$7.3 -$12.3
$7.1 -$11.8
$8.1 -$13.2
5 Percent
$1.9 -$4.7
$2.4 -$6.1
$2.2 -$5.8
$2.6 - $6.5
$3.5 -$5.7
$4.4 - $7.4
$4.3 -$7.1
$4.8 -$7.9
7 Percent
$1.3 -$3.3
$1.7 -$4.4
$1.6 -$4.1
$1.9 -$4.6
$2.5 - $4.0
$3.1 -$5.3
$3.0 -$5.0
$3.5 -$5.7
* Proposed scenarios
8-4
-------
Exhibit 8-3
PRESENT VALUE OF ESTIMATED BENEFITS:
IMPROVED COMMERCIAL SHELL FISHING
(1999 dollars, millions)
Regulatory Scenario
Option 1 -Scenario 1
Option 1 -Scenario 2/3
Option 1 -Scenario 4a
Option 1 -Scenario 4b
Option 2-Scenario 1
Option 2-Scenario 2/3*
Option 2-Scenario 4a*
Option 2-Scenario 4b
Discount Rates
3 Percent
$4.9 - $60.5
$6.6 - $78.6
$5.9 -$73. 5
$7.4 -$85. 5
$5.5 -$71.4
$7.3 - $90.6
$6.5 - $80.6
$8.2 -$98.7
5 Percent
$2.9 - $36.3
$4.0 -$47.1
$3.6 -$44.1
$4.4 -$51.3
$3.3 - $42.8
$4.4 . $54.3
$3. 9 -$48.3
$4.9 - $59.2
7 Percent
$2.1 -$25.9
$2.8 -$33.7
$2.5 -$31.5
$3.2 -$36.6
$2.4 - $30.6
$3.1 -$38.8
$2.8 - $34.5
$3.5 - $42.3
* Proposed scenarios
Exhibit 8-4
PRESENT VALUE OF ESTIMATED BENEFITS:
REDUCED CONTAMINATION OF PRIVATE WELLS
(1999 dollars, millions)
Regulatory Scenario
Option 1 -Scenario 1
Option 1 -Scenario 2/3
Option 1 -Scenario 4a
Option 1 -Scenario 4b
Option 2-Scenario 1
Option 2-Scenario 2/3*
Option 2-Scenario 4a*
Option 2-Scenario 4b
Discount Rates
3 Percent
$1,633.2
$1,635.3
$1,739.4
$1,739.4
$1,735.6
$1,738.2
$1,796.0
$1,796.0
5 Percent
$798.2
$799.2
$850.0
$850.0
$848.2
$849.5
$877.7
$877.7
7 Percent
$475.6
$476.2
$506.5
$506.5
$505.4
$506.1
$523.0
$523.0
* Proposed scenarios
8-5
-------
Exhibit 8-5
PRESENT VALUE OF AGGREGATED BENEFITS
(1999 dollars, millions)
Regulatory Scenario
Option 1 -Scenario 1
Option 1 -Scenario 2/3
Option 1 -Scenario 4a
Option 1 -Scenario 4b
Option 2-Scenario 1
Option 2-Scenario 2/3*
Option 2-Scenario 4a*
Option 2-Scenario 4b
Discount Rates
3 Percent
$1,804.5 -$1,864.9
$1,855.9 -$1,934.1
$1,932.4 -$2,005.9
$1,991.2 -$2,075.6
$4,667.0 - $4,736.4
$5,989.5 - $6,077.7
$5,426.2 - $5,505.0
$6,645.5 -$6,741.2
5 Percent
$900.9 -$937.1
$931. 5 -$978.5
$965.8 -$1,009.9
$1,001.1 -$1,051. 8
$2,607.0 - $2,648.7
$3,400.2 - $3,453.2
$3,055. 9 -$3,103.1
$3,787.4 - $3,844.8
7 Percent
$549.0 - $574.9
$570.7 - $604.2
$589.2 - $620.7
$614.4 - $650.6
$1,761.7 -$1,791.4
$2,328.1 -$2,365. 9
$2,078.8 -$2,1 12.5
$2,601.4 - $2,642.4
* Proposed scenarios
8.4 ANNUALIZED BENEFITS ESTIMATES
In addition to calculating the present value of estimated benefits, EPA has developed an
estimate of the annualized benefits attributable to each of the regulatory scenarios analyzed; these
annualized values reflect the constant flow of benefits over time that would generate the associated
present value. Appendix 8-C provides additional detail on the calculation of annualized benefits.
EPA assumes that benefits related to surface water quality improvements will begin
immediately after the revised regulations are implemented (i.e., because loadings will immediately
decrease), and that these benefits will be constant from year-to-year. As a result, the annualized
benefits in these categories (improved surface water quality, reduced fish kills, and improved
shellfishing) are equivalent to the annual benefits estimated in Chapters 4 through 6, regardless of
the discount rate employed. In the case of private well contamination, however, EPA assumes an
uneven annual stream of benefits. As a result, EPA's estimates of the annualized benefits of reduced
private well contamination depend upon the discount rate employed.
Exhibit 8-6 summarizes the range of annualized benefits for each benefit category under each
of the regulatory scenarios analyzed. Note that the range of benefits associated with reduced
contamination of private wells reflects the variation in discount rates employed in developing the
annualized benefits estimate (three to seven percent). In contrast, the range of values presented for
reduced fish kill and shellfishing benefits is based solely on uncertainty in the underlying analyses.
8-6
-------
There is considerable variation in the range of benefits across regulatory scenarios, as well as some
variation within scenarios with respect to the relative magnitude of surface water and groundwater
benefits. In particular, under scenarios that employ Option 1, the annualized benefits of reduced
private well contamination exceed those associated with other benefits categories. In contrast, under
scenarios that employ Option 2, the recreational and non-use benefits associated with improved
surface water quality account for the greatest share of benefits.
Exhibit 8-6
ESTIMATED ANNUALIZED BENEFITS OF REVISED CAFO REGULATIONS
(1999 dollars, millions)
Regulatory Scenario
Option 1- Scenario 1
Option 1- Scenario 2/3
Option 1- Scenario 4a
Option 1- Scenario 4b
Option 2- Scenario 1
Option 2- Scenario 2/3*
Option 2-Scenario 4a*
Option 2- Scenario 4b
Recreational
and Non-use
Benefits
$4.9
$6.3
$5.5
$7.2
$87.6
$127.1
$108.5
$145.0
Reduced Fish
Kills
$0.1 -$0.2
$0.1 -$0.3
$0.1 -$0.3
$0.1 -$0.3
$0.2 - $0.3
$0.2 - $0.4
$0.2 - $0.4
$0.2 - $0.4
Improved
Shellfishing
$0.1 -$1.8
$0.2 - $2.4
$0.2 - $2.2
$0.2 - $2.6
$0.2 -$2.1
$0.2 - $2.7
$0.2 - $2.4
$0.2 -$3.0
Reduced Private
Well
Contamination
$33. 3 -$49.0
$33. 3 -$49.1
$35. 5 -$52.2
$35. 5 -$52.2
$35.4 -$52.1
$35.4 -$52.1
$36.6 -$53. 9
$36.6 -$53.9
* Proposed scenarios
Exhibit 8-7 presents EPA's aggregated estimate of annualized benefits for each of the
regulatory scenarios analyzed. Under the proposed regulatory scenarios, annualized benefits range
from $146 million (Option 2-Scenario 4a, assuming a seven percent discount rate) to $182 million
(Option 2-Scenario 2/3, assuming a three percent discount rate). Again, note that variation in
discount rates affects only the annualized benefits associated with reduced contamination of private
wells; other annualized benefits remain constant regardless of discount rates.
3-7
-------
Exhibit 8-7
SUMMARY OF ANNUALIZED BENEFITS
(1999 dollars, millions)
Regulatory Scenario
Option 1 -Scenario 1
Option 1 -Scenario 2/3
Option 1 -Scenario 4a
Option 1 -Scenario 4b
Option 2-Scenario 1
Option 2-Scenario 2/3*
Option 2-Scenario 4a*
Option 2-Scenario 4b
Discount Rates
3 Percent
Low
$54.1
$55.7
$58.0
$59.7
$140.0
$179.7
$162.8
$199.4
High
$55.9
$58.0
$60.2
$62.3
$142.1
$182.3
$165.1
$202.2
5 Percent
Low
$45.0
$46.6
$48.3
$50.1
$130.4
$170.0
$152.8
$189.4
High
$46.9
$48.9
$50.5
$52.6
$132.4
$172.7
$155.2
$192.2
7 Percent
Low
$38.4
$39.9
$41.2
$43.0
$123.3
$163.0
$145.5
$182.1
High
$40.2
$42.3
$43.4
$45.5
$125.4
$165.6
$147.9
$185.0
* Proposed scenarios
8.5 LIMITATIONS OF THE ANALYSIS AND
IMPLICATIONS FOR CHARACTERIZING BENEFITS
The results presented above are based on the four separate analyses presented in Chapters 4
through 7, and are subject to the specific uncertainties and limitations that are discussed in detail in
each of these chapters. Beyond these limitations, however, it is important to note that EPA's analysis
does not attempt to comprehensively identify and value all potential environmental changes
associated with proposed revisions to the CAFO regulations. Instead, the Agency focuses on specific
identifiable and measurable benefits. The impacts of the regulatory proposal likely include
additional benefits not addressed in these analyses, such as reductions in the cost of treating public
water supplies; improved recreational opportunities in the Great Lakes, estuaries, and other near-
coastal waters; improvement in the condition of non-boatable waters to boatable condition;
improvements in the quality of water resources that are already suitable for swimming;
improvements in commercial fishing; improvements in near-stream activities; and non-water related
benefits, such as potential reductions in odor from waste management areas. In light of these
limitations, EPA believes that the benefits quantified in this report represent a conservative estimate
of the total benefits of revised CAFO standards.
-------
Appendix 8-A
IMPACT OF ALTERNATIVE TIME FRAMES ON PRESENT VALUE AND
ANNUALIZED BENEFITS ESTIMATES
Exhibit 8A-1
PRESENT VALUE OF BENEFITS: OPTION 2-SCENARIO 4a
(1999 dollars, millions)
Time Period
25 years
50 years
75 years
100 years
Infinite
Discount Rate
3 Percent
Low
$2,466.8
$4,012.2
$4,750.9
$5,103.7
$5,426.2
High
$2,508.0
$4,073.0
$4,821.0
$5,178.3
$5,505.0
5 Percent
Low
$1,958.3
$2,731.5
$2,960.1
$3,027.6
$3,055.9
High
$1,991.6
$2,774.6
$3,006.1
$3,074.5
$3,103.1
7 Percent
Low
$1,589.8
$1,988.6
$2,062.2
$2,075.7
$2,078.8
High
$1,617.3
$2,021.2
$2,095.7
$2,109.4
$2,112.5
Exhibit 8A-2
ANNUALIZED BENEFITS: OPTION 2-SCENARIO 4a
(1999 dollars, millions)
Time Period
25 years
50 years
75 years
100 years
Infinite
Discount Rate
3 Percent
Low
$141.7
$155.9
$160.0
$161.5
$162.8
High
$144.0
$158.3
$162.3
$163.9
$165.1
5 Percent
Low
$138.9
$149.6
$151.9
$152.5
$152.8
High
$141.3
$152.0
$154.3
$154.9
$155.2
7 Percent
Low
$136.4
$144.1
$145.3
$145.5
$145.5
High
$138.8
$146.5
$147.6
$147.8
$147.9
8A-1
-------
Appendix 8-B
CALCULATION OF PRESENT VALUES
The present value (PV) of a benefit (B) to be received t years from now is determined by the
following equation:
PV = Bt/(l+r)t
where r represents the annual discount rate. Thus, the present value of an annual stream of benefits
from Year 1 through Year n is calculated as follows:
=» Bt/(l+rJ
When Bt is constant • i.e., when benefits (B) each year are the same • and n approaches infinity, the
equation above can be simplified to:
PV = B/r
EPA employs the above equation to calculate present values for all categories of benefits that
are assumed to remain constant from Year 1 onward; i.e., for all categories except reduced
contamination of private wells. In the latter case, benefits are assumed to increase in a linear fashion
until Year 27, and then to remain constant. Thus, the value in Year 27 (V21) of the constant, infinite
stream of benefits (B) expected to accrue from that year forward is calculated as:
V21 = B/r
In calculating the present value of reduced contamination of private wells, EPA sets the value of B21
equal to that of V21. The present value of benefits is then determined using the following equation:
PV=- 5,7(1+/•)'
8B-1
-------
Appendix 8-C
CALCULATION OF ANNUALIZED BENEFITS
The constant annual benefit A that, over a period of n years, equals the estimated present
value (PV) of benefits is determined by the following equation:
,4 =PV(r)/(!-[!/(!+/•)"])
where r represents the annual discount rate. As n approaches infinity, this equation simplifies to:
A = PV(r)
EPA uses the equation above to calculate the annualized benefits reported in this analysis.
8C-1
------- |